Download as pdf or txt
Download as pdf or txt
You are on page 1of 75

Atmospheric Environment 43 (2009) 5193–5267

Contents lists available at ScienceDirect

Atmospheric Environment
journal homepage: www.elsevier.com/locate/atmosenv

Review

Atmospheric composition change: Ecosystems–Atmosphere interactions


D. Fowler a, *, K. Pilegaard b, M.A. Sutton a, P. Ambus b, M. Raivonen c, J. Duyzer d, D. Simpson e, f,
H. Fagerli f, S. Fuzzi g, J.K. Schjoerring h, C. Granier i, j, k, A. Neftel l, I.S.A. Isaksen m, n, P. Laj o, p,
M. Maione q, P.S. Monks r, J. Burkhardt s, U. Daemmgen t, J. Neirynck u, E. Personne v,
R. Wichink-Kruit w, K. Butterbach-Bahl x, C. Flechard y, J.P. Tuovinen z, M. Coyle a, G. Gerosa aa,
B. Loubet v, N. Altimir c, L. Gruenhage ab, C. Ammann l, S. Cieslik ac, E. Paoletti ad, T.N. Mikkelsen b,
H. Ro-Poulsen ae, P. Cellier v, J.N. Cape a, L. Horváth af, F. Loreto ag, Ü. Niinemets ah, P.I. Palmer ai,
J. Rinne aj, P. Misztal a, E. Nemitz a, D. Nilsson ak, S. Pryor al, M.W. Gallagher am, T. Vesala aj,
U. Skiba a, N. Brüggemann x, S. Zechmeister-Boltenstern an, J. Williams ao, C. O’Dowd ap,
M.C. Facchini g, G. de Leeuw aq, A. Flossman o, N. Chaumerliac o, J.W. Erisman ar
a
Centre for Ecology and Hydrology, EH26 0QB Penicuik Midlothian, UK
b
Risø National Laboratory, Technical University of Denmark, 4000 Roskilde, Denmark
c
Department of Forest Ecology, University of Helsinki, 00014 Helsinki, Finland
d
TNO Institute of Environmental Sciences, 3584 CB Utrecht, The Netherlands
e
Department Radio and Space Science, Chalmers University of Technology, 41296 Gothenburg, Sweden
f
Norwegian Meteorological Institute, 0313 Oslo, Norway
g
Istituto di Scienze dell’Atmosfera e del Clima – CNR, 40129 Bologna, Italy
h
Royal and Veterinary and Agricultural University, 1870 Frederiksberg C, Denmark
i
UPMC Univ. Paris 06, LATMOS-IPSL; CNRS/INSU, LATMOS-IPSL, 75005 Paris, France
j
NOAA Earth System Research Laboratory, 80305-3337 Boulder, USA
k
Cooperative Institute for Research in Environmental Sciences, University of Colorado, 80309-0216 Boulder, USA
l
Agroscope FAL Reckenholz, Swiss Federal Research Station for Agroecology and Agriculture, 8046 Zurich, Switzerland
m
Department of Geosciences, University of Oslo, Inst. For Geologibygningen, 0371 OSLO, Norway
n
Center for International Climate and Environmental Research – Oslo (CICERO), 0349 Oslo, Norway
o
Laboratoire de Météorologie Physique, Observatoire de Physique du Globe de Clermont-Ferrand, Université Blaise Pascal – CNRS, 63177 Aubière, France
p
Laboratoire de Glaciologie et Géophysique de l’Environnement, Observatoire des Sciences de l’Université de Grenoble, Université
J. Fourier – CNRS, 38400 Saint Martin d’Heres, France
q
Universita’ di Urbino, Istituto di Scienze Chimiche ‘‘F. Bruner’’, 61029 Urbino, Italy
r
Department of Chemistry, University of Leicester, Leicester LE1 7RH, UK
s
University of Bonn, Institute of Crop Science and Resource Conservation – Plant Nutrition, 53115 Bonn, Germany
t
Bundesforschungsanstalt für Landwirtschaft (FAL) Institut für Agrarökologie, 38116 Braunschweig, Germany
u
Research Institute for Nature and Forest, 9500 Geraardsbergen, Belgium
v
INRA, INA PG, UMR Environm & Grandes Cultures, F-78850 Thiverval Grignon, France
w
Department of Meteorology and Air Quality, Wageningen University and Research Centre, 6700 AA Wageningen, The Netherlands
x
Institute for Meteorology and Climate Research, Atmospheric Environmental Research (IMK-IFU), Forschungszentrum Karlsruhe GmbH,
82467 Garmisch-Partenkirchen, Germany
y
Soils, Agronomy and Spatialization (SAS) Unit INRA, 35042 Rennes, France
z
Finnish Meteorological Institute, 00560 Helsinki, Finland
aa
Dipartimento di Matematica e Fisica ‘‘Niccolò Tartaglia’’, Università Cattolica del Sacro Cuore, 25121 Brescia, Italy
ab
Institute for Plant Ecology, Justus-Liebig-University of Giessen, 35392 Giessen, Germany
ac
Institute for Environment and Sustainability, The European Commission, Joint Research Centre, 21020 Ispra, Italy
ad
Istituto per la Protezione delle Piante – CNR, 50019 Sesto Fiorentino, Italy
ae
Botanical Institute, University of Copenhagen, 1353 Copenhagen K, Denmark
af
Hungarian Meteorological Service, 1675 Budapest, Hungary
ag
Istituto di Biologia Agroambientale e Forestale – CNR, 00015 Monterotondo Scalo, Italy
ah
Institute of Agricultural and Environmental Sciences, Estonian University of Life Sciences, 51014 Tartu, Estonia
ai
School of GeoSciences, University of Edinburgh, EH9 3JN Edinburgh, UK
aj
Department of Physical Sciences, University of Helsinki, 00014 Helsinki, Finland
ak
Department of Applied Environmental Science, Atmospher Science Unit, Stockholm University, 10691 Stockholm, Sweden
al
Atmospheric Science Program, Department of Geography, Indiana University, 47405-7100 Bloomington, USA
am
School of Earth, Atmospheric and Environmental Sciences, The University of Manchester, M13 9PL Manchester, UK

* Corresponding author. Tel.: þ44 (0) 131 445 4343; fax: þ44 (0) 131 445 3943.
E-mail address: dfo@ceh.ac.uk (D. Fowler)

1352-2310/$ – see front matter Ó 2009 Published by Elsevier Ltd.


doi:10.1016/j.atmosenv.2009.07.068
5194 D. Fowler et al. / Atmospheric Environment 43 (2009) 5193–5267

an
Department of Forest Ecology, Federal Research and Training Centre for Forests, Natural Hazards and Landscape, 1131 Vienna, Austria
ao
Max-Planck-Institut für Chemie, 55128 Mainz, Germany
ap
Department of Experimental Physics and Environmental Change Institute, National University of Ireland, Galway, Ireland
aq
Climate and Global Change Unit, Research and Development, Finnish Meteorological Institute, 00560 Helsinki, Finland
ar
Energy Research Centre of The Netherlands, 1755 ZG Petten, The Netherlands

a r t i c l e i n f o a b s t r a c t

Article history: Ecosystems and the atmosphere: This review describes the state of understanding the processes involved in
Received 29 January 2009 the exchange of trace gases and aerosols between the earth’s surface and the atmosphere. The gases
Received in revised form covered include NO, NO2, HONO, HNO3, NH3, SO2, DMS, Biogenic VOC, O3, CH4, N2O and particles in the size
27 July 2009
range 1 nm–10 mm including organic and inorganic chemical species. The main focus of the review is on the
Accepted 29 July 2009
exchange between terrestrial ecosystems, both managed and natural and the atmosphere, although some
new developments in ocean–atmosphere exchange are included. The material presented is biased towards
Keywords:
the last decade, but includes earlier work, where more recent developments are limited or absent.
Dry deposition
Trace gas fluxes New methodologies and instrumentation have enabled, if not driven technical advances in measure-
Resuspension ment. These developments have advanced the process understanding and upscaling of fluxes, especially
Biogenic emissions for particles, VOC and NH3. Examples of these applications include mass spectrometric methods, such as
Compensation points Aerosol Mass Spectrometry (AMS) adapted for field measurement of atmosphere–surface fluxes using
micrometeorological methods for chemically resolved aerosols. Also briefly described are some advances
in theory and techniques in micrometeorology.
For some of the compounds there have been paradigm shifts in approach and application of both tech-
niques and assessment. These include flux measurements over marine surfaces and urban areas using
micrometeorological methods and the up-scaling of flux measurements using aircraft and satellite remote
sensing. The application of a flux-based approach in assessment of O3 effects on vegetation at regional scales
is an important policy linked development secured through improved quantification of fluxes. The coupling
of monitoring, modelling and intensive flux measurement at a continental scale within the NitroEurope
network represents a quantum development in the application of research teams to address the under-
pinning science of reactive nitrogen in the cycling between ecosystems and the atmosphere in Europe.
Some important developments of the science have been applied to assist in addressing policy questions,
which have been the main driver of the research agenda, while other developments in understanding have
not been applied to their wider field especially in chemistry-transport models through deficiencies in
obtaining appropriate data to enable application or inertia within the modelling community. The paper
identifies applications, gaps and research questions that have remained intractable at least since 2000
within the specialized sections of the paper, and where possible these have been focussed on research
questions for the coming decade.
Ó 2009 Published by Elsevier Ltd.

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5196
1.1. Scale . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5197
1.2. Reactivity of natural surfaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5197
1.3. Frameworks for analysis and interpretation of trace gas and aerosol exchange . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5197
1.4. Bi-directional exchange . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5198
1.5. Aerosols . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5198
1.6. Ocean–atmosphere exchange . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5198
1.7. Wet deposition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5199
2. Reactive gaseous nitrogen compounds – oxidized nitrogen . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5199
2.1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5199
2.2. Emissions from soils . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5199
2.3. Emissions of NOy from plant surfaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5201
2.4. Canopy atmosphere interactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5201
2.5. Models and measurements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5201
2.6. Exchange of HNO3, HONO, PAN . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5202
2.7. NOx production and emission from snow surfaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5205
2.8. Up-scaling and regional and global trends . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5205
3. Biosphere atmosphere exchange of ammonia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5205
3.1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5205
3.2. Advances in measurement methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5206
3.3. Key controls on biosphere atmosphere exchange of ammonia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5208
3.4. Effects of ecosystem type on ammonia biosphere–atmosphere exchange . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5209
3.5. Modelling surface–atmosphere exchange of ammonia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5209
3.6. Dynamic simulation of ecosystem C–N cycling and ammonia fluxes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5210
3.7. Integrating ammonia exchange processes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5211
3.8. Future challenges for ammonia exchange . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5212
4. Sulphur dioxide . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .5212
4.1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5212
D. Fowler et al. / Atmospheric Environment 43 (2009) 5193–5267 5195

4.2. Worldwide advances in SO2 flux monitoring and modelling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5213


4.2.1. Asia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5213
4.2.1.1. Sulphur dioxide deposition to soils . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5213
4.2.1.2. Micrometeorological measurements over vegetated areas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5214
4.2.1.3. Long-term deposition studies and inferential modelling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5214
4.2.2. North America . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5214
4.2.3. Europe . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5215
4.2.3.1. Long-term flux monitoring in the UK . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5215
4.2.3.2. Other recent European datasets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5215
4.3. Control of surface uptake rates by leaf cuticular chemistry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5216
4.4. Advances in deposition modelling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5217
4.5. Future challenges . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5217
5. Ozone . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .5218
5.1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5218
5.2. Deposition rates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5219
5.2.1. European forests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5219
5.2.2. Crops . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5220
5.2.3. Grasslands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5221
5.2.4. Other vegetated surfaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5221
5.2.5. Non-vegetated surfaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5221
5.2.5.1. Snow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5221
5.2.5.2. Water . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5221
5.3. Non-stomatal deposition processes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5222
5.4. Model development and validation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5222
5.5. Risk assessment methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5223
5.6. Potential effects of climate change . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5224
6. Biogenic volatile organic compounds (BVOC) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5225
6.1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5225
6.1.1. Volatile isoprenoids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5225
6.1.2. Oxygenated volatile compounds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5225
6.2. Environmental controls on BVOC emissions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5225
6.2.1. Physiological and physico-chemical controls of emissions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5225
6.2.2. Physico-chemical controls of emission in species lacking specific storage structures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5226
6.2.3. Uptake and release of volatile compounds by vegetation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5226
6.2.4. CO2-Dependence of emissions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5227
6.2.5. Induced emissions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5227
6.3. Contemporary difficulties in scaling BVOC emissions from leaf to ecosystem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5227
6.4. BVOC fluxes over Europe, by compound and in relation to the needs of photochemical oxidant models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5227
6.4.1. Flux measurement techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5227
6.4.2. Isoprene . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5227
6.4.3. Monoterpenes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5228
6.4.4. Sesquiterpenes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5228
6.4.5. Methanol . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5228
6.4.6. Acetone and acetaldehyde . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5228
6.4.7. Other compounds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5228
6.5. The EU large field campaigns in the Mediterranean area: from BEMA to ACCENT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5228
6.6. Remote sensing of BVOC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5229
7. Deposition and resuspension of aerosols onto and from terrestrial surfaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5230
7.1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5230
7.2. Review of new measurement approaches and instrumentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5231
7.2.1. Flux measurements of particle numbers (size-resolved or total), without information on chemical composition . . . . . . . . . . . . . . . . . 5231
7.2.2. Flux measurements of individual aerosol chemical species . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5232
7.3. Area sources of particles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5232
7.3.1. Resuspension . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5232
7.3.2. Urban emissions of aerosols . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5232
7.4. Dry deposition of particles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5233
7.4.1. Dry deposition rates to vegetation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5233
7.4.1.1. Friction velocity (u*) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5234
7.4.1.2. Surface roughness length (z0) and canopy morphology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5234
7.4.1.3. Particle diameter (Dp) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5235
7.4.1.4. Atmospheric stability (z ¼ 1/L) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5236
7.4.2. Parameterising and modelling deposition rates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5236
7.4.3. Dry deposition rates to urban areas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5236
7.5. Uncertainties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5236
7.5.1. Uncertainties in the application of micrometeorological flux measurement techniques for deriving the local flux . . . . . . . . . . . . . . . 5236
7.5.2. Relating measured fluxes to surface exchange: flux divergence and the effect of chemical interactions . . . . . . . . . . . . . . . . . . . . . . . 5237
7.5.3. Interpretation of measurements for model verification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5238
7.6. Future research needs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5238
7.6.1. Deposition measurements and reporting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5238
7.6.1.1. Standardisation of eddy covariance approaches and data analysis procedures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5238
5196 D. Fowler et al. / Atmospheric Environment 43 (2009) 5193–5267

7.6.1.2. Improved measurements in the accumulation mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5238


7.6.1.3. Understanding the effect of stability and leaf properties on deposition velocities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5238
7.6.1.4. Filtering or accounting for chemical interactions and water uptake . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5239
7.6.2. Deposition models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5239
7.6.2.1. Migration to a probabilistic approach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5239
7.6.2.2. Improvement of modelling approaches . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5239
7.6.2.3. Impact of surface anisotropy on suspension & deposition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5239
7.7. Conclusions – aerosols . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5239
8. Ecosystem–atmosphere exchange of the radiatively active gases – N2O and CH4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5240
8.1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5240
8.2. Global budgets of N2O and CH4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5240
8.3. Biological sources of N2O and CH4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5240
8.3.1. The biology of production and consumption of N2O and CH4 in soils and sediments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5240
8.3.2. Distribution of active microbial populations in soils . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5241
8.3.3. N2O and CH4 fluxes from the main global ecosystems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5241
8.3.4. Plant-mediated transport and production of N2O and CH4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5241
8.3.4.1. Methane from vegetation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5241
8.3.4.2. Nitrous oxide from vegetation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5241
8.4. New developments in measurements of N2O and CH4 and denitrification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5242
8.4.1. Flux chambers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5242
8.4.2. Micrometeorological methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5242
8.4.3. Comparison of eddy covariance with chamber methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5242
8.4.4. Recent methodological advances in measurements of total denitrification rates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5243
8.5. Modelling of N2O and CH4 fluxes at site and regional scales: approaches, applications and uncertainties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5243
8.6. Validation of models by landscape and regional scale measurements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5244
8.7. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5244
9. Exchange of trace gases and aerosols over the oceans . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5245
9.1. New trace gas interactions at the air–sea interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5245
9.1.1. Case studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5245
9.1.1.1. Acetone (ocean uptake) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5245
9.1.1.2. Methanol (ocean uptake) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5246
9.1.1.3. Isoprene (ocean emission) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5247
9.1.1.4. Halogenated organics (ocean emission and uptake) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5247
9.1.1.5. Monoterpenes (ocean emission) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5247
9.1.1.6. Alkyl nitrates (ocean emission) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5247
9.2. Aerosols . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5247
9.2.1. Primary marine aerosol (PMA) source functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5247
9.2.2. Chemical composition of primary sea spray . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5248
9.2.3. Secondary aerosol production . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5250
10. The processes of wet scavenging of aerosols and trace gases from the atmosphere . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .5251
10.1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5251
10.2. Nucleation scavenging of drops and ice crystals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5251
10.3. Impaction scavenging of aerosol particles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5251
10.4. Scavenging of gases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5252
10.5. Clouds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5252
10.6. Orographic precipitation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5252
10.7. Organic N in air and rain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5253
10.8. Conclusions and some priority areas of future research . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5253
11. Ecosystem–atmosphere exchange – concluding remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5254
11.1. Policy needs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5254
11.2. Current understanding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5254
11.3. Future developments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5254
Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5255
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5255

1. Introduction exchange processes is a core activity in understanding the Earth


system. The subject of this review is much narrower than the
The composition of the earth’s atmosphere is unique in the scope of these opening lines, and is restricted to the trace gases
solar system in being largely determined by biological processes and aerosols exchanged between the atmosphere and the earth’s
in soils, vegetation and the oceans interacting with physical and surface. However, as is clear from much of the international
chemical processes within the atmosphere. The physical surface– assessment of changes in atmospheric composition since the
atmosphere exchange of most gases contributing major and trace industrial revolution, these trace atmospheric constituents are
constituents of the atmosphere is coupled to biological produc- changing the earth’s climate (IPCC, 2007), global biodiversity
tion processes and transferred through the surface–atmosphere (Millenium Ecosystem Assessment, 2005) and the biogeochemical
interface. Thus, developing a mechanistic understanding of the cycling of major nutrients including nitrogen, carbon, and
production and destruction processes and their interactions with sulphur.
D. Fowler et al. / Atmospheric Environment 43 (2009) 5193–5267 5197

The earth’s surface is a sink for some atmospheric trace gases the photochemical oxidants in the 1960s and 1970s (Husar et al.,
and aerosols, and a source for many others, and for most the 1978). These early studies were made to determine the importance of
surface–atmosphere interface represents a zone within which most surface removal which is better known as dry deposition (to distin-
of the overall control of fluxes occurs. An understanding of the rate guish the process from removal by precipitation) as a sink for reactive
controlling processes at this interface is therefore vital in describing trace gases (NO2, SO2, O3). Advances in understanding and computing
the exchange process and understanding the global biogeochemical resources have allowed more sophisticated approaches to be adop-
cycles. Applications of science in this field are necessary to quantify ted, in which the processes at the surface recognised different sinks
and model responses to human perturbation of many of the and interactions with other trace gases, allowing rates of dry depo-
biogeochemical cycles (C, N, S, halogens and metals to name but sition to change with time and with surface characteristics.
a few). These perturbations include changes in land use or emis-
sions of trace gases to the atmosphere, through combustion and 1.1. Scale
industrial activities. Taking as an example the global nitrogen cycle,
human activity through combustion processes for oxidized nitrogen Emission or deposition schemes to quantify trace gas fluxes
and the Haber Bosch process for reduced nitrogen now dominates operate at a range of scales depending on the applications, illustrated
the cycling of reactive nitrogen through the atmosphere and back to in Fig. 1.1. For hourly integration the application is primarily
terrestrial and marine ecosystems (Galloway et al., 2004). The total for research purposes and mechanistic study at the small scale
emission of reactive nitrogen (Nr) from human activities at the end (<102 m2). For landscape scale measurements and for assessment of
of the 20th century exceeds that from natural processes by a factor the fate of pollutants at the regional scale (106 km2) the application
of 4 (20.7 Tg of oxidized and reduced reactive nitrogen Nr from has both research and policy objectives. At this scale, spatial and
natural sources within a total of 104 Tg-N in 1993, Galloway et al., temporal integration provides robust parameterisation. The appli-
2004). As nitrogen is a limiting nutrient in many ecosystems, cation in global models to quantify sources and sinks is restricted in
these modifications of the natural cycling have profound effects on spatial resolution, typically to 1  1, and likewise has research and
ecosystem function, biodiversity and atmospheric composition policy application Dentener et al., 2006. For the landscape scale, flux
(Erisman et al., 2008). The human disturbance of the global carbon measurements may be made directly, using micrometeorological
cycle is also extensive, and the quantities involved are very large. methods above canopies of vegetation, soil, or even ocean surfaces
Global emissions of CO2 from fossil fuels since 1700 amount to and have become the method of choice for long-term flux
approximately 600 Gt-C, which have increased the atmospheric CO2 measurement. These techniques provide, in addition to the target
mixing ratio from 280 ppm to 380 ppm in 2006, an increase of about trace gas flux the turbulent exchange characteristics of the under-
30%, (IPCC, 2007). lying surface and the partitioning of available radient energy which
These high level indicators of human influence provide essential enables the processes to be investigated at a sufficiently large scale to
context for this review paper, but conceal the detailed changes integrate canopy scale fluxes over typically 105 m2 (Baldocchi et al.,
taking place and the range of chemical species and interactions 2001).
involved. The subject area includes many different chemical species, The sections focus mainly on individual trace gases or classes of
and it is not possible to be comprehensive in this review for all gases. atmospheric particles, and each considers the surface–atmosphere
In particular the subject of the global carbon cycle and CO2 in exchange over specific ecosystems. The exceptions are the ocean
particular are much too large to cover in this review. The focus of this surfaces and wet deposition, within which a range of relevant
review is on the reactive trace gases and for the greenhouse gases, compounds is considered.
CH4 and N2O. The gases are associated with a range of biological
sources and have varied chemical reactivity in the atmosphere and at 1.2. Reactivity of natural surfaces
surfaces. These differences reveal the range of controls and temporal
and spatial variability in rates of exchange, which are the focus of the For many of the short lived gases (<2 days in the boundary
review. The review moves through a range of chemical species, layer) there are multiple sinks at the surface and these include
identifying the current state of knowledge and, where possible foliar surfaces and soil whose properties as sinks for a range of
applications of the new developments in a policy context. gases vary with humidity and the presence of surface water and
The gases emitted from terrestrial and ocean ecosystems are influenced, sometimes strongly, by the presence of other gases
include all of the major greenhouse gases, H2O, CO2, CH4 and N2O, (Fig. 1.2). The chemical and physical complexity of terrestrial
the nitrogen gases (both in reduced and oxidized forms), sulphur surfaces, illustrated in Figs. 1.1 and 1.2 at the microscopic scale is
compounds, volatile organic compounds (VOC) and halogens. greatly simplified in the parameterisations used in models. The
Quantifying the fluxes of these trace atmospheric components is simplification is necessary in part due to the nature of the flux
a prerequisite within an assessmemnt process leading to the devel- measuring systems, which integrate the net fluxes over large areas
opment of policy on climate change, eutrophication, acidification of these surfaces, and fail to reveal the microscopic scale of
and photochemical oxidant formation. Many research groups have variability of the true exchange.
become involved in the measurement and modelling of emission and
uptake (deposition) fluxes of trace gases and particles. The mecha- 1.3. Frameworks for analysis and interpretation
nistic understanding has developed from two different fields of of trace gas and aerosol exchange
study, the first was concerned with the sources of atmospheric trace
constitiuents, and the greenhouse gases were among the first The measurements of surface–atmosphere exchange provide at
compounds for which surface fluxes were quantified directly by the simplest level the mass exchange per unit area of surface, which
field measurements. These included small scale (0.1 m–0.5 m2) may be ground, water or leaf area, per unit time. To extract useful
measurements of fluxes from soils and vegetation using chamber information on the underlying processes it is necessary to quantify
methods for CO2, CO, CH4, N2O (Junge, 1963). The measurements the contributions each step in the transfer pathway makes to the
showed large spatial and temporal variability so that up-scaling to overall exchange between defined points, which in this scheme is
regions generated very large uncertainties. The other development simplified to vertical levels between a source and a sink. The most
was mainly associated with the atmospheric transport and deposi- widely applied transfer scheme is a resistance analogue (Monteith
tion of pollutants, including nitrogen and sulphur compounds and and Unsworth, 2007), in which the flux of trace gas or particle
5198 D. Fowler et al. / Atmospheric Environment 43 (2009) 5193–5267

Fig. 1.1. A diagrammatic representation of the scales of measurement of trace gas fluxes for process studies (a transverse section through a Phaseolus vulgaris leaf, showing the
palisade and mesophyll cells bounded by epidermal cells and the airspaces for internal exchange between trace gases and intercellular fluid). The field scale at which most of the
micrometeorological flux measurements are made and the continental scale where models provide the emission and deposition fluxes. In this case the emission fluxes of oxides of
nitrogen over Europe are shown, revealing the importance of international shipping to emissions over continental scales.

is treated as an analogue of electrical current flowing through 1.4. Bi-directional exchange


a simple network of resistances, which may act in series if there is
just one sink at the surface, such as a pure water surface, or may For many of the trace gases, regardless of their reactivity, the
have several sinks at the surface, acting in parallel, each repre- exchange fluxes may vary in sign as well as magnitude, with
senting a distinct chemical component of the underlying surface. emission and deposition being commonly observed. The most
A simple resistance network representing three different sinks widely known example of bi-directional exchange is CO2, which
at the surface, and the two atmospheric resistances (Ra and exhibits both deposition and emission fluxes due to photosynthesis
Rb, respectively the turbulent transfer resistance and the leaf and respiration respectively. In this case the concept of compen-
boundary-layer resistance) are illustrated in Fig. 1.3. sation points as mixing ratios at which no net exchange takes place
The atmospheric resistances may be separated from the total is now widely recognised for a range of trace gases (NH3, NO, CO2)
resistance using independent measures of the turbulence above the but all controlled by very different processes.
vegetation. The overall flux may be measured by a variety of micro- The recognition of bi-directional exchange requires modelling
meteorological methods (Monteith and Unsworth, 2007), and thus approaches to simulate the process for application in surface–
the total of the surface or canopy resistances to transfer between the atmosphere exchange schemes, as illustrated for NH3 in Fig. 1.4.
source and sink may be quantified as the residual, as shown in Fig. 1.3.
1.5. Aerosols

The understanding of deposition and emissions of aerosols over


terrestrial surfaces has advanced considerably in the last decade,
after a long period in which application of a model developed by
Slinn (1982) has been a standard for many modelling approaches.
Likewise, the emission of aerosols by resuspension from terrestrial
surfaces has advanced following innovative new measurement
approaches described in this review.

1.6. Ocean–atmosphere exchange

For many years the ocean–atmosphere exchange of trace gases has


been treated in a simplistic way (Liss et al., 1981), in part due to the
relative simplicity of the ocean surface relative to terrestrial surfaces,
but also due to the limited knowledge base to support more complex
treatments. However there has been an accelerating interest in ocean–
Fig. 1.2. Illustrating the importance of different sinks for reaction of trace gases at the atmosphere exchange as new techniques have become available
terrestrial surfaces, notably the external surfaces of vegetation often as in this case
covered by complex layers of epicuticular wax and illustrated in Fig. 1.1, the internal
to make the flux measurements and as very new issues have been
structure of leaves, following uptake through the stomatal apertures and soils greatly identified. Current interest in ocean acidification and ocean eutro-
simplified in this illustration. phication further raise the profile of ocean–atmosphere exchange, and
D. Fowler et al. / Atmospheric Environment 43 (2009) 5193–5267 5199

deposition footprint of pollutants (e.g. with reference to acidification,


χO3 (z-d) eutrophication, metal deposition or photochemical oxidants). These
Ra applications require an understanding of the fate and transformations
of the emitted trace gases within the atmosphere. The developments
Rb O3 in understanding atmospheric chemistry of the trace gases discussed
R = R a + Rb + R c in this paper are reviewed in the companion paper (Monks et al., this
FO3 t
issue) in this volume.
R c2 1
Rc3 cuticle R c1 Rc =
stomata 1 1 1 2. Reactive gaseous nitrogen compounds – oxidized nitrogen
soil + +
R c1 R c2 R c3
2.1. Introduction

χO3 (z 0' ) = 0 Developments in understanding surface–atmosphere exchanges


of NO and NO2 over the last decade have focussed on three specific
issues: the long-term emission of NO from soils; the interaction of
Fig. 1.3. A simple resistance analogy for a trace gas with sinks in stomata, on foliar chemical processing of nitrogen oxides in and above plant canopies;
surfaces and in soil.
and the deposition of NO2 and HNO3 to foliar and soil surfaces. The
measurements have been made over different vegetation, but the
given that these surfaces cover 71% of the earth’s surface, the relatively recent focus has been on forests, in part because the interactions
small section of this review paper on this topic belies its importance in between these processes are greatest for forests, but also because
understanding atmospheric composition change. some of the measurements are simpler to make and interpret for
mature forests. This section outlines the developments in under-
1.7. Wet deposition standing NOy exchange between terrestrial ecosystems and the
atmosphere, concentrating on developments during the last decade.
Process understanding in the scavenging of gases and particles
by precipitation has continued to advance, with important devel- 2.2. Emissions from soils
opments during the last decade. The applications of wet deposition
schemes are very important in the Long-Range Transport (LRT) Soil surface emissions of NO are the result of several biological
models and increasingly in global chemistry-transport models and abiotic processes in the soil producing and consuming NO.
(CTM) (Stevenson et al., 2006). These two applications make very Production and consumption of NO occurs predominantly via the
different demands on available knowledge and understanding. In biological nitrification and denitrification processes. Nitrification is
the case of LRT models in Europe (e.g. EMEP), the applications the oxidation of soil NHþ 
4 to NO3 , and denitrification is the anaerobic

form part of the integrated assessment process and within the user reduction of soil NO3 to N2O and N2. In nitrification, NO is formed as
community the pressure to provide ever finer spatial scale esti- a by-product during the oxidation of NHþ 
4 to NO2 and possibly also as

mates of inputs presents challenges in the capability of LRT models a result of nitrifier reduction of NO2 leading to an NO production of
and the meteorological models on which they depend. Current 1–4% of the NHþ 4 being oxidized (Skiba et al., 1997). The NO produced
demand for assessments of effects at the 1 km  1 km scale allows may be transformed within the soil profile by oxidation to NO 3 or it
the scale of the input estimate to approach the scale of an individual may be released to the atmosphere following diffusion to the
nature reserve, for example. soil surface. In denitrification, NO occurs as an intermediate in the
The focus of this paper on surface–atmosphere exchange processes cascade of reductive processes, and in the soil profile NO reduction
spans a wide range of trace gases and particles. The motives for studies may contribute to the formation of N2O. Abiotic production of
of many of the specific gases were environmental policy related, for NO occurs from oxidation of nitrous acid (HONO) that has been
example to determine the atmospheric lifetime, travel distance and produced by protonation of biologically formed NO 2 (Venterea et al.,

Fig. 1.4. A diagrammatic representation of bi-directional exchange, for NH3 exchange between the atmosphere and vegetation.
5200 D. Fowler et al. / Atmospheric Environment 43 (2009) 5193–5267

2005). Under certain conditions e.g. after application of anhydrous


ammonia to agricultural soils or acidic forest soils, the coupled
biological-abiotic production of NO may constitute the dominant
process for soil NO emissions (Venterea and Rolston, 2000; Gödde
and Conrad, 1998). Factors that increase nitrification and denitrifi-
cation, e.g. substrate and O2 availability, temperature and pH are
thus predicted to influence NO formation. Likewise, factors affecting
transport processes in the soil are predicted to regulate emissions of
NO (and other gases). It has been hypothesized (Davidson, 1991) that
where WFPS (water-filled pore space) is less than 0.6, nitrification is
the dominant process and relatively high emissions of NO may
be observed. Under more reducing conditions, 0.6 < WFPS < 0.9,
denitrification dominates which has a higher potential for NO
production compared to nitrification (Skiba et al., 1997); however
under conditions where anoxic conditions are generated by high soil
water content or by compaction of fine textured soil the probability
of NO being re-consumed by the denitrifying community is greatly
enhanced. Soil water may also play a central role in mediating
chemical processes leading to NO formation (Venterea et al., 2005).
Under most soil conditions, both nitrification and denitrification
occur simultaneously and the net flux of NO between soil and atmo-
sphere is the result of both processes together. As current views of
controls over NO gas emissions are still incomplete and need revision
e.g. with emphasis on the role of abiotic formation (Venterea et al.,
2005) there is a continuous need to further develop and improve
methodologies to identify and characterise the NO formation
processes. Gödde and Conrad (1998) achieved this by a combined
modelling and experimental approach to determine the net NO flux in
relation to NO concentration to quantify production and consumption
rate constants and compensation concentration. Recent advances in
methodological approaches to deepen our understanding of soil based
NO emissions have included application of stable isotope techniques.
Stark et al. (2002) applied a 15N-isotope pool dilution method-
to obtain the simultaneous gross rates of NO forming processes
combined with soil emissions, and Russow et al. (2000) adopted
a kinetic isotope method (KIM) to study the complex N transformation
processes involved in soil NO emissions.
Fig. 2.1. Left: NO emission (mg N m2 h1) as a function of nitrogen deposition
NO and N2O emissions were measured continuously at 15 forest (g N m2 s1). Regression lines (solid ¼ significant, dashed ¼ non significant) for
sites in Europe (Pilegaard et al., 2006) including coniferous and coniferous and deciduous sites, respectively. Right: N2O emission (mg N m2 h1) as
deciduous forests in different European climates, ranging from a function of C/N ratio. The full line represents a linear regression and the dotted line
boreal to temperate continental forests and from Atlantic to Medi- a logarithmic regression.

terranean forests. Furthermore the sites differ in atmospheric N-


deposition ranging from low deposition (0.2 g N m2 yr1) to high
deposition (4 g N m2 yr1). rate of O2 supply and thereby controls whether aerobic processes
The relationships of the emissions of NO and N2O, with the such as nitrification or anaerobic processes such as denitrification
parameters, nitrogen deposition, forest type, age, C/N in the surface dominate within the soil. While N2O emissions are known to
horizon, pH, soil temperature and water-filled pore space (WFPS) increase at higher water contents through larger losses from
were investigated by means of stepwise multiple regression anal- denitrification the relationship between the NO flux and the soil
ysis. NO emission was dependent on forest type and positively water is more complex. Due to limited substrate diffusion at very
correlated with nitrogen deposition (Fig. 2.1). WFPS was tested for low water content and limited gas diffusion at high water content,
curvature by including a quadratic term, but this was not signifi- nitric oxide emissions are suspected to have a maximum at low to
cant. Separately performed regression analyses for deciduous and medium soil water content.
coniferous forests showed that the relationship between nitrogen The effects of soil moisture and temperature on NO and N2O
deposition and NO emission was only significant for the coniferous emission were studied in laboratory experiment with soil cores
forests: (NO (mg N m2 h1) ¼ 13.9 þ 25.5 [N deposition from a range of field sites (Schindlbacher et al., 2004). Soil moisture
(mg N m2 h1)], r2 ¼ 0.82). The N2O emission was significantly and temperature explained most of the variability in NO emission
negatively correlated with both the C/N ratio and the age of the (up to 74%) and N2O (up to 86%) emissions for individual soils. NO
stands; a logarithmic transformation of N2O emission improved the and N2O were emitted from all soils except from a boreal pine forest
significance of the correlation. soil in Finland, where the laboratory experiment showed net NO
Soil temperature is a key variable affecting the emission rates of consumption. NO emissions from a German spruce forest ranged
both gases (Fig. 2.2). Emissions of both NO (Slemr and Seiler, 1984) from 1.3 to over 600 mg NO–N m2 h1 and greatly exceeded
and N2O (Skiba, 1998) increase with soil temperature due to the emissions from other soils. Average N2O emissions from this soil
positive effect of temperature on enzymic processes as long as tended also to be largest (170  40 mg N2O–N m2 h1), but did not
other factors (e.g. substrate or moisture) are not limiting. Soil water differ significantly from other soils. NO and N2O emissions showed
acts as a transport medium for NO þ
3 and NH4 and influences the a positive exponential relationship with soil temperature.
D. Fowler et al. / Atmospheric Environment 43 (2009) 5193–5267 5201

400 In general, relationships between nitrogen oxides emission and soil


moisture and soil temperature can be found within a single locality
when studying short-term variations. However, when comparing
NO emission [μg NO-N m-2 h-1]

annual values from different localities within a large region, parame-


300 terisations differ appreciably between locations and other factors such
as soil properties, stand age, and site hydrological conditions interfere.
This scale of variability has slowed the development of improved soil
NO emission inventories, and their application in global chemistry-
200
transport models. So while it is disappointing to observe current
generation CTMs using 1995 soil NO estimates, it is understandable
and identifies a need for better parameterisations and soil emission
100 databases for global application.

2.3. Emissions of NOy from plant surfaces

0 Production of NOy on Scots pine branch surfaces by ultraviolet


-5 0 5 10 15 20 radiation has been observed in Hyytiälä, southern Finland (Hari
Forest floor temperature [°C] et al., 2003) (Fig. 2.4). Other studies have shown that irradiance-
dependent NOy emissions from snow and different chamber surfaces
Fig. 2.2. Relationship between daily mean forest floor temperature and daily mean NO have been observed to originate from HNO3 or nitrate photolysis. In
emissions at the Höglwald Forest (spruce, control) for the observation period January 1,
Hyytiälä forest, Raivonen et al. (2006) investigated whether the NOy
2004–December 31, 2006. For details on measurement and site characteristics see
Gasche and Papen (1999). emitted from pine shoots could originate from photolysis of HNO3
attached to the needle surface. Field data of several years from
Hyytiälä were used to test this hypothesis. The HNO3 deposition,
estimated for the Hyytiälä site, has been high enough to account for
The results from the annual averages of field data did not show the NOy emission rates observed from the chambers. The particular
significant relationships with soil temperature for either NO or for characteristics of the daily pattern of CO2 exchange or stomatal
N2O emission. Schindlbacher et al. (2004) showed that N2O emis- control was not reflected in the NOy flux. When a pine branch
sions increased with increasing WFPS or decreasing water tension, was rinsed, which reduced the amount of water-soluble nitrogen
respectively. Maximum N2O emissions were measured between 80 compounds (e.g. HNO3, nitrates and HONO) from the needle surface,
and 95% WFPS or 0 kPa water tension. The optimal moisture for NOy emissions from that branch decreased compared to another
NO emission differed significantly between the soils, and ranged non-rinsed branch. Therefore, it was concluded that the results
between 15% WFPS in sandy Italian floodplain soil and 65% in loamy support the hypothesis and that HNO3 photolysis on plant surfaces
Austrian beech forest soils. For the field data WFPS was not needs to be taken into account both from air chemistry and plant
a significant parameter for N2O emission, but had a positive signif- sciences point of view.
icant effect on NO emission (Fig. 2.3). The annual average WFPS in
the field was higher than the optima found for NO in the laboratory 2.4. Canopy atmosphere interactions
experiment, but since not all field sites were studied in the labora-
tory it is difficult to provide a general conclusion. The inter-annual The interaction between chemical reactions of nitrogen oxides
variation within single sites clearly showed relationships to both taking place in the canopy and trunk space of a forest is a special
temperature and soil moisture. An important factor for N2O emis- case because in this area chemical and turbulent timescales change
sion is freeze-thaw events which can produce a significant outburst substantially leading to a very complex situation in which even the
of N2O (Kitzler et al., 2006). direction of fluxes may change (Duyzer et al., 1995) (Fig. 2.5).
This makes it nearly impossible to interpret measurements of the
turbulent fluxes of some reactive trace species above the canopy from
single point eddy covariance measurements. Several models have been
Sandy loam developed to simulate the overall exchange and show the magnitudes
100

Silty loam of the different competing processes. These models describe the
Sandy clay loam
Loam(1) coupled processes of atmospheric transport and chemical processes
NO emission (% of maximum)

Loam(2) above and in canopies in detail. Over forests the situation is even more
80

complex. Flux measurements are usually carried out near the top of
rough canopies leading to potential inaccuracies in the K-theory
60

approximation. This theory is relatively easy to combine with vertical


atmospheric transport phenomena with fast chemical reactions.
40

2.5. Models and measurements


20

Measurements of small fluxes of NO and NO2 have shown spurious


results especially at low concentrations due to a lack of specificity of
monitors and a lack of instrument sensitivity, but other problems may
0

well have contributed, including violation of conditions under which


0 20 40 60 80 100
such fluxes may be measured above canopies and the complexity of
WFPS(%)
interactions; soils and sunlight driven reactions may both be sources
Fig. 2.3. The relationship between NO emission and water-filled pore space at different of NOy and these interact with the stomatal sinks and the chemical
localities in the NOFRETE project (based on data in Schindlbacher et al., 2004). processing within the canopy trunk space Duyzer et al., 1983.
5202 D. Fowler et al. / Atmospheric Environment 43 (2009) 5193–5267

Fig. 2.4. Effect of UV radiation on the NOy concentration in a small Teflon chamber that enclosed a clean pine branch or a branch that had been treated with NH4NO3 solution. The
branches were dead and dry, cut from the tree. UV wavelengths were filtered away using a Plexiglas plate (Raivonen et al., 2006).

As a result of these limitations there are only limited data a compensation point above which the flux is directed towards the
available for verification of models. Duyzer et al. (2004) described surface and below which the flux is away from the surface.
the analyses of a dataset acquired in the framework of a European In summary the flux of NO2 above a forest can be described with
project from an experiment carried out in a 20 m high coniferous the following function:
forest (Speulderbos, The Netherlands). A 1D multi-layer model of  
a forest canopy was used to analyse the field data. In each layer 1
FNO2 ¼ f FNO;soil  CNO2 
vertical transport was described using K-theory; canopy uptake Rc;NO2
was described using a resistance layer model. Simple chemical
where all variables have their common meaning and CNO2 denotes
reactions between ozone and nitric oxide and photolysis of
the concentration of nitrogen dioxide above the canopy. This
nitrogen dioxide were described. The coupled differential equa-
equation is rather qualitative but indicates the sensitivity of the flux
tions were solved numerically. Input to the model calculations were
of NO2 above the canopy. More quantitative model runs are needed,
concentrations of nitrogen oxides and ozone at the highest level
but these require a large amount of input data and the results are
above the forest, levels of radiation, temperature, humidity, wind
still uncertain.
speed, turbulence parameters and an estimate of the emission of
nitric oxide. Output of the model is the concentration and fluxes of A simple resistance model (Duyzer et al., 2005a,b) was tested in
the relevant components at the height of each level in and above a deciduous forest (Sorø, Denmark) and is illustrated in Fig. 2.6.
the canopy. These may be compared with measured fluxes of these Generally the understanding of the various processes and their
components at two levels above and one below the canopy. It is fair interaction is increasing. Nevertheless many uncertainties
to say that the comparison between measured and modelled fluxes remain and there is a need for further improvement of models,
is not impressive. There are many possible explanations for this especially for Lagrangian models incorporating chemical reac-
observation but no clear single cause has been identified. tions. On the other hand, the accuracy of the results of field
Depending on the magnitude of this soil flux the NO2 flux is measurements has been rather low. It should be noted that
either downward or upward. In the case of the coniferous forest although the interaction between atmospheric chemical reac-
described here and the conditions during the experiment the result tions and exchange between the canopy and the atmosphere is
was that when the NO flux from the soil exceeded 10 ng m2 s1, easy to understand its importance may be limited. In cases where
the NO2 emission was upward (i.e. away from the forest). fluxes of nitrogen oxides are small the corrections could be large
At high concentrations the NO2 flux is directed towards the in a relative sense but still rather small in an absolute sense. The
forest and at small concentrations the flux is more likely to be currently available models could very well be capable of making
directed towards the free atmosphere. This may be interpreted as estimates of the magnitude of these effects. In view of all the
uncertainties hindering improved estimates in testing of models
the limited quality of the description of atmospheric transport
processes within the canopy may not be a serious problem here.

2.6. Exchange of HNO3, HONO, PAN

The deposition of HNO3 to terrestrial surfaces has been shown to


be primarily controlled by the rates of turbulent exchange in the
atmospheric boundary layer and the leaf boundary layer (Huebert
and Robert, 1985). The highly reactive and soluble nature of gaseous
HNO3 leads to large rates of deposition, approaching the maximum
rates of deposition limited by turbulent exchange when each
molecule arriving at terrestrial surfaces is immediately absorbed at
the surface. In these conditions the surface is considered to be
acting as a perfect sink, canopy resistance is zero and the numerical
value for the deposition velocity becomes:
Fig. 2.5. A schematic of the various canopy interactions involved in the exchange of
VgðNHO3 Þ ¼ Vmax ¼ 1=ra þ rb
nitrogen oxides between the free atmosphere and forests.
D. Fowler et al. / Atmospheric Environment 43 (2009) 5193–5267 5203

Fig. 2.6. Profiles of CO2, NO2, O3 and NO in a beech forest near Sorø, Denmark. The profiles clearly show the effects of stable conditions during night and daytime turbulence mixing
the full air column. CO2 is built up during night due to soil respiration; O3 is depleted during night due to deposition and chemical reactions. At the soil surface high concentrations
of NO are seen due to emission from the soil.

The values for deposition velocity in these conditions are very US pine forests at Duke Forest, North Carolina, (RH > 75%) and
sensitive to wind velocity values and approach several cm s1 even Blodgett Forest, California, (RH < 30%) (Farmer et al., 2006; Turn-
over relatively short vegetation. The consequence of these large rates ipseed et al., 2006; Wolfe et al., 2008).
of deposition are that even in areas with small HNO3 concentrations, At Duke Forest fluxes of PAN, PPN and MPAN were measured with
dry deposition contributes a substantial quantity if nitrogen. Taking an a CIMS technique (Turnipseed et al., 2006). There were no significant
ambient concentration of 0.1 ppb HNO3, the annual deposition of N for differences in the Vd of the three different PAN compounds, but all
a forest would be of the order 3 kg N ha1 annually from HNO3 alone. three species deposited about four times faster than predicted by the
The close coupling between rates of turbulent exchange and model of Wesely (1989) during the day, and nearly an order of
dry deposition rates for HNO3 also generates substantial spatial magnitude faster during the night, indicating that aqueous solubility
variability in N deposition in the landscape, with hotspots for N considerations are insufficient to predict the behaviour of PAN on
deposition being forests and especially forest edges, hedgerows and surfaces. The average Vd was 2.5 mm s1 during day and 8 mm s1
isolated, exposed hills, where wind speeds are larger. during night. In contrast to the considerations of Wesely (1989), wet
Several studies have recently attempted to measure total surfaces showed a smaller non-stomatal resistance (Rns ¼ 125 s m1)
oxidized nitrogen (NOy) fluxes or even total reactive nitrogen than dry surfaces (Rns ¼ 250 s m1).
(Nr ¼ NOy þ NHx) to ecosystems (Turnipseed et al., 2006). These At the much drier Blodgett forest site, the flux of the sum of all
approaches offer the prospect to apply eddy covariance techniques PANs was measured by TD-LIF, based on thermal conversion and NO2
for the robust and relatively cost effective determination of total detection (Farmer et al., 2006). PAN was derived as the difference
atmospheric N deposition, but they do not provide the chemical between the ambient temperature and 180  C channel. They found
speciation needed to further process understanding. There have, upward fluxes in summer and on average bi-directional exchange
however, also been advances in the understanding of individual N with afternoon deposition in winter, when noon-time deposition
compounds other than NH3, NO and NO2. velocities averaged 8 mm s1. More recently, these measurements
A recent lab study (Sparks et al., 2003) has confirmed that PAN have been repeated with the more selective CIMS technique (Wolfe
deposition through the stomata can make a significant contribution et al., 2008). Here measurements indicated larger average midday
to plant uptake of atmospheric N. In addition, recent instrument values of Vd for PPN (12 mm s1) than for the PAN and MPAN
developments in chemical ionisation mass spectroscopy (CIMS) (4 mm s1), while both compounds deposited slowly at night
and thermal-dissociation laser induced fluorescence (TD-LIF) have (Vd < 2 mm s1). The authors of this study attribute the difference in
enabled the application of eddy covariance to the biosphere/ the Vd between compounds to MPAN and PAN production inside the
atmosphere exchange of preoxy acyl nitrates (PANs such as PAN, canopy and suggest that the PPN fluxes are a better descriptor of the
PPN and MPAN). Measurements were made over two contrasting surface deposition. They suggest that the non-stomatal uptake is
5204 D. Fowler et al. / Atmospheric Environment 43 (2009) 5193–5267

dominated, but not fully explained, by thermochemical decompo- governed by the thermodynamic equilibrium with NH4NO3 on leaf
sition, and thus strongly linked to canopy temperature. surfaces or fertilizer pellets. Fig. 2.7 shows an example of reduced
In summary, this recent measurement evidence suggests that deposition of HNO3 and HCl over a Dutch heathland.
deposition rates of PANs in warm climates are at least a factor of 5 The view that HNO3 normally deposits with a near zero canopy
larger than predicted by commonly used models and non-stomatal resistance still holds. There is an increasing measurement database
deposition is larger to wet and humid surfaces than to dry surfaces. of HNO3 concentrations in national and regional networks suitable
During the same TD-LIF study, Farmer et al. (2006) measured for inferential modelling of HNO3 deposition (Tang et al., 2009),
fluxes of total alkyl nitrates (gas and aerosol phase), from the differ- which now provides independent confirmation from the model
ence between the 180  C and 330  C channels. These compounds results, that HNO3 deposition makes a very significant contribution
showed large winter-time midday deposition velocities of 20 mm s1, to nitrogen deposition across Europe. In addition, Europe-wide
approaching those of HNO3 (25 mm s1). Even higher Vd of 30 mm s1 monitoring activities have produced the first hourly monitoring
was derived by Horii et al. (2005) for what they interpret as isoprene- datasets of HNO3, which allows for a much more in-depth assess-
derived hydroxyalkyl nitrates. ment of the performance of oxidized nitrate chemistry in atmo-
Nitric acid (HNO3) has traditionally been believed to deposit at spheric transport models (Tarrason and Nyiri, 2008).
the maximum rate possible according to turbulence (Vmax) and its Much development has occurred in measurement techniques
flux measurement continues to be used to derive quasi-laminar for nitrous acid (HONO), e.g. based on long path absorption
boundary-layer resistances for vegetation (e.g. Pryor and Klemm, photometry (LOPAP) and differential optical absorption spectrom-
2004). This view has been challenged by recent measurements etry (DOAS). This has contributed to the improvement of process
that indicated non-negligible canopy resistances in the range of 50 understanding of sources of HONO in the atmosphere, e.g. revealing
to >200 s m1 during midday (Nemitz et al., 2004b, 2008). This larger daytime sources than previously thought and identifying
has been attributed to non-zero chemical compensation points NO2 reactions with humic acid as a novel production mechanism

10

0
Fχ [ng m s ]
-2 -1

-10

-20 HNO3 HCl

-30

-40
15:00 18:00 21:00 00:00 03:00 06:00 09:00 12:00 15:00 18:00 21:00 00:00
30
25 Vmax(HNO3)
Vd [ mm s ]

Vmax(HCl)
-1

20
Vd(HNO3)
15
Vd(HCl)
10
5
0
15:00 18:00 21:00 00:00 03:00 06:00 09:00 12:00 15:00 18:00 21:00 00:00
500

400
Rc [s m ]
-1

300
HNO3
200
HCl
100

0
15:00 18:00 21:00 00:00 03:00 06:00 09:00 12:00 15:00 18:00 21:00 00:00
0.5 120

0.4 RH u* 100
RH (z0') [%]
u* [m s ]

80
-1

0.3
60
0.2
40
0.1 20
0.0 0
15:00 18:00 21:00 00:00 03:00 06:00 09:00 12:00 15:00 18:00 21:00 00:00
GMT

Fig. 2.7. A time series of HNO3 and HCl exchange measured above a Dutch heathland with a denuder gradient system with online analysis by ion chromatography. The panels show:
(a) fluxes, (b) deposition velocities of HNO3 and HCl in comparison with their maximum values and (c) Rc for HNO3 and HCl, (d) friction velocity (u*) and relative humidity (h). Data
represent 2.5 h running means of 30 min. Vd(HNO3) and Vd(HCl) are reduced compared with their maximum values, presumably due to non-zero chemical compensation points
originating from deposited NHþ 4 salts. From Nemitz et al. (2004a).
D. Fowler et al. / Atmospheric Environment 43 (2009) 5193–5267 5205

(Kleffmann et al., 2005; Stemmler et al., 2006). By contrast, appli- based on intensity of rainfall) to estimate soil NO emissions.
cations of these approaches to flux gradient measurements are still Yienger and Levy also provide a canopy reduction factor to include
rare, which nevertheless confirm surface sources of HONO (Vitel estimates of the chemical conversion and re-deposition of NO as
et al., 2002; Kleffmann et al., 2003). NO2 within the canopy. Compared to the methodology by Yienger
and Levy, the Skiba-EMEP/CORINAIR approach is more simplistic.
2.7. NOx production and emission from snow surfaces Based on a literature review by Skiba et al. (1997) this approach
postulates that 0.3% of any form of nitrogen is volatilized as NO, i.e.
Snow lying on the Earth’s surface has traditionally been viewed regardless whether it originates from inorganic or organic fertil-
as a chemically inert medium, whose influence on the overlying ization or atmospheric N deposition. Furthermore, a background
atmosphere was exerted through its albedo effect, and by emission of 0.1 ng NO–N m2 s1 (z0.032 kg NO–N ha1 a1) was
restricting exchange of gases between the air and land/sea surfaces. assumed (Simpson et al., 1999). In addition, EMEP/CORINAIR also
The a priori view was that the boundary layer and troposphere over use a more detailed methodology (BEIS-2), which originates from
Antarctica would be somewhat uninteresting, with low concen- the work of Novak and Pierce (1993) and considers soil temperature
trations of reactive radicals such as OH, HO2, NO and NO2, and as well as different land use classes. A statistical summary model
a composition dominated by longer-lived chemical species. The was developed by Stehfest and Bouwman (2006), which is based
equivalent regions of the Arctic atmosphere were assumed domi- on the most extensive literature review currently available.
nated by long-range transport of anthropogenic emissions from This methodology for calculating soil NO emissions on global and
lower latitudes. However, recent research has shown that this regional scales considers land-use, N fertilization rate [Fertilizer],
picture is far from the truth, and that snow is a highly photo- soil N content (three different classes, estimated as 1:10 of soil
chemically active medium. Snow-pack impurities, of which there organic carbon content) [SON] and climatic regions. The method-
are many, can be photolysed to release reactive trace gases to the ology was recently adapted to calculate a European wide inventory
atmosphere. These processes are likely to be active anywhere that of NO emissions from forest soils (Kesik et al., 2006). Kesik et al.
sunlight irradiates snow. The importance of these processes to used the process-oriented ecosystem model Forest-DNDC. The
boundary layer composition varies with geographical location; in model was extensively tested for its performance to predict NO
regions with a high background of radicals, for example arising emissions at the various NOFRETETE field sites, which were located
from anthropogenic pollution, emissions from snow are of lesser across Europe and, thus, were covering different climatic condi-
importance. But in the remote polar regions, emissions from snow tions (Pilegaard et al., 2006). Regionalisation was finally achieved
can be the dominant source of reactive trace gases and have a major by linking the model to a detailed GIS database holding all relevant
influence on boundary layer chemical composition. This conclusion information for initializing and driving the model such as data on
was first reached for NOx (NO þ NO2), which was measured in vegetation (e.g. forest type) and soil properties (e.g. texture, soil pH,
the boundary layer at Summit, Greenland at surprisingly high organic C content) and climate (either present day conditions or
concentrations, and with a ratio to NOy that suggested a local projected future climate predictions). This approach demonstrated
source. Measurements of NOx within the snow-pack interstitial air for the first time the huge regional differences in NO emissions
revealed concentrations that were higher still, suggesting that the from forest soils across Europe as shown in Fig. 2.8, to estimate its
snow-pack itself was the source, with a gradient to the atmosphere. significance on a regional scale and to unravel the importance of
Subsequent measurements made in Antarctica confirmed that atmospheric N deposition for the magnitude of forest soil NO
NOx production within the snow-pack was a feature of both emissions.
polar regions (Jones et al., 2000). Additional measurements
confirmed that NOx generated within the snow-pack was released 3. Biosphere atmosphere exchange of ammonia
to the overlying boundary layer (Jones et al., 2001), contributing
to the higher than expected NOx concentrations that were 3.1. Introduction
encountered.
The air chemistry of polar regions is a rapidly developing field Substantial progress has been made during the last five years in
and extends substantially beyond interest in oxidized nitrogen, in understanding ammonia biosphere–atmosphere exchange. Experi-
which surface processes clearly provide the major source of reac- mental studies have included controlled laboratory analysis, while
tive oxidized nitrogen in Antarctic regions. Discussion of other trace a series of micrometeorological studies have assessed net fluxes
gases, including halogen and mercury compounds are beyond the occurring under field conditions. In particular, major advances have
scope of this review, however a valuable review of polar halogen been made in modelling the different aspects of ammonia exchange.
chemistry and links to oxidized nitrogen chemistry is provided by This has included not just analysis of the drivers of the vertical flux
Simpson et al. (2007). densities, but also a consideration of non-stationarities, such as
advection effects and chemical interactions. Traditionally, micro-
2.8. Up-scaling and regional and global trends meteorological experiments were designed to avoid these effects,
focusing as far as possible on ‘ideal’ micrometeorological conditions,
The complexity of processes involved in NO emissions from soils so as to better quantify the vertical exchange processes, and develop
has resulted in a significant uncertainty in the regional and global parametrisations for ‘dry deposition schemes’ in regional models
source strength of soils for NO. However, different methodologies (Fowler and Duyzer, 1989; Fowler et al., 1998; Sutton et al., 1994;
have been developed, e.g. relatively simple statistical models as Simpson et al., 2006). However, for ammonia, it has become
well as process-based model approaches, to cope with the problem increasingly clear that these non-stationarities represent important
of regionalisation of soil NO fluxes. The most widely used approach effects that are widespread in the real environment and need to be
to calculate regional NO emissions from soils is based on the work quantified (Sutton et al., 2007, 2008c).
of Yienger and Levy (1995). These authors consider land use and The developments in the last decade have arisen from a wide range
respective background emission strengths, nitrogen fertilization of national and international projects. National studies have particu-
rate (2.5% loss of applied nitrogen), temperature effects (three larly addressed exchange with key ecosystems of regional importance,
classes: cold-linear, exponential and optima) as well as the pulsing such as ammonia losses from agricultural systems (e.g. Milford et al.,
of NO emissions following prolonged dry periods (four classes, 2001a; Walker et al., 2006; Wichink Kruit et al., 2007) and the
5206 D. Fowler et al. / Atmospheric Environment 43 (2009) 5193–5267

were complemented by the LIFE project, which added long-term


ammonia flux data for a number of grassland, moorland and forest
ecosystems (e.g. Flechard and Fowler, 1998; Erisman et al., 2001;
Spindler et al., 2001).
As understanding of ammonia exchange has improved and
scientific ambition developed, increased attention has been given to
integrating the different drivers of ammonia exchange processes. This
has, for example, been reflected in the Braunschweig Integrated
Experiment of GRAMINAE (e.g. Sutton et al., 2002, 2008a, 2009a,b),
which linked a wide range of biospheric, atmospheric and manage-
ment interactions as these control ammonia exchange with managed
grassland. This integration has developed substantially under the
NitroEurope Integrated Project (Sutton et al., 2007), in which inter-
actions between the different components of nitrogen fluxes,
including ammonia and oxidized nitrogen and their effect on the net
greenhouse gas balance have been investigated. In parallel, major
advances have been made in spatial modelling of ammonia fluxes,
from individual forest edges to global scales (Theobald et al., 2004;
Dentener and Crutzen, 1994; Hertel et al., 2006; Bleeker et al., 2006).

3.2. Advances in measurement methods

Before considering the developments outlined above in more


detail, it is important to highlight that the advances have been
critically dependent on improvements in measurement technology
(See Table 1). At the start of the 1990s, ammonia flux measure-
ments were still being made using wet chemistry and manual batch
sampling with time integration of typically 2 h (e.g. Sutton et al.,
1993a; Duyzer, 1994). The most important advance was the intro-
duction of continuous wet chemistry methods for measuring
ammonia profiles, including the AMANDA wet rotating denuder
(Wyers et al., 1993) and the mini-Wet Effluent Diffusion Denuder
(e.g. Blatter et al., 1993; Neftel et al., 1999). Although these tech-
Fig. 2.8. Importance of atmospheric N deposition for NO emissions from forests soils. The niques are liable to malfunction, with effort and careful operation
map shows the difference in NO emissions between a scenario with zero atmospheric
they have produced many key datasets over the last 15 years
N deposition and present day atmospheric N deposition. In large parts of central Europe
but also Scandinavia forest NO emissions are likely to decrease significantly if atmo- (e.g. Erisman and Wyers, 1993; Sutton et al., 1995, 1997; Fowler
spheric N deposition can be reduced to background levels. et al., 1998; Flechard and Fowler, 1998; Neftel et al., 1998; Milford
et al., 2001a,b) and still represent the state-of-the-art as regards
precise measurement of small ammonia fluxes (Wichink Kruit et al.,
ammonia inputs into semi-natural ecosystems of conservation value 2007; Neirynck and Ceulemans, 2008; Sutton et al., 2008b).
(e.g. Wyers and Erisman, 1998; Neirynck and Ceulemans, 2008). A unique inter-comparison of four continuous wet chemical
Collaborative international projects have sought to integrate and systems was made at the GRAMINAE Braunschweig Experiment
extend these interests, making the comparison between ecosystem (Sutton et al., 2007, 2009b; Milford et al., 2009), which highlights
types and looking at the interactions (e.g. Sutton et al., 2009a). the potential and limitations of the approach. Fig. 3.1 shows the
The first European collaborative project dedicated to ammonia ammonia flux measured before and after cutting an agricultural
exchange was ‘EXAMINE’. Attention was given to quantifying grassland, and following subsequent fertilization with calcium
ammonia exchange with a range of European ecosystems, under both ammonium nitrate. The measurement systems were able to detect
experimental and field conditions (e.g. Sutton et al., 1995; Schjoerring the wide range of ammonia fluxes, but the degree of agreement
et al., 1998; Neftel et al., 1998; Meixner et al., 1996; Nemitz et al., varied greatly between days. This was a result of varying perfor-
2009b), including analysis of the surface gas-particle interactions mance of the different analysers, highlighting the need for highly
between ammonia, nitric acid and hydrochloric acid (Nemitz and intensive instrument maintenance.
Sutton, 2004). As part of EXAMINE a major collaborative analysis was Despite the improvements that have been made in the automation
made in the North Berwick experiment, which provided a uniquely and reliability of the continuous wet chemical gradient methods (e.g.
detailed examination of the processes controlling ammonia exchange Wichink Kruit et al., 2007; Flechard et al., 2007; Sutton et al., 2007),
with an oilseed rape canopy (Husted et al., 2000; Nemitz et al., their remain several limitations, which have encouraged researchers
2000a,c; Sutton et al., 2000a,b). to seek alternative ammonia flux measurement approaches. In prin-
In the second major European collaboration dedicated to ciple, many refinements have allowed the automated wet chemical
ammonia, the GRAMINAE project analyzed the processes controlling methods to become more reliable and comprehensive (such as being
ammonia exchange with grassland ecosystems across Europe (Sut- able to measure aerosol and acid gas gradients simultaneously, Trebs
ton et al., 2001). This included assessment of both the bi-directional et al., 2006). However, the use of many moving parts can be consid-
fluxes of ammonia with agricultural grassland – as these affect ered as inherently liable to faults. Similarly, the response times of
atmospheric ammonia balance (e.g. Milford et al., 2001a; Mosquera these instruments are typically >5 min, which means that they
et al., 2001), and with semi-natural grasslands as these are impacted are normally limited to the measurement of mean concentration
by the atmosphere (e.g. Horváth et al., 2005). These studies differences and vertical gradients.
D. Fowler et al. / Atmospheric Environment 43 (2009) 5193–5267 5207

Table 1
Practical suitability of systems to measure ammonia biosphere–atmosphere exchange.

Chemical approach Advantage Disadvantage Application References

Box AGM REA EC


Simple, cheap, high
Batch filter packs Uncertain gas – aerosol split, batch J J L L Sutton et al. (1993a,b)
air volume

Simple, cheap good


Batch denuders Low air volume, batch J J L L Duyzer (1994)
gas-aerosol split

Automated batch Automated in field, medium High laboratory processing cost, J J L L Loubet et al. (2006, 2009)
annular denuders cost, precise, high air volume only hourly, need two
systems for fluxes

Continuous annular Automatic sensitive, Cost, complexity, fault liable, JJ JJ L L Wyers et al. (1993), Erisman and Wyers (1993),
denuders precise, high air volume gradient only Sutton et al. (1995, 2000b, 2001a)
and Nemitz et al. (2001b)

Continuous parallel Automatic, sensitive,


Cost, complexity, fault liable J J JJ L Nemitz et al. (2001a) and Hensen et al. (2008)
plate denuders high air volume REA

Continuous Automatic, Cost, complexity, fault liable J JJ J L Hensen et al. (2008)


mini-WEDD sensitive, precise

Continuous Automatic, sensitive, Cost medium complexity J JJ J L Flechard et al. (2007) and Hensen et al. (2008)
membrane precise, reliable
denuder AIRmonia

Photo-accoustic Automatic, sensitive, Cost, complexity, not reliable JJ J L L Whitehead et al. (2008)
in principle reliable

Tunable diode laser Automatic, sensitive, Very high cost, complexity, J J L J Shaw et al. (1998), Famulari et al. (2004),
fast response (>10 Hz) maintenance Twigg et al. (2005) and Whitehead et al. (2008)

AGM: Aerodynamic Gradient Method; REA: Relexed Eddy Accumulation, EC: Eddy Covariance.

The benefits of quantifying ammonia fluxes using the gradient


technique have been clearly demonstrated by the many papers
published using this approach. In terms of informing our under-
50
standing of ammonia exchange processes and model development,
NH3 flux (ng m s )

Pre-cut
-1

CEH
this has almost exclusively been provided by measurements
-2

25 FRI
using the aerodynamic gradient method (90%), with a few studies
0 (in continental climates) applying the modified Bowen Ratio
method (5%). By contrast, the key disadvantage of this method is
-25 that it depends on micrometeorological stationarity, with no
change in the vertical flux with height. These flux/gradient methods
-50
are not suitable for the study of exchange fluxes where advection of
-75 ammonia from local sources is of interest (e.g. Loubet et al., 2001,
NH3 flux (ng m s )

2006) and where gas-particle ammonia–ammonium interactions


-1

22/05/00 23/05/00 24/05/00 25/05/00


Post-cut are significant (e.g. Brost et al., 1988; Nemitz et al., 1996, 2004b).
-2

CEH
FRI
FAL-D
To address some aspects of advection and air chemistry interac-
500
FAL-CH tions, determination of fluxes at a single height offers a way forward.
If this can be achieved, in principle, deployment of replicate
measurement systems at several heights could then be able to
0 determine vertical flux divergences (Sutton et al., 2007). Both the
Eddy Covariance (EC) method and Relaxed Eddy Accumulation (REA)
allow fluxes to be determined from measurements at one height, and
31/05/00 01/06/00 02/06/00 03/06/00
NH3 flux (ng m s )

have therefore been the subject of several recent studies. The


-1

6000 Post-fert
advantage of REA is that slow response ammonia measurements can
-2

CEH
FRI be combined with fast response switching, as has recently been
FAL-D
4000 demonstrated in an inter-comparison of 4 REA systems for ammonia
FAL-CH
(Hensen et al., 2008). A further advantage is that programmed
2000 periods of random switching between air up- and down-drafts
allows automatic zero checks and the correction of any bias (Nemitz
0 et al., 2001a; Hensen et al., 2008). By contrast, the challenge for REA
06/06/00 07/06/00 08/06/00 09/06/00 and ammonia is that the concentration differential to be measured is
typically much smaller than for the gradient method, which to a large
Fig. 3.1. Inter-comparison of continuous profile systems for measuring ammonia fluxes extent cancels the benefit of auto-referencing.
by the aerodynamic gradient method (AGM), from the GRAMINAE Braunschweig
Experiment. Although highly scattered, this flux inter-comparison is unique and repre-
sents the current state-of-the-art in chemical detection systems for ammonia fluxes.
Several recent studies have demonstrated the potential of fast
Increased emissions due to cutting of the underlying grass sward (29 May) and the effect response tunable diode laser absorption spectroscopy (TDLAS)
of N fertilization with (100 kg N ha1, 5 June) are clearly shown (Sutton et al., 2007). for measurement of ammonia fluxes by eddy covariance
5208 D. Fowler et al. / Atmospheric Environment 43 (2009) 5193–5267

1200 average concentration in the ecosystem canopy, both of which


TDLAS vary in time and in space (e.g. Sutton et al., 1995; Asman et al.,
1000 QC-TDLAS 1998). Within the canopy, several sources and sinks combine
together to determine the average ammonia concentration in the
Flux NH3 (ng m-2 s-1)

800 canopy, including exchange with plant tissues through stomata,


with leaf cuticles and with decomposing leaf litter and the soil
600
surface (e.g. Denmead et al., 1976; Sutton et al., 1993b, 1998).
Ammonia within or immediately above the canopy airspace may
400
undergo chemical reactions, for example forming particulate
matter, while depletion of gases within a plant canopy coupled
200
with altered microclimate can lead to evaporation of ammonium
containing aerosol (Brost et al., 1988; Nemitz et al., 2004b).
0
Finally, the complex nature of ammonia sources and sinks in rural
-200 landscapes means that strong horizontal gradients of ammonia
29/04/2005 29/04/2005 30/04/2005 30/04/2005 30/04/2005 30/04/2005 occur. The result is that ammonia is not simply deposited from
14:24 20:24 02:24 08:24 14:24 20:24
above, but fluxes are often significantly influenced by advection
Date, Time (GMT)
effects, for example where advection from a ground level source
Fig. 3.2. Fluxes of NH3 measured by eddy covariance over intensively managed beneath a micrometeorological reference height adds substan-
grassland (Easter Bush, Scotland) several days after the application of liquid manure to tially to a net deposition flux (Loubet et al., 2001, 2008, 2009;
the grassland (Sutton et al., 2007). Milford et al., 2001b).
It is relevant to summarize the main influences on the primary
drivers of exchange, the atmospheric ammonia concentration and
the mean concentration of ammonia within the canopy. The first
(Shaw et al., 1998; Famulari et al., 2004; Whitehead et al., 2008). of these is influenced partly by dispersion from adjacent
In principle, reliable flux measurements can now be made for ammonia sources and partly by exchange with the surface itself.
periods of large ammonia fluxes (e.g. after manured application), Over a surface which acts as an ammonia sink, above-canopy
as has recently been demonstrated in an inter-comparison of ammonia concentrations are depleted relative to background
two laser systems (Fig. 3.2). However, there was little correlation concentrations, while above canopy concentrations may be
for fluxes <50 ng m2 s1, while the AMANDA systems have been significantly enhanced if the surface is a net source (e.g. Sutton
shown to be able to measure <10 ng m2 s1 (e.g. Sutton et al., et al., 2000a).
1998). Table 1 provides an overview of these and other systems The mean ammonia concentration of the canopy itself results
for measuring ammonia fluxes. In principle, TDL and EC has the from the resolution of competing emission and deposition
potential to be rated as high as the continuous gradient methods, processes with leaf cuticles, through stomata and with the ground
but this still needs to be demonstrated by a more substantial body surface. The concept of ‘compensation point’ concentrations has
of published measurements, particularly over longer time often been used to describe these relationships. The earliest view of
periods and of a suitable quality for testing of models. a compensation point for ammonia related it to exchange through
plant stomata with the leaf apoplast (Lemon and Van Houtte, 1980;
3.3. Key controls on biosphere atmosphere exchange of ammonia Farquhar et al., 1980). Under this interpretation, net ammonia
fluxes would depend on the difference between what has since
Fig. 3.3 summarizes the main processes affecting the net been termed the ‘stomatal compensation point’ (cs) and the
exchange of ammonia with the atmosphere (Sutton et al., 2007). atmospheric concentration (ca). By contrast, subsequent studies
The primary driver of ammonia exchange is the difference highlighted the fact that ammonia deposition rates were often
between the atmospheric ammonia concentration and the faster than feasible by stomatal uptake, demonstrating the impor-
tance of ammonia deposition to leaf cuticles (e.g. Sutton et al.,
1993a,b; Duyzer, 1994). The resolution between these positions was
provided in the development of the concept of the ‘canopy
compensation point’ (cc), which accounts for both bi-directional
stomatal exchange and deposition to leaf cuticles (Sutton and
Fowler, 1993; Sutton et al., 1995). Such canopy compensation point
concepts have since been further developed to include bi-direc-
tional exchange with leaf surfaces and exchange with the ground
surface under the canopy (e.g. Sutton et al., 1998; Flechard et al.,
1999; Nemitz et al., 2001b).
These inter-relationships are developed quantitatively in a two-
layer canopy compensation point model (Nemitz et al., 2001b). One
of the key points to note about ammonia compensation points is
that they depend on the net solubility of ammonia in aqueous
solution, which is largely dependent on its equililbrium with
ammonium ions. By combining the temperature dependence of the
Henry equilibrium and the ammonium dissociation equilibrium,
Fig. 3.3. Summary of the key issues affecting the net land–atmosphere exchange of the gaseous ammonia concentration can be compared with a given
ammonia. Each of these interactions can lead to ammonia fluxes changing with height ratio of [NHþ þ
4 ]/[H ], which has been termed G (Nemitz et al., 2000c,
above the ground. Ideally, flux measurements, based on e.g. relaxed eddy accumulation
or eddy covariance, made at several heights above the canopy would be used to
2001a; Sutton et al., 2000b). On this basis, G can be used to provide
quantify these effects, though until now such assessments have had to focus on the use temperature-normalized compensation points, for example
of vertical profiles in mean ammonia concentration. cs ¼ f(T, Gs) where Gs ¼ [NHþ þ
4 ]apoplast/[H ]apoplast.
D. Fowler et al. / Atmospheric Environment 43 (2009) 5193–5267 5209

3.4. Effects of ecosystem type on ammonia It is feasible that these emission periods represent events
biosphere–atmosphere exchange of desorption of previously deposited ammonia occurring in
dry conditions. Conversely, it is also feasible that they represent
It has long been established that ecosystem type affects net apparent ‘emissions’, being an artefact whereby horizontal
ammonia fluxes (cf., Denmead et al., 1976; Horváth, 1983; Sutton ammonia concentration gradients away from an adjacent ground-
et al., 1993b, 1995; Duyzer, 1994). Overall, unfertilized ecosystems, based ammonia source (e.g. manure spreading, farms etc) lead to
such as forest and moorlands are generally sinks for atmospheric an advection error. This would reduce the measured deposition
ammonia, while fertilized and grazed agricultural ecosystems tend rate and could explain apparent ammonia upward fluxes in this
to show bi-directional fluxes with some periods of deposition and context. This illustration emphasizes the complexity of measuring
some periods of emission. Of course, the distinction is not absolute, ammonia exchange processes and highlights the need for further
as smaller ammonia emissions may also occur from semi-natural investigation of each option.
ecosystems (e.g. Sutton et al., 1995; Flechard and Fowler, 1998). The above example of mainly ammonia deposition to a forest
However, such a general difference is clear, and can be explained by ecosystem may be contrasted with recently published measure-
the increase in cs and cground that occurs in fertilized and grazed ments of ammonia fluxes over an intensively managed grassland in
ecosystems. Two recently published examples of ammonia exchange the Netherlands (Wichink Kruit et al., 2007). The diurnal patterns in
provide a useful basis to highlight these differences. ammonia concentration and net exchange flux are illustrated
Neirynck et al. (2005) report ammonia flux measurements made in Fig. 3.5. Hourly ammonia concentrations in the air at this site
using the AMANDA technique (Wyers et al., 1993) over a coniferous were again very large, 1–50 mg m3, with an overall mean of around
forest in Belgium. Their forest site occurs in an area of intensive 10 mg m3. In this case, net emission occurred for around 40% of the
livestock rearing, so that ammonia concentrations from some wind diurnal period (10:00–20:00), with net deposition at other times.
directions are very large (5–25 mg m3) while for other wind sectors Wichink Kruit et al. (2007) also estimated the canopy
ammonia concentrations were more moderate (2–4 mg m3). Even compensation point (cc) based on profile estimation of c (zo0 ). They
considering the effects of canopy wetness, in all conditions the then combined this with estimates of surface temperature to esti-
mean diurnal profiles show consistent net deposition to the forest mate G(zo0 ) or ‘Gc’ from the measurements (Fig. 3.6). Estimated
canopy. Curiously, the largest deposition fluxes occurred in dry values of cc were in the range 1–30 mg m3, which is comparable
conditions, which is unusual, as Rw would be expected to be smaller with other studies for managed grassland (e.g. Milford et al., 2001a;
when the canopy is wet (Sutton et al., 1995; 1998; Nemitz et al., Sutton et al., 2001; Loubet et al., 2006), and substantially smaller
2001b). Although this difference is partly explained by different than the upper values implied for the forest in Fig. 3.4. Normalized
values of Fmax during conditions of different canopy wetness, this for canopy temperature, the values of Gc were in the range
does not to fully explain the difference. Further analysis by Neir- 200–11,000 through a period of May to October 2004, with a mean
ynck et al. (2005) showed differences in the overall canopy resis- value of just over 2000.
tance (Rc) for ammonia deposition with different canopy wetness
and temperature, and with larger values of Rc occurring at higher 3.5. Modelling surface–atmosphere exchange of ammonia
ammonia concentrations.
Neirynck et al. (2005) did report periods of net ammonia Over recent years, the canopy compensation point approach has
emission from their forest canopy (Fig. 3.4). These were recorded become the standard technique to model bi-directional ammonia
during periods with winds from the high ammonia wind sector and surface–atmosphere exchange. Starting with the 1-layer models
found to only happen at very large ammonia concentrations, which offsetting bi-directional stomatal exchange against deposition to
occurred when air temperatures were larger than 15  C and relative leaf surfaces (Sutton and Fowler, 1993; Sutton et al., 1995, 2007),
humidity less than 60%. Fig. 3.4 presents an intriguing result, since subsequent models have developed in several directions. The main
according to the concepts of ammonia compensation points subsequent developments can be summarized as follows:
a different picture should emerge, namely that periods of ammonia Treatment of multiple canopy layers: In addition to ammonia
emission occur when atmospheric ammonia concentrations are exchange with the top part of the canopy, leaf litter and the soil
small. By contrast, such a relationship is possible when emissions surface have been shown to be important sources of ammonia
from a canopy are strong (and not compensation point driven), so emission into the plant canopy (e.g. Nemitz et al., 2000a). For an
that it is the emissions from the surface that generate increased oilseed rape canopy Nemitz et al. (2000b) also highlighted the
ammonia air concentrations. This phenomenon was also observed importance of an upper and lower part of the main foliage, dis-
following harvest of an oilseed rape field (Sutton et al., 2000a,b). tinguishing the main foliage from an over canopy of oilseed ‘siliques’.
In practice this three layer model becomes complex to parametrize,
and there has now developed consensus that a two-layer model
represents an appropriate balance of realistic description while
avoiding excess complexity. A recent implementation of the 2-layer
model is that of Personne et al. (2009) for the GRAMINAE Integrated
Experiment (Sutton et al., 2009b). They used measured bioassay
estimates of Gs and Glitter (Mattsson et al., 2008a,b; Herrmann et al.,
2008) combined with an energy balance approach to calculate
component resistances, showing close agreement with measured
ammonia fluxes (Fig. 3.7).
Treatment of cuticular fluxes: The initial parametrisations of
the cuticular resistance (Rw) allowed only for deposition, depen-
dent on relative humidity (Sutton and Fowler, 1993; Sutton et al.,
1995) or vapour pressure deficit (Nemitz et al., 2000c, 2001b). As
noted above for the forest example, ammonia deposited to a canopy
Fig. 3.4. Dependence of ammonia flux on concentration in the high ammonia wind surface may also be re-emitted to the atmosphere, particularly
sector during warm daytime conditions with dry canopy (Neirynck et al., 2005). under drying conditions. A first approach to simulate this effect
5210 D. Fowler et al. / Atmospheric Environment 43 (2009) 5193–5267

Fig. 3.5. NH3-concentration (upper panel) and NH3-flux (lower panel) measurements above managed grassland in The Netherlands from 18 July until 15 August 2004 (summer
period). The horizontal axis represents time of the day (UTC). Local time is UTCþ2. The vertical axis represents the NH3-concentration (mg m3) or NH3-flux (ng m2 s1). Diamonds
A are calculated values for the half-hourly NH3-concentration or NH3-flux; the solid line (dd) (with vertical 25 and 75 percentile bars) is the median of all half-hourly fluxes for
that time. The dashed line (- - -) in the lower panel is the mean leaf wetness signal during this period (Wichink Kruit et al., 2007).

treated the leaf surface as a humidity dependent capacitance 3.6. Dynamic simulation of ecosystem C–N cycling
(Qd), which would be in equilibrium with a non-zero leaf surface and ammonia fluxes
concentration (cd) (Sutton et al., 1998). In this case an adsorption/
desorption resistance (Rd) is also defined. This first dynamic A disadvantage of the compensation point scheme for simu-
approach had the advantage of being able to simulate ammonia lating ammonia fluxes outlined above is that empirical values of
charging and discharging of the cuticle, but had the disadvantage G must be provided. The only way forward from this position is to
that the leaf surface pH needed to be specified as an input. The develop models of carbon–nitrogen cycling that can simulate
approach was further developed by Flechard et al. (1999) who G values for the different pools based on an understanding of the
considered the full aqueous chemistry on leaf surfaces, dependent pool dynamics (cf. Massad et al., 2008). To date, the only such
on multiple air pollutant inputs and potential leaching of base model to attempt this coupling is the PaSim model of Riedo et al.
cations from leaf surfaces. In this model, leaf surface pH is solved by (2002). The model distinguished plant nitrogen pools into struc-
ion balance, and the model is able to take account of the effects of tural nitrogen, substrate nitrogen and apoplastic nitrogen (a sub-
other trace components such as SO2 on ammonia fluxes. Burkhardt pool of substrate nitrogen), linking these with plant uptake and
et al. (2008) have recently extended this model to incorporate the growth processes. The model was parametrised based on measured
two-layer approach with bi-directional exchange for each of the fluxes for a Scottish grassland (Milford et al., 2001b) and has
leaf surface, stomata and ground surface. The cuticular resistance recently been tested for the Braunschweig Experiment (Sutton
clearly responds to the chemistry of the liquid film on vegetation et al., 2009b). Overall, the model was able to simulate the larger
and the combination of reactive gases present (Flechard et al., net emissions that occurred after cutting and after fertilization, as
1999). Even in the absence of additional reactive trace gases, the well as the decline in the 10 day period following fertilization. By
cuticular resistance declines with increasing NH3 concentration. contrast, the component fluxes were less well described. Bioassays,
In a series of chamber experiments Jones et al. (2007) quantified chamber measurements and within-canopy profiles during the
the relationships between ambient NH3 concentration and the bulk Braunschweig Experiment (Mattsson et al., 2008a,b; Herrmann
canopy resistance for a range of moorland vegetation as shown in et al., 2008; Nemitz et al., 2009b) highlighted leaf litter as being
Fig. 3.8 in which the non-stomatal ‘cuticular’ resistance is seen to a key source of emission following cutting. This source is currently
increase lineary with NH3 concentration, leading to much smaller not simulated in PaSim, which simulated that increased emissions
deposition at high concentrations than if deposition velocity after cutting were due to an increase in apoplastic ammonium. The
remained constant with concentration as is usually assumed. bioassays indicated that the foliage was more likely to be a sink of
D. Fowler et al. / Atmospheric Environment 43 (2009) 5193–5267 5211

Fig. 3.6. Derived canopy compensation points (cc ¼ c(zo0 )) (upper panel) and ratios between NHþ þ 0
4 and H concentration (Gc ¼ G(zo )) (lower panel) from the end of May until the
end of October 2004 (diamonds) and the constant value (2200) that is normally assumed for modeling (line) (Wichink Kruit et al., 2007).

soil/litter ammonia emissions, highlighting the need for improved 1. That unreplicated ammonia flux measurements in most studies
ecosystem modelling of ammonia exchange that accounts for litter are highly uncertain and need to be considered with caution
decomposition processes (Sutton et al., 2000b). when compared with model estimates.
2. That advection effects can significantly influence measured
3.7. Integrating ammonia exchange processes ammonia fluxes, both due to dispersion away from nearby
point sources (correction for advection effects increases net
The preceding sections have highlighted the many processes deposition) and due to emissions from an emitting field itself
and interactions that combine to regulate ammonia fluxes between (correction for advection increases net emission).
vegetation and the atmosphere. It thus becomes a major challenge 3. That gas-particle interactions had a minor effect on measured
to integrate each of these processes to develop a holistic view. It is ammonia fluxes, though the relative effect on calculated aerosol
necessary to quantify the interactions in each case in order that deposition rates was significant (being the cause of apparent
valid conclusions can be obtained. This creates a major challenge aerosol emissions).
for experimentalists to be able to address all the questions in the 4. That reasonable agreement can be made between relaxed eddy
field. For example, in the absence of measurements of horizontal accumulation for ammonia and the aerodynamic gradient
concentration profiles, it is difficult to quantify the potential for method, though measurements are not yet sufficiently precise
advection effects to have influenced the results presented in to detect flux divergence (except for possible cases of extreme
Fig. 3.4. Similarly, it remains an open question in most studies advection errors).
whether gas-particle interactions have a significant influence on 5. That net emissions from this grassland canopy are controlled by
measured ammonia fluxes. the recapture of leaf litter ammonia emissions by overlying
It was with such interactions in mind that the GRAMINAE foliage and the interaction of cuticular exchange pools with
Integrated Experiment was designed. A number of findings from mainly stomatal uptake of ammonia from the leaf litter emis-
this experiment have already been mentioned, but the experiment sions. Net emissions increase following cutting due to exposure
demonstrates both the challenges and the power of developing an of the litter and cutting induced senescence, with a similar
integrated approach. Fig. 3.9 summarizes each of the issues and recapture process affecting net emission following fertilization.
measurement methods that were investigated during this experi-
ment (Sutton et al., 2008a). The key conclusions from this study A range of models is able to simulate the dynamics of net
have been summarized by Sutton et al. (2009b) and include: ammonia exchange with the managed grassland, but further
5212 D. Fowler et al. / Atmospheric Environment 43 (2009) 5193–5267

4500 key challenges include the climate dependence of net ammonia


Measured flux
4000 emission and deposition, and the characteristic fluxes of other
Modelled flux (Gamma Litter)
3500 ecosystems in the world.
NH3 fluxes (ng m s )
-1

In principle the models of ammonia exchange incorporate the


3000
-2

main features of environmental conditions and could therefore be


2500
applied in different climates. Here the limitations include the lack of
2000
available data for empirical factors such as G values and the back-
1500 ground data to extrapolate to conditions with different climates.
1000 Currently, the estimates of G have mainly been derived for cool
500 European conditions and for a very limited number of ecosystems.
0 Although there have been many studies of ammonia emission from
-500
fertilized tropical systems, such as rice and maize, there are few
22-May 29-May 05-Jun 12-Jun published studies of ammonia fluxes over semi-natural unfertilized
tropical ecosystems. The rates of ammonia deposition or emission
Fig. 3.7. Comparison of ammonia fluxes simulated by a two-layer canopy compensa- in these situations are thus highly uncertain. Given the differences
tion point model (SURFATM-NH3) with measured fluxes (Fmg) during the GRAMINAE
Braunschweig Experiment. For this model scenario, the ground emission is assumed to
in biology of these systems, measurements are required to underpin
originate from leaf litter based on measured Glitter (Personne et al., 2009). modelling approaches.
A modest degree of climate change (e.g. þ2  C) is a much easier
matter to simulate, for example based on the analysis of temperature
attention is needed to develop dynamic treatments of ammonia effects within existing datasets. The thermodynamics of ammonia
emissions from leaf litter decomposition. solubility and dissociation are rather straightforward, indicating for
example a doubling in cs every 5  C increase for a given value of
3.8. Future challenges for ammonia exchange G (Sutton et al., 2001). However, caution is needed before making
climate change simulations on this basis. Analysis of the PaSim model
The results from the GRAMINAE Integrated Experiment provide under different temperature regimes showed that net ammonia
a microcosm of some of the key challenges to measure ammonia fluxes for Easter Bush in Scotland (cf. to measured fluxes of Milford
fluxes and model the process interactions. In a wider perspective et al., 2001b) were quite insensitive to temperature. For example,
increased temperature (in the absence of moisture limitation) led
to increased grass growth which diluted available nitrogen pools,
thereby reducing G values (Sutton and Milford, unpublished simu-
lations). Similarly, increases in wetness, while favouring smaller
values of Rw may also lead to increased rates of leaf litter decompo-
sition, favouring ammonia emissions. To take another example, in
colder conditions, NH3 from manure application to the land surface
tends to be emitted at smaller rates, but the emission lasts longer,
especially if a waterlogged or frozen soil conditions prevent infiltra-
tion. With these illustrations in mind, it becomes a major future
challenge to generalize how ammonia fluxes might change in the
future under different climatic regimes.

4. Sulphur dioxide

4.1. Introduction

There are three very different spatial scales relevant to the


exchange of SO2 at terrestrial surfaces, first the micro-scale, at which
the chemical and biological interactions occur (Fig. 1.1). Second is the
spatial scale at which most of measurement and interpretation
takes place, which is the field scale (averaged over 103–105 m2) for
measurements using micrometeorological methods. Lastly, the
application of knowledge of the surface exchange process is primarily
at regional to continental scales to characterise the fluxes and
budgets within chemistry transport models (CTM) and comparisons
with the concentration fields observable from satellites.
The measurements have mainly been made at the field scale
using micrometeorological methods, although there have been
some laboratory studies, mainly in the early days of SO2 dry
deposition research. The initial measurements were used to esti-
mate the regional scales of dry deposition, often using a fixed
deposition velocity as a key variable within long-range transport
Fig. 3.8. The relationship between ambient ammonia concentrations and the cuticular models (e.g. Fisher et al., 1978). With larger datasets of measure-
resistance to deposition for moorland vegetation. A significant difference was found ments covering a wide range of conditions, it is clear that rates
between day and night for the bulk canopy resistance (Rc), which included both
stomatal uptake and deposition to the leaf surfaces. Once the effect of the stomatal
of dry deposition vary considerably in time and space (Fowler
resistance (Rs) was accounted for, the cuticular resistance (Rw) was found to approx- and Unsworth, 1979) in particular because the sinks available at
imately constant day and night (Jones et al., 2007). terrestrial surfaces, including the stomata in vegetation, leaf
D. Fowler et al. / Atmospheric Environment 43 (2009) 5193–5267 5213

Estimation of Advection of Quantification of Interactions with acid


farm-scale NH3 NH3 from nearby energy balance & gases and ammonium
emissions sources & effects environ controls on particles & effects
from plume on vertical fluxes NH3 exchange on net NH3 fluxes
measurements

Continuous
measurement of NH3
fluxes by gradient
Effects of dew and REA approaches
Effects of cutting
& leaf surface
& N fertilization
chemistry
events & choices
on NH3 fluxes NH3 compensation
points of foliage
Plant bioassay
Within-canopy
determination of
cycling of NH3 fluxes
NH3 emission potential

Determination of Effects of leaf senescence


within-canopy NH3 release from and plant species on
turbulent exchange litter decomposition NH3 emission potential

Soil chemistry interactions


with plant N uptake & NH3 fluxes

Fig. 3.9. Overview of issues addressed by the GRAMINAE Integrated Experiment (Sutton et al., 2008a).

surfaces and the presence of liquid water on vegetation from dew 4.2. Worldwide advances in SO2 flux monitoring and modelling
or rain, present a variable absorbing surface. The data have shown
the role of atmospheric composition and surface leaf water chem- 4.2.1. Asia
istry in controlling canopy resistance. Sulphur dioxide dry deposition to vegetated surfaces is largely
Most dry deposition measurements of sulphur dioxide over the controlled by non-stomatal processes, but in many arid ecosystems
last 30 years have been made in N. America and Europe, and have and deserts of the world where vegetation is sparse, the nature and
served as a basis for the parameterisation of dry deposition models pH of soils determine the sink strength. In Asia, substantial efforts
(Erisman, 1994; Smith et al., 2000; Zhang et al., 2002), which in turn have for example gone into the characterization of SO2 uptake by
have been applied to ecosystems in different parts of the globe. loess soils, given their large geographical representation in
However, most SO2 emission and deposition now occurs outside Northern China, their alkaline nature and their ability to neutralize
N. America and Europe. Asia’s contribution in 1985 of 20% to global atmospheric acidity and to serve as an oxidation medium for
anthropogenic SO2 emissions has doubled since then, reaching 37% SO2. Both micrometeorological and laboratory- or field-based
by the year 2000, of which 23% is emitted by China alone and 5% by flow reactor methods were deployed. New micrometeorological
India. measurements over forests and short vegetation have also been
Southern China is one of the world’s most sulphur polluted reported over the last 10 years in the region, reflecting the growing
areas. Paradoxically, in Northern China, where ambient SO2 concern over increasing sulphur emissions and deposition to
concentrations are very large, rainfall is generally alkaline, and the ecosystems.
areas polluted by acid rain do not necessarily correspond to the
areas of high SO2 emissions. One of the reasons for this discrepancy 4.2.1.1. Sulphur dioxide deposition to soils. Utiyama et al. (2005)
is the presence of alkaline soils (yellow sand) distributed over measured dry deposition to loess soil and dead grass in Beijing
the arid areas of N.W. China (e.g. the loess plateau and Gobi desert), using the aerodynamic gradient method, though in neutral condi-
the windborne erosion of particles with high base cation concen- tions 22% of the time. In stable or unstable thermal stratification,
trations can neutralize atmospheric acidity (Utiyama et al., 2005). they used a surface reaction concept for inferring dry deposition.
Loess soil, which covers vast areas of the Eurasian continent Two surface kinetics models were considered: either i) the reaction
extending from N.E. China to Central Asia, contains Ca in large occurs in soil pores and SO2 molecules diffuse through porosity
quantities, and calcium carbonate (CaCO3) reacts with atmospheric while reacting with alkaline sites on the pore surface; or ii) the
SO2, to form calcium sulphate (CaSO4). Thus, even bare soil without adsorption mechanism is of Langmuir–Hinshelwood type, where
vegetation may be a significant sink (Sorimachi and Sakamoto, the partial pressure of SO2 and its desorption pressure from the site
2007), which may affect the regional SO2 budget if the process is are in equilibrium. The model parameters are then fitted so that the
inadequately quantified in dry deposition models. resulting (modelled) vertical SO2 concentration gradient matches
In this section we review research and monitoring from the last the observations. Measured deposition velocities (Vd) were in the
decade, including SO2 dry deposition measurements from Asia, range 1–12 mm s1.
North America and Europe, as well as findings from long-term flux Sorimachi and Sakamoto (2007) conducted laboratory-based
monitoring experiments. The current state of knowledge concern- flow-reactor measurements of SO2 deposition to soil samples from 12
ing mechanisms of SO2 dry removal from the atmosphere is sites in the arid loess plateau and deserts of Northern China. Canopy
reviewed, with consequences for temporal trends in atmospheric resistances in the range 28–650 s m1 (with a mean of around
concentration and deposition, and key future research areas are 200 s m1) were found to be dependent on RH, as was S(IV) oxidation
identified. to S(VI). It was hypothesized that Northern China soils, which are
5214 D. Fowler et al. / Atmospheric Environment 43 (2009) 5193–5267

much more alkaline than in Southern China, are a greater sink for SO2 The latter can only provide crude estimates of deposition rates, as
and a neutralizing buffer for acidifying atmospheric deposition surrogate surfaces do not adequately account for the complexity of
(Sorimachi et al., 2004). By comparison, in modelling SO2 deposition natural surfaces, but they do allow continuous monitoring at
to Asia, Xu and Carmichael (1998) used a fixed Rc for deserts of a number of sites and help to detect trends. The use of inferential
500 s m1, which is clearly too high in the case of Northern China methods requires underpinning measurements of surface resistance
deserts. A flow-reactor was also used by Sakamoto et al. (2004) to to representative surfaces to parameterise the models used. It is also
determine SO2 dry deposition to yellow sand and soil-mediated SO2 important to note that long-term changes in canopy resistance are
oxidation by O3. The deposition velocity for SO2 increased with RH likely, especially in regions in which the relative ambient concen-
due to the positive effect of RH on the SO2 oxidation rate. trations of SO2 and NH3 change with time, as for example in Europe
during the period 1990–2005 (Fowler et al., 2001c, 2007).
4.2.1.2. Micrometeorological measurements over vegetated Ta et al. (2005) thus provided long-term sulphur dioxide dry
areas. Matsuda et al. (2006) reported micrometeorological (aero- deposition estimates across Gansu Province, China, using K2CO3-
dynamic gradient) flux measurements of SO2 and O3 over a tropical coated surrogate sulfation plates. Samples were taken monthly for
(teak) forest in Northern Thailand in dry and wet seasons. The 11 years at 48 sites distributed across 11 cities in the province.
deposition velocity for SO2 in the dry season was rather low, in the The data showed that cumulative SO2 dry deposition fluxes were
range 1–3.1 mm s1 in daytime and 0.8–1.1 mm s1 in night-time. closely related to local SO2 emissions, and had seasonal variations
In the wet season, however, Vd was much higher due to enhanced with maxima in winter and minima during summer as a result of
non-stomatal uptake in wet conditions, with values in the range higher winter and lower summer SO2 emissions and concentra-
9.5–13.9 mm s1 in daytime and 2.6–4.2 mm s1 in night-time. The tions. Monthly average SO2 deposition velocities, however, peaked
data were compared with a recent non-stomatal resistance scheme in April–July at 11–27 mm s1, and minimum values were observed
(Zhang et al., 2003a), and it was concluded that extended experi- in January at 2–10 mm s1.
mental SO2 dry deposition studies are needed in the tropics, while Inferential models (Erisman, 1994; Smith et al., 2000; Zhang
Zhang et al. (2003a) recommend more studies to quantify the et al., 2002) may be used to estimate dry deposition at observation
different effects of dew and rain on SO2 deposition. sites, where single-height ambient concentration measurements
Sulphur dioxide dry deposition was also measured by Matsuda are available together with standard meteorological data. Model
et al. (2002) over a red pine forest located in Oshiba Highland, parameters, however, have been largely derived from European and
Nagano, Japan, using a Bowen ratio technique. The median daytime N. American studies and may not necessarily be adequate for Asian
(12:00 to 14:00) deposition velocity was 9 mm s1. Measurements vegetation and soils, and numerical evaluations need to be carried
compared favourably with estimates by an inferential model for out. Thus Takahashi et al. (2002) simulated the dry deposition of
wet conditions, but for dry or mixed wet-dry surfaces there were SO2 to a Japanese cedar (Cryptomeria japonica) forest located in
large differences between model and measurements. The authors Gumma Prefecture, based on the results of 1-year’s concentration
ascribed the discrepancy to a relative humidity threshold value measurements. The mean modelled Vd at this site was 8.8 mm s1
used in the inferential scheme to characterise canopy wetness, and (Takahashi et al., 2001). The inferential estimate of the dry sulphur
pointed to the need for a refined parameterisation of the cuticle or deposition flux was 11.1 mmol m2 yr1 (3.6 kg S ha1 yr1), which
external leaf surface resistance. compared well with the net throughfall flux (12.4 mmol m2 yr1,
In a study of SO2 and O3 dry deposition to short grassy vegeta- or 4.0 kg S ha1 yr1). Over a broadleaf forest on typical red soil of
tion over an alkaline soil (pH ¼ 9.2) near Beijing, using the Southern China, Xu et al. (2004) simulated Vd for SO2 and partic-
aerodynamic gradient method, Sorimachi et al. (2003) measured ulate SO24 , as well as their atmospheric deposition fluxes. The
mean Vd values of 2 (1) mm s1 and 4 (2) mm s1 in late summer simulations indicated that about 99% of the dry sulphur deposition
and early winter, respectively. Although the grass was lush and flux in the forest resulted from SO2, which contributed over 69% of
thick in the late summer, and senescent and leafless in the early the total (wet þ dry) annual sulphur deposition.
winter observation period, there was no difference in the mean Rc By comparison, Wang et al. (2003) computed dry deposition
(180  270 s m1 and 180  300 s m1, respectively), but the fluxes of SO2 and SO2 4 for 1 year to agricultural land over red soil
uncertainties given reveal a large variability in measured Rc. The (pH ¼ 5.3–5.8) in the Jiangxi province of Central China. The crops
difference in Vd stemmed from the higher aerodynamic (Ra) and grown were rice paddies and oilseed rape. Sulphur dioxide
quasi-laminar sub-layer (Rb) resistances in late summer than in concentrations were measured 8 times day1, 7 days month1, using
early winter. The absence of vegetation and stomatal uptake in a bubbler method. Annual mean modelled estimates of Vd were
early winter, which might otherwise have reduced the SO2 sink 3.8 (0.16) mm s1 for SO2 and 0.20 (0.12) mm s1 for SO2 4 .
strength, seems to have been compensated for by the soil alkalinity. Measured monthly mean concentrations ranged from 9 to
As the soil was more exposed and the in-canopy aerodynamic 163 mg S m3 (6.7–121 ppb), with an annual mean of 64 mg S m3
resistance was reduced, the soil surface offered more adsorption (47 ppb). Estimates of total monthly wet and dry deposition of SO2
and reaction sites for SO2, with the result that the field was an and SO2 4 ranged from 2.2 to 20.3 kg S ha
1
with an annual total
1
equally efficient SO2 sink in early winter as in summer. deposition of over 100 kg S ha , of which 83% was via dry deposi-
The deposition velocity for SO2 was measured by Jitto et al. tion, accounting for over 90% of total S input to farmland in this area.
(2007) during a 1-year experiment over a canopy of irrigated rice
paddy in Thailand using the Bowen ratio technique. The deposition 4.2.2. North America
velocity was highest around noon and lowest at night. Seasonally- Few long-term datasets of SO2 dry deposition monitoring have
averaged values of Vd were 6.7, 12.5, and 15.1 mm s1 in the winter, emerged over the last 10 years, reflecting the declining importance of
summer, and rainy seasons, respectively. SO2 as an acidifying input relative to NOy and NHx. Advances have
nonetheless been made in inferential modelling of SO2 uptake,
4.2.1.3. Long-term deposition studies and inferential modelling. As especially regarding the quantification of the non-stomatal (external)
alternatives to costly and labour-intensive micrometeorological leaf surface resistance, which serve as a basis for simulating regional
measurements of dry deposition, several authors in Asia have esti- patterns of SO2 deposition (Zhang et al., 2002, 2003b).
mated long-term SO2 deposition using monitored concentration data Micrometeorological SO2 flux data from 5 sites (2 forests, a corn
and inferential models, or long-term artificial collection devices. field, a soybean field and a pasture) in eastern USA (Finkelstein et al.,
D. Fowler et al. / Atmospheric Environment 43 (2009) 5193–5267 5215

2000; Meyers et al., 1998) were compared with modelled data by that the flux remained nearly constant. The measurements of the
Zhang et al. (2003a), with the specific objective of evaluating the new atmospheric terms (Ra and Rb) revealed no trend, thus the trend in
non-stomatal resistance scheme of the new Canadian model (Zhang Rc is not caused by changes in turbulence, and are clearly a conse-
et al., 2002, 2003b). Over the forest sites, Finkelstein et al. (2000) had quence of the chemical affinity of the surface changing with time.
noted that wetness tended to increase deposition velocity, but that These dry deposition measurements have proved valuable in
the nature of wetness (rain or dew) and its chemistry also controlled explaining the consistently larger decline in ambient SO2 concen-
canopy resistance. Non-stomatal surfaces like leaf surface, stem, tration than in emissions in Europe. In the absence of these flux
trunk and ground were important sinks for SO2, and the authors measurements it would be a matter of speculation as to the under-
concluded that a better understanding of surface chemistry and lying cause of the faster decline in ambient concentration than
water film chemistry was needed. emission. Even with these measurements there remains the possi-
Dew formation has long been recognised as an important sink for bility that SO2 oxidation rates have increased due to the growing
SO2 (Fowler and Unsworth, 1974, 1979). In more recent work Meyers oxidizing capacity of the atmosphere and have contributed to the
et al. (1998) show that dew is the reason for the relatively high early relative changes in emission and deposition (non-linearity). It will be
morning deposition rates at 2 of the 3 low vegetation sites studied in necessary in the further analysis and interpretation of European
Eastern USA. Recognizing the weakness of existing North American pollution climate data to carefully examine the relative importance
parameterisations (e.g. Meyers et al., 1998) in predicting SO2 depo- of the different contributors to the observed trends in concentration
sition rates to non-stomatal surfaces, especially in wet canopies, and deposition and quantify the relative importance of changes in
Zhang et al. (2003a) demonstrate that the AURAMS scheme (Zhang dry deposition and oxidation rates in the long-term trends.
et al., 2002) performed well at these 5 sites, using different resistance
values for dew and rain. The revised non-stomatal resistance scheme 4.2.3.2. Other recent European datasets. The SO2 flux–gradient data
(Zhang et al., 2003b) includes a treatment of in-canopy transport, soil obtained over short vegetation by Feliciano et al. (2001), collected
and cuticle terms, and is a function of relative humidity, leaf area over a period of 3 years in the mid 1990s at 3 different sites in
index and friction velocity. For wet canopies, the cuticular resistance Portugal, were important in providing Rc estimates for the Medi-
is treated differently for dew and rain. terranean region of Southern Europe. The 3 sites had contrasting
pollution climates, with a coastal, oceanic, humid meadow in
4.2.3. Europe N. Portugal, a hot and semi-arid pseudo-steppe and a site located in
4.2.3.1. Long-term flux monitoring in the UK. Sulphur dioxide fluxes a mostly dry, intensive agricultural area, both in S. Portugal. Median
have been monitored continuously since the mid 90s at two rural canopy resistances varied from 140 s m1 to 200 s m1 and although
sites in the UK, over agricultural land at Sutton Bonnington in the stomatal uptake was important when vegetation was biologically
English Midlands, and over moorland at Auchencorth Moss in active, the annual deposition was dominated by non-stomatal
S. Scotland (Fowler et al., 2001c, 2005, 2007). The dry deposition mechanisms on wet surfaces. The night-time canopy resistance,
measurements have continued to bring surprises over the last 10 a proxy for the non-stomatal resistance, increased with decreasing
years. At Sutton Bonnington, the ambient concentrations have relative humidity at all 3 sites. A comparison of nocturnal Rc for the
declined from about 2.8 ppb in 1996 to current values close to southern sites showed that, for a given level of relative humidity, the
1.4 ppb and yet the deposition velocity continues to increase due to Rc at the intensive agricultural site was systematically lower than at
continued reduction in the canopy resistance (Rc) (Fig. 4.1). Over the pseudo-steppe site, which is used more extensively for grazing
the monitoring period the canopy resistance has almost halved and and hay production. Although the authors make no mention of NH3
is now about 70 s m1. The consequence of the steady decline in being measured at these sites, it might be hypothesized that a higher
canopy resistance along with a decline in ambient concentration is NH3 concentration at the intensive agricultural site may have been

10
9
Mixing ratio (ppb)

8
7
6
5 NH
NH33

4 SO
SO2
2

3
2
1
0
1995 1996 1997 1998 1999 2000 2001 2002 2003 2004
SO2 /NH3 molar ratio (ppb ppb -1)

0.6 140

0.5 120
SO2/NH3
SO2/NH3
Rc SO2 (s m-1)

100
0.4 Rc (Wheat)
80
0.3
60
0.2
40
0.1 20
0 0
1995 1996 1997 1998 1999 2000 2001 2002 2003 2004

Fig. 4.1. Changes in the mean concentrations (ppbV) and ratio of ammonia and sulphur dioxide and in the May–July canopy resistance for SO2 deposition on Wheat at Sutton
Bonnington between 1996 and 2003.
5216 D. Fowler et al. / Atmospheric Environment 43 (2009) 5193–5267

responsible for the observed lower nocturnal Rc, compared with the
extensively-managed, steppe-like grassland.
The development of low-cost systems for the long-term moni-
toring of SO2, NH3 and other trace gas fluxes holds promise for
widening the range of dry deposition datasets for comparison with
inferential models. Hole et al. (2008) present an 18-month dataset of
SO2 fluxes acquired with a conditional time-averaged gradient
(COTAG) system (Fowler et al., 2001b; Famulari et al., in press) in
a semi-alpine ecosystem in Southern Norway. The mean annual SO2
deposition velocity was 4.0 mm s1, although the dataset included
some negative deposition velocities (upward fluxes), and the annual
mean Vd was 13.0 mm s1 if only the positive values were included.
The authors report evidence of enhanced SO2 deposition rates during
an episode in November 2005 when the NH3/SO2 ratio was high, and
conversely of decreased SO2 uptake and increased NH3 uptake in
November 2004 when the NH3/SO2 ratio was low. Comparison with
the inferential model by Zhang et al. (2002, 2003b) was satisfactory
but the model could not reproduce the large observed variability in
exchange rates, which may result from NH3–SO2 co-deposition
processes not being included in their resistance scheme.
More experimental evidence of the mutual influences of
NH3 and SO2 concentrations on their deposition rates was obtained
by Derome et al. (2004), though not by micrometeorological
measurements but using bulk precipitation collectors and through-
fall measurements in Scots pine canopies in SW Finland. The study
was conducted over a 6-year period (1993–1998) in the vicinity of
a Cu–Ni smelter, which emitted large amounts of gaseous NH3. These
emissions were shown to have strongly enhanced the scavenging of
atmospheric SO2 by the pine canopy, resulting in increased levels of N
and S deposition and increased foliar N and S concentrations. In
an NH3 fumigation experiment, Cape et al. (1998) had previously
described similar findings over a Scots pine forest in Central Scotland,
with the canopy resistance for SO2 decreasing with elevated NH3
concentration. Although NH3 concentrations were not measured in Fig. 4.2. A schematic representation of the dynamic canopy compensation pollution
the Finnish study (Derome et al., 2004), they were likely higher than model for SO2 and NH3 exchange over vegetation (from Flechard et al., 1999).
normally encountered in the countryside, except near animal
housing in areas of intensive agriculture, where such processes could
be significant. increased to 2 mg m3 to provide sufficient NH3 to neutralize the
acidity from the ambient SO2 oxidation in solution. The conse-
4.3. Control of surface uptake rates by leaf cuticular chemistry quence of the increase in ambient NH3 is to decrease the canopy
resistance for SO2 and increase the deposition rate of SO2 to the
A number of authors have addressed the issue of the chemical maximum under the prevailing atmospheric conditions. Demon-
control of surface pollutant uptake rates (e.g. Flechard et al., 1999). strating a close link between SO2 deposition and ambient NH3 is not
The most important finding for SO2 deposition is that the rates of new, as this was predicted in earlier work in the Netherlands (Van
deposition are controlled mainly by the chemistry at the vegeta- Hove et al., 1989). However, this work quantified the process
tion–atmosphere interface, and that as the surfaces are wet most of
the time, the processes are regulated by chemical processes within
the thin film of moisture. In principle, many compounds influence 10
the chemistry of this surface layer, including plant exudates and soil
0
SO2 flux (ng m s SO2)

derived compounds, but the key reactant for SO2 is NH3. Thus the
ambient concentrations of SO2 and NH3 essentially regulate the pH -10
-2 -1

of the surface moisture and thus control the uptake of SO2. The -20
full surface chemistry of the process has been incorporated into Meas. FSO2
-30
a dynamic mechanistic model shown in Fig. 4.2 (Flechard et al., Mod. FSO2 (χNH3 = ambient)
1999). The chemistry of the surface water film is initialised in the -40 Mod. FSO2 ( χNH3 = 2 μ g m-3)
model using measured precipitation chemistry, the model then -50 FSO2, max
simulates the dynamic responses of the net land–atmosphere
-60
exchange of SO2 as the ambient concentrations of the reactive trace
21/03/95 12:00

22/03/95 00:00

22/03/95 12:00

23/03/95 00:00

23/03/95 12:00

24/03/95 00:00

24/03/95 12:00

25/03/95 00:00

25/03/95 12:00

26/03/95 00:00

26/03/95 12:00

gases and meteorological conditions change. The model has been


shown to provide good agreement with observed 30 min average
fluxes for several days. An example is provided in Fig. 4.3, for a five-
day period at Auchencorth Moss in the Scottish Borders. The
general agreement between measured and modelled fluxes is Fig. 4.3. A comparison between measured and modelled SO2 fluxes at Auchencorth
excellent during the three-day period, 21st March to 23rd March Moss over the period 21-3-95 to 26-3-95 showing the influence of increasing ambient
1995. From the 24th March, the observed NH3 concentrations are NH3 concentration on SO2 flux (from Flechard et al., 1999).
D. Fowler et al. / Atmospheric Environment 43 (2009) 5193–5267 5217

correctly for the first time, demonstrated the effects in field period has meant that the SO2/NH3 ratio has decreased dramatically,
conditions at ambient concentrations and provided a mechanistic resulting in a reduced Rc for SO2 (Fig. 4.1). The data for Fig. 4.1 are of
model incorporating the full chemical scheme. course site specific and the footprint of the measurements in Not-
tinghamshire is only 104–105 m2, but this trend may be regarded as
4.4. Advances in deposition modelling representative for the regions in which ambient SO2 concentrations
have declined by up to an order of magnitude since 1970 and
The magnitude of dry deposition at the national and regional includes much of central and eastern England and the industrial
scales requires that process-based, rather than empirical, parame- regions of Germany, France, the Netherlands and Belgium. While
terisations be implemented in atmospheric models, accounting for ambient SO2 was a relatively good proxy for total atmospheric and
variations in surface chemical characteristics driven by local pollu- leaf surface acidity 15 or 25 years ago, the relative share of SO2
tion climates. The observation of changes in deposition velocity from compared to other inorganic atmospheric acids (e.g. HNO3 and HCl)
the European SO2 deposition studies of the 90s is now widely known is now much smaller. The NitroEurope network of 56 DELTA
and is being used by EMEP to explain growing discrepancies in the samplers across the European continent currently provides monthly
model-measurement comparisons over Europe. The work has led to mean concentrations of HNO3 and HCl as well as SO2 and NH3 and
2-
modifications of the EMEP model (Simpson et al., 2003) to simulate aerosol NHþ 
4 , NO3 and SO4 (Tang et al., 2009), with a view to vali-
the temporal trends, resulting in an increase in Vd for SO2 over dating European concentration fields of concentration and deposi-
many parts of the continent, and driven by the long term, large scale tion for these species. The data show (Fig. 4.5) that the geometric
decrease in the SO2/NH3 ratio (Fig. 4.4). The new scheme, for non- mean mixing ratios of SO2, HNO3 and HCl across the network are 0.4,
stomatal resistance of both NH3 and SO2, incorporates an acidity to 0.35 and 0.15 ppb, respectively, so that, on average, SO2 makes up
alkalinity (SO2/NH3) molar ratio as a scaling factor for resistances. For only about 40% of the sum of acids (SO2 þ HNO3 þ HCl). Further, the
SO2, two non-stomatal resistances Rns,wet and Rns,dry are calculated as data indicate that at some sites (e.g. most Danish, French and Italian
a function of the SO2/NH3 ratio, and a function of relative humidity is sites), the acidity is largely dominated by HNO3 and HCl, which are
used for the transition from dry to wet when the surface cannot be considered in most models (e.g. Simpson et al., 2003) to be deposited
considered fully wet or fully dry. at the maximum rates allowed by turbulence (Rc w 0 s m1). Under
The EMEP non-stomatal deposition scheme has also been used such conditions, the proxy (SO2 þ HNO3 þ HCl)/NH3 would seem
in field-scale inferential modelling of N and S dry deposition as part more appropriate to quantify the relative importance of surface
of the NitroEurope project, using low-cost, long-term atmospheric acidity and alkalinity in model parameterisations, than the ratio of
trace gas and aerosol DELTA samplers (Tang et al., 2009). Another SO2 alone to NH3. Clearly the surface affinity for SO2 uptake will
implementation of the parameterisation was made by Zimmer- depend on the presence of fast-depositing, strong acids as the acidity
mann et al. (2006) for the simulation of atmospheric deposition to is no longer SO2-dominated, and this needs to be accounted for in
Norway spruce, using the SPRUCEDEP SVAT model, and comparison extended surface resistance parameterisations.
with throughfall measurements and a canopy base cation budget
model. The agreement between (inferential & canopy budget) 4.5. Future challenges
modelling and observations was very good for S and oxidized
N. Here, the contribution of dry to total (dry þ wet) deposition was The principal controls over SO2 deposition to terrestrial surfaces
around 60% for S and for both reduced and oxidized N. have been identified from field, mainly micrometeorological
The Dutch IDEM model (Bleeker et al., 2004; Erisman, 1994) also measurements. These studies have enabled the controlling steps in
uses an NH3/SO2 molar ratio as a proxy for surface acidity. For NH3, the deposition pathway to be separated and their response to
a range of default Rext values are used for 3 classes of the N/S ratio (very environmental variables quantified. In turn the data and responses
low, low and high), depending on surface wetness, land-use and time have been used to develop process-based models and applied to
of day, while for SO2 the only effect implemented is to add 50 s m1 to quantify regional deposition budgets at country and continental
the non-stomatal resistance when the N/S ratio is very low (<0.02). scales. There have been surprises, notably in the last decade. The
The large reduction in European SO2 emissions and ambient largest surprise has been the recognition that long term (w1 year)
concentrations over the last 25 years, and the relative stagnation in average deposition velocities change with time due to changes
NH3 emissions and concentrations in Western Europe over the same in the chemical climatology at the regional scale. Thus a few

Fig. 4.4. Modelled (EMEP) dry deposition velocity of SO2 (cm s1) over Europe for 1980 and 2000, taking into account the effect of the change in the SO2/NH3 ratio on the canopy
resistance (Fagerli pers com).
5218 D. Fowler et al. / Atmospheric Environment 43 (2009) 5193–5267

2.5
HCl

2 HNO3
SO2
Concentration (ppb)

1.5

0.5

0
CZ-BK1

NL-Ca1

PT-Mi1
CH-Oe1

FI-Hyy

FI-Lom
DE-Gri

DE-Hoe

NL-Spe
FR-Pue

IT-Col

PL-wet

UK-EBu
UK-ESa
DE-Hai
DE-Kli

FR-Gri

IE-Dri

IT-BCi

IT-Cas
DK-Ris

ES-VDA

FR-LBr

IT-MBo

IT-SRo

SE-Nor

DE-Meh
FR-Lq2

NL-Loo

SE-Sk2
UK-AMo
DE-Geb

ES-ES1

FR-Fon

HU-Bug

IT-Amp

PT-Esp
IE-Ca2

IT-Ren
IT-Ro2
BE-Vie

DE-Wet
DE-Tha
CH-Lae

DK-Lva

UA-Pet
FI-Kaa
BE-Bra

ES-LMa

RU-Fyo
FI-Sod

UK-Gri
FR-Hes

NL-Hor
BE-Lon

DK-Sor

3.5
SO2/NH3
Acid/NH3 ratio (ppb ppb-1)

3 (SO2+HNO3+HCl) / NH3

2.5

1.5

0.5

0
CZ-BK1
CH-Oe1

DE-Tha

ES-ES1

FI-Hyy
DE-Wet

FI-Lom
BE-Vie

ES-VDA

PL-wet

SE-Nor

UA-Pet
DK-Sor
DE-Gri
DE-Hai
DE-Kli

RU-Fyo

UK-EBu
FI-Sod

UK-Gri
DK-Ris

FR-Gri

IE-Dri

IT-BCi
IT-Col

NL-Hor

IT-Cas
BE-Bra
BE-Lon

FR-Hes
FR-LBr

IT-Ro2
CH-Lae

DE-Geb

FR-Pue

PT-Esp

UK-ESa
HU-Bug

IT-MBo

IT-SRo

NL-Loo

NL-Spe
DE-Meh
FR-Fon

FR-Lq2

IE-Ca2

IT-Amp

IT-Ren

NL-Ca1

SE-Sk2
UK-AMo
PT-Mi1
FI-Kaa
DK-Lva

DE-Hoe
ES-LMa

Fig. 4.5. Top: Annual mean concentrations of SO2, HNO3 and HCl across the NitroEurope network; Bottom: Annual mean Acid/NH3 molar ratios calculated from SO2 alone or from
SO2 þ HNO3 þ HCl.

measurements of SO2 deposition rates to parameterise models will Ozone deposition to external surfaces of vegetation is important
not be satisfactory in the long term if emissions of any of the acifi- as a removal pathway for ground level ozone but is of little
fying or alkaline gases change. It is necessary therefore to underpin consequence for plant effects. The primary potential for injury to
estimates of regional SO2 deposition with measured deposition vegetation, requires stomatal uptake of ozone molecules (Fig. 5.1)
fluxes and estimates of the surface resistance to quantify the long- followed by reaction with the internal plant tissue generating
term trends. The same logic means that deposition velocities from highly reactive oxidants that interfere with physiological processes
one region will not necessarily apply elsewhere. The most important (e.g. Matyssek et al., 2008). As ozone is a strong oxidant, it can also
region globally for sulphur emissions and deposition is currently react with leaf cuticles and other external plant surfaces or with
East Asia, and China specifically. While the methodologies and volatile compounds emitted by vegetation and non-stomatal ozone
models developed in Europe and North America are applicable for deposition is a substantial fraction of the total flux. In addition to
measurements of SO2 deposition, the use of parameters deduced in vegetation, ozone molecules may be deposited at any surface
Europe or North America are not applicable, and measures fluxes providing a chemical sink or acting as a surface for heterogeneous
and surface resistances in these regions are necessary to underpin decomposition (Cape et al., 2009). Quantifying the stomatal uptake
the assessments. rates is central to understanding the ozone-induced risk to vege-
tation, but the non-stomatal deposition needs to be quantified to
5. Ozone correctly partition the total deposition flux.
Over terrestrial surfaces ozone, confined within the atmospheric
5.1. Introduction boundary layer, has a relatively short lifetime scale of the order of
one day due to dry deposition at the surface (Wesely and Hicks,
Ozone is a gaseous, phytotoxic secondary air pollutant with 2000). Thus surface removal represents an important control on the
widespread effects on human health, vegetation and materials. It is near-surface ozone concentrations, and is the main cause of diurnal
also a greenhouse gas (GHG), third behind CO2 and CH4 in impor- variation in rural areas (Garland and Derwent, 1979). Dry deposition
tance. Its deleterious effects on plants pose a large-scale risk to crop constitutes a major term in the global mass balance of tropospheric
production and forest vitality in many regions of the Northern ozone, the mean global dry deposition sink calculated with 20
Hemisphere (Fowler et al., 1999; Cape, 2008), which have been chemistry-transport models (CTMs) (1000 Tg y1) clearly exceeding
widely studied in Europe and North America (e.g. Hayes et al., 2007; the net stratospheric input (550 Tg y1) (Stevenson et al., 2006).
Karnosky et al., 2007), there is also evidence of ozone impacts in Asia, The first measurements of ozone deposition fluxes were made
Africa and Latin America (e.g. Ashmore, 2005). in the 1950s using the micrometeorological gradient method
D. Fowler et al. / Atmospheric Environment 43 (2009) 5193–5267 5219

Fig. 5.1. The sinks for ozone at terrestrial surfaces and processes regulating the exchange.

(Regener, 1957). The earliest investigations were aimed at quanti- (Garland and Derwent, 1979). The first long-term measurements
fying the surface sink term of the tropospheric ozone budget. Based showed a seasonal cycle over vegetated surfaces that followed the
on these, Galbally (1971) concluded that bulk surface resistance (Rc) growing season of the plants (Colbeck and Harrison, 1985), leading
of dry soil and short grass surfaces was approximately 100 s m1. to the assumption that ozone deposition is mainly controlled by
These and other pioneering studies (e.g. Turner et al., 1974) showed stomatal uptake and deposition to non-vegetated surfaces is
that vegetation and soil constitute important pathways by which constant, only depending on the material and surface area. More
ozone is removed from the atmosphere, while water and snow recent studies have shown that although deposition rates are partly
surfaces are rather inefficient sinks. However, the early studies governed by stomatal uptake over a plant canopy, it only accounts
provide rather little information about the processes that control for ca 40–60% of total deposition on average and that the non-
ozone deposition to terrestrial or marine surfaces. stomatal component is not constant (Fowler et al., 2001a; Coyle
There has also been an interest in measuring surface fluxes et al., 2009; Hogg et al., 2007). These observations are described in
prompted by ecological concerns. The importance of environ- more detail in the following sections.
mental conditions on plant injury due to ozone through the regu-
lation of stomatal uptake was recognised in 1960s by Mukammal 5.2.1. European forests
(1965), who observed that the presence of high concentrations was Ozone deposition fluxes to forests, as well as other vegetated
not a sufficient condition for plant injury. Indeed, a few years later surfaces, are largely controlled by the physiological activity and
the close coupling between ozone and water vapour fluxes was associated gas exchange of the vegetation, with solar radiation, air
demonstrated by Rich et al. (1970). However, while ozone effects on temperature, air humidity and soil moisture as the main controlling
vegetation have been closely associated with stomatal uptake for variables. Thus the deposition velocities (Vd) observed above forests
decades, only recently have practical risk assessment methods been typically exhibit diurnal and seasonal cycles that depend on the
formulated in terms of stomatal uptake rather than ambient structure, physiological responses and phenological state of the trees.
concentration (UNECE, 2004). According to present understanding, however, there are other signif-
Micrometeorological techniques have been in use since the icant processes in addition to stomatal regulation of gas exchange that
first flux measurements. In the late 1970s, Eastman and Stedman control the magnitude and variation of the ozone deposition efficiency
(1977) developed a fast-response ozone sensor that facilitated of forests (Altimir et al., 2006; Cieslik, 2004; Dorsey et al., 2004;
direct ozone flux density measurements by the eddy covariance Goldstein et al., 2004; Hogg et al., 2007; Lamaud et al., 2002; Tuovi-
method. This resulted in a series of measurement campaigns in nen, 2000; Tuovinen et al., 2008a; Zhang et al., 2002).
the eastern United States, including various vegetated and The temporal patterns are clearly demonstrated by the long-
other surface types (Wesely, 1983). These studies improved the term flux measurement data available from a few sites, such as
understanding of deposition processes and formed the basis for those reported for a temperate spruce forest by Mikkelsen et al.
the detailed surface resistance parameterisation of Wesely (2004) and a boreal pine forest by Keronen et al. (2003) and Altimir
(1989). This parameterisation has been implemented into et al. (2006). In the boreal region in winter, the dormancy of
numerous CTMs. In Europe, the eddy covariance technique was vegetation and below-zero temperatures result in low and rela-
adopted somewhat later, but rapidly became popular with the tively stable deposition rates (Vd z 0.1 cm s1, Fig. 5.2a), while the
introduction of a commercial fast-response ozone sensor mean diurnal cycle at the temperate forest shows a weak midday
(Güsten et al., 1992). enhancement related to gas exchange superimposed on a relatively
high and seasonally invariable base level (Vd z 0.5 cm s1). The
5.2. Deposition rates decrease of Rc in spring correlates with the onset of photosynthesis,
and a well-defined, symmetrical diurnal cycle ensues in summer,
Early estimates of ozone dry deposition were obtained from with a Vd maximum of 0.7–0.9 cm s1 (Keronen et al., 2003; Mik-
measurements of the diurnal cycle of ozone in rural areas kelsen et al., 2004). However, Altimir et al. (2006) observed that the
5220 D. Fowler et al. / Atmospheric Environment 43 (2009) 5193–5267

a Boreal Pine Forest Median b Temperate Oak Forest


0.7 Winter (Nov-Feb) 1.4
0.6 Summer (May-July) 1.2 July -A ug
0.5 1.0

vd [cm s-1]

vd [cm s-1 ]
0.4 0.8
0.3 0.6
0.2 0.4
0.1 0.2
0.0 0.0
:00
:00
:00
:00
:00
:00
:00
:00
:00
:00
:00
:00
:00

:00
:00
:00
:00
:00
:00
:00
:00
:00
:00
:00
:00
00
02
04
06
08
10
12
14
16
18
20
22
00

01
03
05
07
09
11
13
15
17
19
21
23
c Potatoes d Temperate Grassland
1.4 0.7
1.2 July-Aug 0.6 Apr-Sep
vd [ cm s -1 ]

1.0 0.5

vd [cm s -1 ]
Oct-Mar
0.8 0.4
0.6 0.3
0.4 0.2
0.2 0.1
0.0 0.0
0 1:00
0 3:00
0 5:00
0 7:00
0 9:00
1 1:00
1 3:00
1 5:00
1 7:00
1 9:00
2 1:00
2 3:00

0 0:00
0 2:00
0 4:00
0 6:00
0 8:00
1 0:00
1 2:00
1 4:00
1 6:00
1 8:00
2 0:00
2 2:00
0 0:00
Fig. 5.2. Median diurnal cycles in deposition velocity at (a) boreal scots pine forest, Hyytiala, Finland, 2002–2003 (Altimir et al., 2006), (b) oak forest, Alice Holt, England, 16th
July-05–18th August-05 (Coyle et al., 2006), (c) potatoes, Gilchriston, Scotland, 9th July-06–3rd Aug-06, (Coyle et al., 2009) (d) intensively managed lolium perene grassland, Easter
Bush Scotland, 2002–2003 (Coyle, 2005).

correlation with physiological activity is poorer in autumn and occurrence and persistence of drought. For example, in August
concluded that the deposition rate is modified by the frequent 2003 the mean Gst derived for the Italian oak forest by Gerosa et al.
wetting of the forest canopy. Both Mikkelsen et al. (2004) and (2008) was only 35% of that in the cooler and wetter August of
Altimir et al. (2006) attribute a major part of the total annual ozone 2004, and the effect of drought persisted even after soil water was
deposition to non-stomatal pathways. replenished by rainfall.
High non-stomatal fluxes are also observed in Mediterranean
forests (Gerosa et al., 2005, 2008). However, the diurnal and 5.2.2. Crops
seasonal variations differ from those of the northern forests in many For agricultural crops, ozone deposition rates exhibit pronounced
aspects. The measurements above oak forests in central Italy seasonality that results from the distinct phenological stages of the
(Gerosa et al., 2005, 2008; Cieslik, 2009) and south-eastern France growing season. The flux measurements above an Italian barley field
(Michou et al., 2005) show that dry and hot conditions can signifi- by Gerosa et al. (2004) demonstrate how Vd increases during the first
cantly affect the diurnal courses of Vd and Rc. On the other hand, in growth stages (seedling growth, tillering, stem elongation). The
this region there is a potential for high deposition rates throughout maximum is reached soon after anthesis, during the grain filling
the year, and higher stomatal uptake may take place during winter period, when photosynthetic activity is at its highest level. Gerosa
than summer months, in spite of the lower ozone concentrations in et al. (2004) observed an average Rc of about 75 s m1 for this period,
winter (Cieslik, 2009). while the bulk stomatal resistance (Rst) was about 150 s m1. After
During dry periods stomata are either almost completely closed that, deposition rates decrease gradually with ripening of the crop
or the cycle of stomatal conductance (Gst ¼ R1 st ) is less symmetrical, and leaf senescence, as Rst increases, and are further reduced by
with a rapid increase from the nocturnal levels to a maximum in harvest. A slightly higher minimum Rc but a similar decrease was
the morning, rather than around noon, and a gradual decrease observed at the same site for wheat. In this case, the decline was
towards the evening. The latter behaviour may take place in more amplified by the rapid drying of soil (Gerosa et al., 2003).
northern forests as well, if leaf temperatures reach sufficiently high For both barley and wheat, the diurnal cycles of Vd and surface
levels (e.g. Dorsey et al., 2004; Tuovinen et al., 2008a). At the French conductances were symmetrical during the photosynthetically
site, Vd starts increasing very early (at 3–4 a.m. local time) and it active period with a maximum (Vd z 0.7–0.9 cm s1) around noon
was suggested that this could be related to either stomatal response (Gerosa et al., 2003, 2004). During the latter part of the growing
to blue light, resulting in uptake already in the predawn hours, or to season, the midday deposition rates are strongly reduced, due to
non-stomatal deposition enhanced by surface wetness (Michou increased Rst, resulting in an earlier maximum and a diurnal course
et al., 2005). At the Italian site, however, the maximum (in Gst) that is skewed towards morning. In the afternoon, values compa-
occurs later and seems to have a non-stomatal origin, possibly due rable to those after harvest were observed for the barley field
to the reaction with the NO accumulated within the forest canopy (Gerosa et al., 2004).
(Gerosa et al., 2005) or leaf wetness (Gerosa et al., 2008). Similarly, For other crops, data are rather limited. Michou et al. (2005)
the measurements taken by Coyle et al. (2006) in an oak forest report a symmetrical diurnal cycle for the Vd measured over
in England in summer show a steep increase to the maximum at a rapidly growing maize field, with a mean minimum of 0.05 cm s1
8 a.m. (Vd z 1.0 cm s1) and a approximately linear decrease to and a mean maximum of 0.50 cm s1. In an earlier North American
a nocturnal level (Vd z 0.1 cm s1), Fig. 5.2b. study, a similar diurnal cycle, with slightly higher values throughout
In addition to the phenological development of plants, the the day, was observed over a maize field during the period of most
seasonal cycle can be strongly influenced by the soil moisture active growth (Meyers et al., 1998). During senescence, the hourly
conditions, especially in southern Europe. Consequently, there may mean Vd only reached 0.2 cm s1 in the morning. The patterns were
be large annual variations in ozone uptake rates depending on the also similar for soybean, but the Rc of soybean seems to be
D. Fowler et al. / Atmospheric Environment 43 (2009) 5193–5267 5221

significantly lower than that of maize, as the maximum mean primary productivity at modest O3 exposure concentrations
hourly Vd was almost 1 cm s1 during the active growth period (King et al., 2005).
(Meyers et al., 1998). Coyle et al. (2009) reported an asymmetrical
diurnal cycle in Vd over a potato field during the phase of tuber 5.2.4. Other vegetated surfaces
initiation and growth through to harvest, with a median value of Measurements have been made over a variety of other plant
0.6 cm s1, day-time values from 0.5 to 2 cm s1 decreasing to canopies from moorlands and subartic mires to tropical forests. They
0.4 cm s1, or less, at night (Fig. 5.2c). all exhibit diurnal and seasonal cycles driven by variations in
As was the case with forests, stomatal uptake rates derived from stomatal activity and climate, as described for the previous canopies.
the evapotranspiration fluxes only explain a part of the total ozone For example Fowler et al. (2001) for moorland reported summer
flux and a significant proportion must be attributed to non-stomatal diurnal cycles ranging from w0.3 cm s1 at night to w0.6 cm s1
sinks. Even during the active growth period, no more than 50–60% during the day while in the winter the afternoon peak was only
of the total flux to a wheat field was stomatal, and this fraction w0.4 cm s1 with night-time values also w0.3 cm s1. Tuovinen et al.
decreased during the senescence (Gerosa et al., 2003). The same was (1998) measured a small diurnal cycle over flark fen, 300 km north of
true for barley, and even though the day-to-day variation in the the Artic circle in the late summer, ranging from 0.1 to 0.15 cm s1 at
inferred non-stomatal flux fraction is large, the corresponding non- night to only w0.2 cm s1 during the day. For tropical forests there
stomatal surface conductance remains relatively stable throughout are often two seasons, wet and dry: During the wet season, Rummel
the summer (Gerosa et al., 2004). Similar results were obtained for et al. (2007) reported diurnal cycles over Amazonian rain forest
onion, but in this case the variation can be explained by irrigation ranging from w0.4 cm s1 during the night to w1 cm s1 at day, with
that clearly enhanced both stomatal and total fluxes (Cieslik, 2009). a symmetrical diurnal cycle.
In the case of potatoes, the non-stomatal component was only w15%
when the canopy was well-watered but increased to w80% when the 5.2.5. Non-vegetated surfaces
crop became senescent. (Coyle et al., 2009). Deposition to the soil underlying the vegetation layer may signifi-
cantly contribute to the vertical ozone flux observed above the canopy
5.2.3. Grasslands (e.g. Dorsey et al., 2004). Especially in arid regions, the surface resis-
Grasslands can be used to describe a wide variety of habitats tance of soil (Rsoil) is the key determinant of the surface removal of
from intensively managed pastures that are usually dominated by ozone (e.g. Michou et al., 2005). As the literature review by Massman
a single species, such as Lolium perenne, to natural grasslands which (2004) and the more recent data of Sorimachi and Sakamoto (2007)
contain a rich diversity of grasses, forbs and legumes and are often indicate, Rsoil is highly variable, ranging from 10 to >1000 s m1. It has
of high conservation value. The canopies that have been studied to been observed that Rsoil decreases with increasing organic content
date all exhibit similar patterns of deposition as forests and agri- and porosity of soil. Clearly, wet soils have a considerably higher Rsoil
cultural crops in that they have phenologically driven seasonal and (w500 s m1) than dry soils (w100 s m1) (Galbally and Roy, 1980;
diurnal cycles. There are few long-term measurements of ozone Massman, 2004), as increasing the moisture content of soil decreases
deposition to grasslands reported in the literature at present its porosity and thus reduces the area of reactive surface available to
(Colbeck and Harrison, 1985; Grünhage and Jäger, 1994; Pio et al., ozone molecules (Sorimachi and Sakamoto, 2007). However, it is
2000). All these studies measured during winter and summer so difficult to disentangle individual effects based on the current data, and
that the growing season and dormant periods were observed. The the situation is further complicated by biogenic NO emissions from
results of several short-term campaigns over active and dormant or soils. Removal through the reaction with NO potentially constitutes
dry grasslands have also been reported and are summarized below a significant sink for ozone, especially in forests at night (Pilegaard,
(Bassin et al., 2004; Garland and Derwent, 1979; Meyers et al., 1998; 2001; Dorsey et al., 2004), even though this reaction does not take
Sorimachi et al., 2003). place specifically at the air–soil interface.
Over active, green grasslands the daytime Vd is only 0.5 cm s1
on average, decreasing to w0.1–0.2 cm s1 at night, although peaks 5.2.5.1. Snow. The dry deposition velocity over snow- and ice-
over 1 cm s1 are often observed (Fig. 5.2d). The diurnal cycle is covered surfaces and water is known to be smaller than 1 mm s1.
often symmetrical with a steady increase in the morning after However, the measurement data are highly variable. It is possible
sunrise and decrease is the afternoon as light and temperatures that the measured Vd is affected by chemical reactions taking place
decrease. However (Grünhage and Jäger, 1994) reported an asy- in the snowpack, which, combined with dynamic transport
metrical diurnal cycle in Vd with a steep increase after 6 a.m. until 9 processes, can result in the observed variability. Helmig et al.
a.m. when it steadily declined. This is attributed to an increase in (2007) modelled ozone concentrations in high northern latitudes
water vapour pressure deficit (VPD) at midday and the afternoon with different values of Rc and demonstrated how even small
causing stomata to close, as has been observed for many other deposition rates, if effective over large areas, can significantly affect
canopies. Where measurements have been made over dormant the near-surface concentrations and that a high Rc is required for
(i.e. during winter) or dry grasslands the diurnal cycle is far less snow in order to reproduce the observed concentrations; the best
pronounced with daytime Vd only reaching w0.2 cm s1 although agreement was obtained when limiting Vd to 0.01 cm s1.
night-time values are similar at all times of year (Fig. 5.2d).
Coyle, (2005) showed that non-stomatal uptake was w60% of 5.2.5.2. Water. For water surfaces, current understanding of
the total budget over an improved grassland in Central Scotland. controlling processes is more coherent, involving both turbulent and
Pleijel et al. (1995) found that the non-stomatal sink is enhanced by molecular mixing, and chemical reactions. It has been observed that
surface wetness while the work of Coyle (2005) also demonstrated wind-induced turbulence greatly enhances deposition rates by
that the non-stomatal component was not constant but varied with increasing atmospheric mixing, surface roughness, wave breaking
wetness, surface temperature, solar radiation and wind speed as and spray generation (e.g. Gallagher et al., 2001). The observations of
has been suggested for other canopies. Rc range from 1000 to 10,000 s m1 (Gallagher et al., 2001). At low
Ozone deposition to forests has also been extensively studied in wind speeds, Vd increasingly depends on the molecular gas transfer
North America, and in recent years a unique study in elevated CO2 near the air–water interface. Even in the absence of turbulence,
using a Free Air Enrichment Experiment (FACE) has demonstrated significant deposition rates are possible, since the chemical reactions
strong interactions between O3 and CO2, the former decreasing taking place in the aqueous phase enable more efficient deposition
5222 D. Fowler et al. / Atmospheric Environment 43 (2009) 5193–5267

than the solubility of ozone alone would suggest. The modelling logarithm of the surface resistance in s m1 is taken as the inverse
results of Fairall et al. (2007) show that the oceanic turbulent mixing reaction rate). The results for four vegetated canopies are also plotted
also plays an important role in enhancing ozone deposition by up to in Fig. 5.3 and the slope of their linear regression lines is of the
a factor of three. Iodide (Chang et al., 2004) and chlorophyll (Clifford same order of magnitude as the artificial surfaces. This shows that the
et al., 2008) have been suggested as the main reagents controlling dependence on temperature and activation energies are similar for all
ozone destruction in the organic surface microlayer. As the distri- surfaces but the absolute reaction rates differ. The simplest conclusion
bution of these compounds in water bodies is related to that of is that heterogeneous decomposition of ozone to molecular oxygen is
phytoplankton, the chemical enhancement of ozone deposition can the underlying process responsible; hence the similar activation
be expected to be highly variable both temporally and spatially. energies in Fig. 5.3 and the variation in reaction rates can be attributed
Similarly, coastal ozone deposition to the iodide-rich macroalgae to differences in effective surface area (Cape et al., 2009).
surfaces depends on the tidal phase, and fluxes can be further Cape et al. (2009) also tested the hypothesis that the surface
enhanced by photochemical destruction of ozone during the iodine- reactivity of vegetation may be enhanced by reaction of biogenic
mediated particle formation events (Whitehead et al., 2009). volatile organic compounds (BVOCs) dissolved in cuticular waxes
and showed no enhancement due to surface reaction. However,
5.3. Non-stomatal deposition processes ozone does react in the gase phase with BVOCs emitted by vege-
tation, with gas-phase reaction rate coefficients varying between
Although the stomatal uptake of ozone is an important sink over 1018 and 1016 cm3 molec1 s1, potentially leading to apparent
vegetated surfaces it accounts for only a fraction (typically 1/3–2/3) of non-stomatal deposition velocities of similar magnitude to those
the total deposition. In most studies it has been assumed that the measured in the field (Coyle, 2005). The emission of BVOCs by
non-stomatal sink is constant, depending only on the surface mate- vegetation is also light and temperature dependent, increasing
rial and area, although it was expected that the presence of surface with both parameters which may explain some of the variation in
water would inhibit deposition as the solubility of ozone is quite non-stomatal deposition. Nevertheless, most measurements indi-
small. However, field measurements have indicated that surface cate that the concentrations or reactivity of the emitted BVOCs are
temperature, solar radiation, surface wetness and wind speed may all not sufficient to explain the magnitude or variation in non-stomatal
influence the magnitude of the non-stomatal flux. The influence of deposition in all circumstances. In can be concluded that although
wind speed is simply explained, as when wind speed increases more BVOCs may play a part in the non-stomatal deposition they are only
air will penetrate the canopy, increasing the surface area available for significant for species that emit currently unidentified compounds
deposition. The mechanisms that have been proposed for the other that react very rapidly with ozone (Hogg et al., 2007).
parameters are: Although some studies have shown surface water inhibits
ozone deposition to a vegetated canopy the consensus is now that
C Temperature: thermal decomposition ozone on plant surfaces, it increases deposition (Hogg et al., 2007; Coyle et al., 2008a) in
mediated by waxes and other substances on the surface most circumstances. Although Fuentes (1992) found more organic
(Coyle et al., 2009; Hogg et al., 2007; Fowler et al., 2001; Cape compounds in water from maple leaves compared to poplar, specific
et al., 2009) compounds that may be responsible have not been identified.
C Solar radiation: ozone photolysis also mediated by the surface Coyle et al., (2009) suggested that non-stomatal deposition is gov-
(Coyle et al., 2009; Hogg et al., 2007 and references therein) erned by three main regimes: ozone deposition increasing as the
and reaction with VOCs emitted by vegetation (Hogg et al., temperature and solar radiation increases on a dry surface due to
2007; Coyle, 2005) thermal decomposition; decreased deposition on surfaces with
C Surface wetness: aqueous reactions in water-films on plant a thin film of water present as thermal decomposition is blocked by
surfaces (Coyle et al., 2009; Altimir et al.,) the water film; enhanced deposition on a fully wetted surface due to
aqueous reactions in the water.
Principal component analysis of data from several field While non-stomatal deposition has become widely recognised
campaigns indicated the following order of precedence: tempera- as an important sink, the concept of stomatal uptake at night is less
ture, surface wetness (measured as vapour pressure), solar radia- widely appreciated and this complicates the simplistic view of day–
tion then wind speed although for forests temperature and wetness night differences in ozone sinks in vegetation (Grulke et al., 2004).
were of almost equal importance (Coyle et al., 2008a).
Some preliminary work exposing wax coated and metal surfaces to 5.4. Model development and validation
ozone in controlled environment chambers (Cape et al., 2009) showed
that Vd is temperature dependent. An Arrhenius analysis of the change The recent European measurement data on ozone fluxes have
in reaction rate with temperature is shown in Fig. 5.3 (the natural been used for testing and improving the DO3SE deposition

14 stainless steel y = 40.1x - 3.28, R2 = 0.84


aluminium foil y = 24.9x + 0.14, R2 = 0.29
12 paraffin wax y = 22.8x + 0.85, R2 = 0.18
ln(Rns)

10 beeswax y = 16.1x + 2.18, R2 = 0.44


moorland vegetation y = 36.4x - 9.62, R2 = 1
8 Potatoes y = 49.9x - 14.84, R2 = 0.06
6 Grassland y = 52.6x - 16.26, R2 = 0.06
Forests y = 56.2x - 16.56, R2 = 0.14
4
0.39 0.4 0.41 0.42 0.43
1000/RT

Fig. 5.3. Arrenhius relationship analyses for ozone deposition to various surfaces: stainless steel, Aluminium foil, paraffin wax and beeswax from Cape et al. (2009); moorland
Fowler et al. (2001); potatoes Coyle et al. (2009); Grassland and Forests, Coyle et al. (2008a).
D. Fowler et al. / Atmospheric Environment 43 (2009) 5193–5267 5223

module, which has been incorporated into the Unified EMEP CTM AOT40 is still the most common risk indicator used in Europe for
developed for European policy-making applications (Tuovinen setting environmental objectives and defining the so-called critical
et al., 2001, 2004, 2008b; Emberson et al., 2007). DO3SE makes it levels, Ashmore and Fuhrer (2000). For a proper application of AOT40,
possible to calculate stomatal and non-stomatal ozone fluxes to ozone concentration must be known at the height of the canopy
different vegetation types, taking into account both plant top (UNECE, 2004). This means that the concentration determined
phenological and meteorological factors (irradiance, temperature, above this height (as is typically the case with measurements) must
humidity, soil moisture) on an hourly basis (Simpson et al., 2007; be transformed to the correct reference height, because of the
Emberson et al., 2000, 2007). Many of the validation studies deposition-sink generated vertical concentration gradient; a failure
have been focussed on the leaf-scale stomatal conductance of to correct for this gradient may seriously overestimate the risk
a specific plant species only, since parameterisations of this kind are metrics (Tuovinen and Simpson, 2008b). Even if the profile correction
needed to relate the ozone uptake to plant response (e.g. Pleijel compensates for a significant bias, the near-surface concentration
et al., 2007). There are also smaller-scale CTMs covering a part of may prove a poor surrogate for the effect-inducing flux. High ozone
Europe, which have been used for high-resolution mapping of ozone concentrations are frequently connected with conditions that poten-
fluxes (e.g. Lagzi et al., 2004), and local-scale soil–vegetation– tially limit the stomatal uptake, such as high temperature and VPD
atmosphere-transfer models that include ozone (e.g. Grünhage and (e.g. Solberg et al., 2008). This co-variation means that the concen-
Haenel, 2008). Inferential modelling (IM) has been used for tration at leaf surface is not necessarily connected with a proportional
calculating regional ozone budgets (Coyle et al., 2003) and mapping stomatal uptake and therefore plant response (Cieslik, 2004). This is
exposures and doses on a high (1–2-km) spatial resolution for one reason for the different accumulation rates of AOT40 and stomatal
national-scale risk assessment of ozone effects (e.g. Keller et al., uptake observed in many studies (Fig. 5.4).
2007). As a compromise between the less data demanding exposure
The global-scale CTMs (Stevenson et al., 2006), as well as many indices and the more realistic dose metrics, solutions based on an
regional models (e.g. Vautard et al., 2005), typically include a variant ‘effective’ concentration have been suggested (e.g. Gerosa et al.,
of the parameterisation of Wesely (1989). This parameterisation 2004; Karlsson et al., 2004; Pleijel et al., 2004). Common to all these
does not include the stomatal effect of soil moisture conditions, is the idea that the concentrations for AOTX are first modified to
which has been shown to be a significant modifier of ozone fluxes, accommodate environmental factors controlling stomatal uptake,
especially in the Mediterranean region (Gerosa et al., 2008), yet such as VPD in the definition of the modified AOT30. In some cases,
proved challenging to model within CTMs (Emberson et al., 2007). all the main modifiers, such as those in the DO3SE model, are
Another key problem with large-scale CTMs is related to the aggre- considered (Gerosa et al., 2004; Pleijel et al., 2004). However, this
gated land cover classes, making it difficult to parameterise sub-grid
processes. For example, the dry deposition module of the recently
developed global Modular Earth Submodel System (MESSy) does not
differentiate between different vegetation types, even though a soil
moisture stress function is included (Kerkeweg et al., 2006).
The non-stomatal deposition processes are parameterised in
CTMs and IM systems in a much simpler way than the stomatal
component, typically by defining constant values for the relevant
resistances. However, a preliminary parameterisation has been
developed for the EMEP CTM for surface wetness effects for
northern European coniferous forests (Tuovinen et al., 2008). This
parameterisation is derived from the observations of Altimir et al.
(2006) and represents a gradual increase in the surface sink with
increasing surface wetness, parameterised as a function of relative
humidity. This results in higher and more variable ozone removal
rates within the model. In the future, integrated models are needed
to couple the surface exchange of energy, carbon and trace gases.
In particular, a multi-layer structure facilitating an explicit simu-
lation of vertical mixing and other in-canopy processes would be
useful for interpreting flux measurements (e.g. Duyzer et al., 2004;
Simon et al., 2005a).

5.5. Risk assessment methods

European abatement strategies are founded on effects-based


approaches, which involves different numerical indicators for
different air pollution effects on vegetation and human health
(UNECE, 2004). For potential ozone effects on vegetation, the
AOTX (Accumulated exposure Over a Threshold of X ppb) expo-
sure index replaced simpler concentration averages in the 1990s
(Fuhrer et al., 1997). Present definitions also include a metric
based on the stomatal uptake flux, AFstY (Accumulated stomatal
flux Fst above a threshold of Y nmol m2 s1), which represents
the absorbed dose and is thus considered biologically more
meaningful than the concentration-based AOTX (UNECE, 2004).
AFstY is more complex than AOTX in that it entails modelling the Fig. 5.4. Accumulation of AOT40 and stomatal dose over a Holm oak forest (6 August–11
stomatal conductance. October) in 2003 and 2004 in Italy (Gerosa et al., 2008).
5224 D. Fowler et al. / Atmospheric Environment 43 (2009) 5193–5267

approach is very close to actually calculating the stomatal flux, as it changes in the meteorological input data. This resulted in
involves a multiplicative stomatal conductance model. Even the a decreased dry deposition sink, which increased the near-surface
profile corrections required for AOTX are based on flux–gradient ozone concentrations by 2–8 ppb on a seasonal basis. However,
relationships of ozone and thus entail deposition modelling (Tuo- a reduction in Gst, due to increased CO2 or any other effect, does not
vinen and Simpson, 2008b). From a micrometeorological point of proportionally reduce the stomatal uptake flux. This results from the
view, it would thus appear natural to aim at developing accurate effect of decreased surface removal on the overall mass balance, with
parameterisations for partitioning the total ozone flux into the higher concentrations partly compensating for the suppressed Gst.
stomatal and non-stomatal components and applying flux-based While it would seem plausible that in warm climates increasing
risk metrics because of their superior biological basis. temperatures reduce stomatal exchange, the opposite is true for
The model calculations of Simpson et al. (2007) indicate that cooler regions. However, based on a modified version of the DO3SE
AOT40 and AFstY (for both crops and forests) show very different model, Karlsson et al. (2007) concluded that in northern Scandinavia
regional patterns of exceedance of critical levels across Europe with the most significant impact of the higher temperatures may be
much smaller south–north gradients and larger exceedance area related to an earlier onset of the growing season and the associated
for AFstY (Fig. 5.5). Even though there are still numerous uncer- phenological development, rather than their direct enhancement of
tainties involved in this kind of modelling, there is evidence that Gst. This together with elevated and increasing ozone concentrations
the flux-based risk maps better correlate with observed plant in spring may amplify the risk of negative ozone effects on vegeta-
damage (Hayes et al., 2007). tion in these areas. On the other hand, there may be a counteracting
effect on the stomatal uptake mediated by the concurrent higher
5.6. Potential effects of climate change VPD (Karlsson et al., 2007; Harmens et al., 2007).
The summer of 2003 was exceptionally warm in Europe, espe-
Changes in the climatic conditions and chemical composition of the cially in the central part of the continent, and may be taken as an
atmosphere are expected to have a wide range of effects on the analogue of the future summers to be expected in the latter part of
interactions between tropospheric ozone and the biosphere. First, the 21st century. A series of heat waves produced meteorological
ozone exposures are changing globally due to changes in precursor conditions highly favourable for the net formation of ozone and its
emissions. Second, the long-term changes in meteorological condi- build-up over large areas; indeed, record-high near-surface concen-
tions affect atmospheric transport patterns and the rates of tropo- trations were observed in many locations (Solberg et al., 2008). It is
spheric chemical reactions and dry deposition processes, and also very likely that reduced dry deposition played a significant role in the
modify plant phenology. In addition to the rising temperature and formation of the ozone episodes in 2003. The high temperatures and
changes in precipitation distribution, elevated CO2 and ozone soil moisture deficits (SMDs) most probably decreased the Gst of
concentrations may act as significant modifiers to stomatal exchange vegetation and thus ozone removal from the atmosphere in central
(e.g. Ashmore, 2005). Finally, the characteristics of vegetation cover and southern Europe, as indicated by the Italian eddy covariance
and land use may be altered on various scales as a result of human measurements (Gerosa et al., 2008) discussed above. Fig. 5.4 shows
activities, effects of climate and ozone on plant species composition that the ozone dose absorbed by a Holm oak forest in August–
and ecosystem function, and natural disturbances, all of which September 2003 was less than 50% of that during the same period in
potentially generate feedbacks to the surface removal of ozone. 2004, in spite of the much higher concentrations in 2003.
So far, few studies have addressed the projected changes beyond The sensitivity runs with a global CTM by Solberg et al. (2008)
the ozone precursor emissions and atmospheric dynamics. In the demonstrate the potentially large effect of dry deposition on near-
multi-model ensemble simulations of future concentrations by surface concentrations. Similarly, doubling of Rc within the model-
Stevenson et al. (2006), the projected Vd was only altered by the ling study of Vautard et al. (2005), partly because of the expected
meteorological responses, neglecting the effects of water stress and increase in SMD which is not taken into account in the deposition
elevated CO2 concentrations, for example. In some studies, indi- parameterisation, improved the model performance by increasing
vidual, uncoupled effects have been included in the models. For the modelled concentrations. Considering the accumulation of
example, Sanderson et al. (2007) investigated the impact of the ozone dose of plants over the whole growing season, the significance
stomatal conductance changes as directly induced by the rising CO2 of drought periods much depends on their timing with respect to the
levels. In this experiment, a global CTM was run with an assumed phenological stage of plants and the occurrence of elevated ozone
reduction in Gst due to a doubling of CO2 concentrations, but with no concentrations (Harmens et al., 2007). In addition, prolonged

Fig. 5.5. Ratio of the AOT40 (left) and AFst1.6 (right) risk metric to the corresponding critical level for forests (not shown for values < 1) (Simpson et al., 2007).
D. Fowler et al. / Atmospheric Environment 43 (2009) 5193–5267 5225

drought stress may result in sustained impairment of the hydraulic 6.1.2. Oxygenated volatile compounds
conductivity of plants, challenging the traditional dry deposition While the production and emission of volatile isoprenoids, and
models (Gerosa et al., 2008). in particular isoprene and monoterpenes, is strongly species-
It has been estimated that exposure of plants to even the current specific, all plants emit oxygenated volatiles. At the global scale, the
levels of ozone may significantly increase water use of forest trees emission source strength of these compounds is generally smaller
(McLaughlin et al., 2007) and reduce plant productivity in the most than that for volatile isoprenoids, and many of these oxygenated
polluted areas of the world, with exacerbating effects projected for compounds are less reactive than isoprenoids in the atmosphere.
the future (Felzer et al., 2005). With elevated ozone concentrations, However, the emissions of oxygenated compounds, which may
the reductions in carbon sequestration may also lower soil carbon be induced by developmental and stress factors, may be large at
formation rates and alter the below-ground carbon cycling (Loya certain periods of the year and by certain vegetation types (see also
et al., 2003). Sitch et al. (2007) suggest that the ozone-induced flux measurements below).
suppression of the global land-carbon sink gives rise to additional Methanol, acetaldehyde and C-6 compounds are often emitted in
accumulation of anthropogenic CO2 emissions in the atmosphere large quantities, especially in the presence of mechanical wounding
and thus should be considered an indirect radiative forcing, which or other stresses (Loreto et al., 2006). Methanol formation is prob-
could exceed the direct radiative forcing due to ozone increases. ably due to the demethylation of pectins in cell walls (Galbally and
Kirstine, 2002) and does not have any known protective role for
plants. The release of methanol into the atmosphere is therefore
6. Biogenic volatile organic compounds (BVOC) associated with cell wall damage occurring because of wounding
(Karl et al., 2001a). Methanol is also emitted by growing plant tissues
6.1. Introduction (Nemecek-Marshall et al., 1995; Harley et al., 2007; Hüve et al.,
2007), and senescing tissues (Fall, 2003). Large fluxes of methanol
Biogenic volatile organic compounds (BVOC) are emitted by could be measured, as an example, from rapidly expanding leaves of
almost all plants. In higher plants, emissions range from close to zero the Mediterranean vegetation during the spring ACCENT–VOCBAS
to 10–20% of the carbon fixed by photosynthesis. Global emissions 2007 campaign (see below).
have been estimated at around 800 Tg C y1 although this figure Large fluxes of acetaldehyde have been observed in conditions
has been regularly revised based on improvements in scaling up of of root anoxia such as under waterlogging stress. Short-lived bursts
laboratory results, spatial and temporal integrations, and large-scale of acetaldehyde are sometimes also observed from darkened leaves
monitoring at whole ecosystem levels (Lathiere et al., 2006; Arneth (Karl et al., 2002b). Interestingly, acetaldehyde is also emitted
et al., in press). Approximately half of the emissions are believed to following wounding (Fall et al., 1999) and ozone stress episodes,
be isoprene (Guenther et al., 2000, 2006). Monoterpenes, the other and large fluxes of this compound can be observed under natural
large class of volatile isoprenoids, contribute another 10–15% of the conditions (Lathiere et al., 2006).
total BVOC emissions. Sesquiterpenes, a third class of isoprenoids, Finally, C-6 oxygenated compounds are emitted from leaves
are emitted in small quantities from non-stressed vegetation, except subject to various stresses such as wounding, e.g. as a consequence of
from flowers. The remainder is emitted as oxygenated volatile cutting hay, insect feeding, ozone stress and heat stress (Hatanaka,
compounds, including alcohols, aldehydes and ketones, particularly 1993). Aldehydes, (Z)-3-hexenal, (E)-3-hexenal and (E)-2-hexenal
during periods of plant development or in response to environ- with characteristic green leaf (‘cut grass’) odour are formed first,
mental stress (Heiden et al., 2003; Seco et al., 2007). and then transformed into corresponding alcohols by alcohol
dehydrogenases. Esters such as hexenolacetates can also be formed
6.1.1. Volatile isoprenoids and emitted during these processes. Proton-transfer reaction mass-
Volatile isoprenoids have been extensively studied compounds spectrometry has provided detailed insight into the associated
because of their relevant functions in plant vs. environment inter- time-sequence of events (Beauchamp et al., 2005).
actions and their role in the atmosphere. Isoprene and mono-
terpenes are formed in plastids via methylerythritol phosphate
(MEP) pathway (Lichtenthaler, 1999). 6.2. Environmental controls on BVOC emissions
While isoprenoid emissions mainly rely on newly synthesized
photosynthetic metabolites in chloroplasts, extra-chloroplastic 6.2.1. Physiological and physico-chemical controls of emissions
sources can feed carbon to sustain isoprene or monoterpene Models of BVOC emissions consider that emissions are controlled
biosynthesis, including xylem-transported sugars and chloroplastic either by physiological factors (‘‘isoprene’’ algorithm) or by physico-
starch (Karl et al., 2002a; Kreuzwieser et al., 2002) as well as chemical factors (‘‘monoterpene’’ algorithm) (Guenther et al.,
refixation of respired CO2 (Loreto et al., 2004). These additional 1993). By considering only physiological factors, the synthesized
carbon sources of isoprenoid biosynthesis can become significant compounds are immediately released from the foliage. In contrast,
especially when photosynthesis is constrained by environmental by considering only physico-chemical factors, the compounds are
stresses. Under extreme stress conditions, such as drought stress, emitted from specialized storage structures such as resin ducts
the leaf carbon budget can become negative, as leaves release more present in conifers, glandular trichomes in species from Lamiaceae
carbon in the form of isoprenoids and respiratory CO2 than they (peppermint), and oil glands in Rutaceae (lemon, orange) and
gain through photosynthesis (e.g. Brilli et al., 2007; Schnitzler et al., Myrtaceae (eucalypts).
2004). Physiological controls operate at the level of compound
Monoterpenes and sesquiterpenes are active compounds in plant synthesis. Light and temperature, the key environmental drivers,
interactions with other organisms. Monoterpenes may either attract affect the rate of intermediate production; temperature also affects
or deter herbivores or carnivores, and attract pollinators (Gershenzon the activity of flux controlling enzymes (Fig. 6.1, Niinemets et al.,
and Dudareva, 2007). Indirect evidence suggests that isoprene 1999) Over the longer term, synthesis and turnover of rate-limiting
and monoterpenes as lipid-soluble molecules may protect plant enzymes can control the flux rate (Monson et al., 1994; Lehning
membranes from thermal and oxidative stress (reviewed by Sharkey et al., 2001; Fischbach et al., 2002). In species with large storage
and Yeh, 2001). This favourable action can also assist in adapting pools, the emissions can be independent of the rate of compound
isoprenoid-emitters to increasing oxidative pressures (Lerdau, 2007). synthesis, being controlled by temperature effects on the
5226 D. Fowler et al. / Atmospheric Environment 43 (2009) 5193–5267

Fig. 6.1. The emission of volatile organic compounds from plants is limited by both physiological and physico-chemical characteristics. Physiological factors control the rate of
synthesis of VOC and operate at the level of intermediate production and the activity of flux controlling enzymes such as isoprene synthase or terpene synthases for volatile
isoprenoids. Physico-chemical factors, in particular, low compound volatility and rate of diffusion, can limit the release of synthesized VOC from the leaves and determine the degree
to which the synthesized compound can be stored in the leaves.

evaporation and diffusion of compounds from the storage pools concentration and atmospheric reactivity in morning hours. Current
(Fig. 6.1, e.g. Tingey et al., 1991). steady-state models predict zero nightime monoterpenes emissions
Often there is no clear separation between physiological and for species lacking specialized storage structures.
physico-chemical controls. Although it is generally belived that
only evaporation from pools controls the emissions in species with 6.2.3. Uptake and release of volatile compounds by vegetation
specialized storage pools, there is increasing evidence that the A relevant implication of the non-specific storage is the uptake of
emissions can be partly controlled by physiological factors also in volatile compounds from ambient air when air concentrations are
classical ‘‘storage’’ species. On the other hand, physico-chemical higher than those in equilibrium with plant liquid and lipid phases.
controls on emissions often interact with physiological controls in The compounds taken up during the periods with high atmospheric
species without specialized storage pools for BVOC (Fig. 6.1). BVOC concentrations may be released back into the ambient air
when air concentrations are small if they are not metabolized or
6.2.2. Physico-chemical controls of emission in species lacking translocated to the roots. The uptake of water-soluble compounds
specific storage structures is expected to scale with leaf water content, while the uptake of
Every BVOC species can be non-specifically stored in leaf liquid lipid-soluble compounds with leaf lipid content (Noe et al., 2008a).
and lipid phases, with the non-specific storage capacity depending on Thus, even species considered ‘‘non-emitting’’ can emit several
the compound physico-chemical traits such as the gas/liquid aqueous BVOCs at trace level from the non-specific storage built up from
phase partition coefficient (Henry’s law constant, H) and lipid/liquid ambient sources.
phase partition coefficient (octanol/water partition coefficient, KOW).
For instance, compounds that are highly water-soluble such as
methanol and ethanol can accumulate in leaf aqueous phase, espe-
cially if gas-phase diffusion out of the leaves is hindered due to limited
opening of stomatal pores. Although the rate of compound synthesis
may respond very quickly to environmental perturbations, a build up
of water-soluble compounds reduces the sensitivity of the emission
responses to variation in environmental factors.
Analyses of the emissions of strongly lipid-soluble BVOC species
such as non-oxygenated monoterpenes indicates that the non-specific
storage of these compounds, mainly in leaf lipid phase, significantly
alters the emission kinetics in species lacking specialized mono-
terpene storage pools. In these species, to reach steady-state rates of
monoterpene emission can take from minutes to hours depending
on monoterpene physico-chemical characteristics (Niinemets and
Reichstein, 2002; Noe et al., 2006).
The delayed emission responses due to non-specific storage can
also result in modified sensitivity of the emissions to environ-
mental factors. For instance, a sigmoidal light-response curve can
result if non-specific storage pools have not yet reached a steady-
state with each light level. Under such experimental conditions, the
emission rate recorded is lower than the monoterpene synthesis
rate. Given the long time periods, often on the order of 1 h required
to reach the steady-state, non-steady-state conditions are common
in monoterpene measurements.
In addition to alteration of the immediate environmental controls,
Fig. 6.2. Monoterpene emissions from Quercus ilex dominated forest in Castelporziano,
non-specific storage can modify the emissions at daily and weekly Italy for six days in June simulated using the standard Guenther et al. (1993) model and
timescales. Another important implication of the non-specific a model considering non-specific monoterpene storage in leaf liquid and lipid pools
storage is significant nocturnal emissions, which was observed e.g. in (modified from Niinemets and Reichstein, 2002; Niinemets, 2008). The standard emis-
the case of monoterpenes (Loreto et al., 2000; Niinemets et al., sion model predicts that the emission rate responds immediately to changes in light and
temperature, but non-specific storage of lipid-soluble non-oxygenated and water-soluble
2002b). Simulation analyses indicate substantial nocturnal emissions oxygenated monoterpenes results in time-lags between terpene synthesis and emission.
at the ecosystem scale (Fig. 6.2). Significant night-time BVOC emis- As the result of these time-lags, the emissions are predicted to continue also at night,
sions from non-specific storage can strongly influence OH-radical although synthesis has ceased.
D. Fowler et al. / Atmospheric Environment 43 (2009) 5193–5267 5227

6.2.4. CO2-Dependence of emissions Although the available emission information has been recently
The classical Guenther (1993) algorithm considers strong light and collated, data for many key emitting species are still lacking. Recent
temperature dependencies of isoprene and ‘‘non-stored’’ mono- observations have indicated that several important species such as
terpene emissions. CO2 was not considered as an important factor in cork oak (Quercus suber) (Staudt et al., 2004; Pio et al., 2005) and
controlling the emission of volatile isoprenoids. More recent labora- European beech (Fagus sylvatica) (Moukhtar et al., 2005; Holzke
tory and field studies have established that volatile isoprenoids are et al., 2006) previously considered non-emitters are moderate to
sensitive to ambient CO2 and that the emissions decrease in plants strong emitters of monoterpenes.
grown at CO2 concentrations higher than ambient (Loreto et al., 2001; The emission potentials of many ornamental, alien species used
Rosenstiel et al., 2003). This effect may be observed at CO2 concen- in urban landscaping and in gardens are missing (Owen et al., 2003;
trations that are likely to be reached in the future (double or less than Noe et al., 2008b). Understanding the emissions of ornamental
double the current concentrations) and should be therefore addressed species is also important in light of potential changes in vegetation in
in future modelling efforts. urban landscapes driven by global change and the growth of urban
There is now sufficient information to believe that the negative areas. Global warming and associated increases of evergreen emit-
effect of CO2 on isoprene emission is ubiquitous and not species- ting exotic species in northern urban landscapes of northern hemi-
specific. Arneth et al. (in press) included an empirical CO2-depen- sphere can importantly enhance the winter emissions (Niinemets
dence into the Niinemets et al. (1999) model, and predicted that and Peñuelas, 2008).
CO2 reduction of isoprene emission could partially compensate for In addition, stress- and time-dependent modifications of emission
the emission increases with rising temperature. However, elevated potentials are only partly understood, but such adaptive responses can
CO2 will enhance photosynthesis and growth of plants, in particular vastly affect ecosystem fluxes. Apart from gradual changes in BVOC
vegetation leaf area, which may in turn increase the emission of emission capacity in response to day-to-day and seasonal differences
isoprene by vegetation. The net effect of these interactions await in weather conditions (Guenther,1999; Sharkey et al.,1999), emissions
experimental confirmation. triggered by biotic stresses such as herbivory or pathogen attack or
Monoterpene emissions are also likely to be influenced by CO2, by abiotic stress factor such as elevated ozone concentrations
but there is less experimental evidence than for isoprene (e.g. (e.g. Beauchamp et al., 2005), can potentially greatly influence whole
Loreto et al., 2001). More studies are clearly needed to assess ecosystem fluxes.
whether the different physico-chemical controls (see above) and
the presence of small internal pools buffer the effect of CO2 and 6.4. BVOC fluxes over Europe, by compound and in relation
make monoterpenes less sensitive to CO2 control. to the needs of photochemical oxidant models

6.2.5. Induced emissions 6.4.1. Flux measurement techniques


In addition to the constitutive emissions, recent work demon- The development of disjunct eddy covariance (DEC) methods
strates that synthesis of volatile isoprenoids is induced in many (Rinne et al., 2001; ) alongside the development of proton-transfer
species in response to biotic (e.g. attack of herbivores and pathogens) reaction-mass spectrometry (PTR-MS, Lindinger et al., 1998) have
and abiotic (e.g. ozone stress, heat stress) stresses (Beauchamp et al., enabled direct flux measurements of multiple VOC species simul-
2005; Blande et al., 2005; Loreto et al., 2006). taneously. The PTR-MS has also been used in a more traditional
Previous work has shown that the emissions of volatile organic continuously sampling eddy covariance technique, in this mode the
compounds are induced in response to stress in practically all plant flux of just one VOC species can be measured at a time (Karl et al.,
species, also in those not emitting volatile isoprenoids under optimal 2001b, 2001c). The DEC methods have been utilized in different
growth conditions (e.g. tobacco and sunflower, Heiden et al., 1999; European ecosystems (Rinne et al., 2007; Davison et al., 2008).
Beauchamp et al., 2005). In constitutively emitting species, the Inter-comparison experiments to validate these new techniques
volatile isoprenoid blend differs between induced and constitutive have been conducted partly under ACCENT-BIAFLUX (Ammann
emissions. For instance, European aspen (Populus tremula) emits et al., 2006; Neftel et al., 2007; Rinne et al., 2008). For isoprene
isoprene as the main product in non-stressed conditions, but there also exists a fast isoprene sensor (FIS) based on chem-
induced emissions are dominated by monoterpenes limonene and iluminescence enabling the application of eddy covariance
a-pinene and sesquiterpenes (Blande et al., 2005). In non-stressed measurements (Guenther and Hills, 1998).
conditions, the emissions of the monoterpene-emitting species High reactivity of several plant BVOCs imposes significant diffi-
Pinus pinea are dominated by limonene, but the emissions in culties in determining whole canopy terpene emission fluxes by
conditions of high temperature and low water availability are micrometeorological techniques. In particular, some monoterpenes
dominated by linalool and ocimene; these emissions significantly and most sesquiterpenes have atmospheric lifetimes on the order of
exceed the emission in non-stressed conditions (Staudt et al., 2000). minutes. High reactivity of these compounds can imply that before
Current emission models do not consider ‘‘non-emitting’’ reaching the BVOC detector, a significant fraction of the emitted
species, which, in addition to emissions from non-specific storage, compounds has already reacted in the atmosphere, resulting in
may have significant induced emissions. Furthermore, modification underestimated emission fluxes (Rinne et al., 2007). To determine
of gene expression profiles in response to stress and upon adap- the emissions of reactive terpenes, atmospheric chemistry models
tation to stress may in many cases explain the modified emission have been inverted (Bonn et al., 2007). However, the lack of reactiion
compositions. Accordingly, understanding induction mechanisms rate coefficients for many terpenes and the dependencies of these on
and consideration of induced emissions is crucial in explaining and humidity and temperature seriously hamper the overall assessment
predicting emission profiles. of the rates of emission and contribution to atmospheric reactivity.

6.3. Contemporary difficulties in scaling BVOC emissions 6.4.2. Isoprene


from leaf to ecosystem Isoprene is the most studied biogenic VOC. European isoprene
emissions have been studied at the ecosystem scale in a range of
Parameterisation of models at ecosystem scales is bounded by landscapes. In some European ecosystems considerable isoprene
a series of uncertainties. A key uncertainty is currently the lack of emission fluxes have been measured (Ciccioli et al., 1997; Davison
reliable information of emission potentials (Arneth et al., in press). et al., 2008; Simon et al., 2005b). In contrast European boreal
5228 D. Fowler et al. / Atmospheric Environment 43 (2009) 5193–5267

coniferous forests are generally very low emitters (Rinne et al., 1999). an orange orchard revealed substantial emission of b-caryophyllene,
In measurements conducted above homogeneous conifer forest, the while the flux measurements conducted by REA method show the
isoprene emissions from isoprene emitters concentrated in e.g. fluxes to be close to zero indicating rapid within and below canopy
ditches, lakeshores and roadsides, will not be observed. Landscape chemical degradation (Ciccioli et al., 1999).
scale emissions over southern Finland, estimated by a boundary
layer budget method, show much smaller isoprene than mono- 6.4.5. Methanol
terpene emissions (Spirig et al., 2004). In general, measuring forest The development of the DEC-PTR-MS has recently enabled
sites, characterized by mixed composition, must account for the ecosystem scale flux measurements of methanol leading to
plant species composition within the flux footprint. This has been considerable progress in our knowledge of ecosystem scale emission
shown by Spirig et al. (2005), who observed the variation in of this compound. Methanol seems to be emitted from all ecosys-
normalized ecosystem scale isoprene emission from a central tems and also from drying and decaying plant material. Methanol is
European mixed broadleaf forest to be dependent on the abundance typically the second or third most abundantly emitted biogenic VOC
of Quercus robur, which is a high isoprene emitter, within the flux after isoprene and monoterpenes in most ecosystems.
footprint area. The measured ecosystem scale methanol emissions do not
Flux measurements of isoprene often correlate well with leaf- generally fit the emission algorithms, used for monoterpene fluxes. A
level measurements, reflecting the relatively long atmospheric recent empirical temperature dependent formulation has been pre-
lifetime of isoprene (w1 h) compared with other shorter lived sented for methanol by Harley et al. (2007). From a mechanistic
BVOCs. However, when the ecosystems are not homogenous, understanding of the physico-chemical control (see above), meth-
correspondence between emissions at plant and ecosystem levels is anol emission should be regulated by stomata, in contrast to isoprene
not straightforward. For example, the study conducted at Siikaneva and monoterpene emission. In field flux experiments, this kind of
fen ecosystem using soil chambers (Hellén et al., 2006) and REA behaviour has not been generally observed. Brunner et al. (2007)
technique (Haapanala et al., 2006) shows a considerable discrepancy have observed the methanol emissions from agricultural grassland
between the isoprene emissions measured by these two techniques. ecosystem to be large in the morning relative to emissions in the
The fluxes measured by Haapanala et al. (2006) under low CTCL evening. A similar observation was found in the recent ACCENT–
values were substantially smaller than the model values, which may VOCBAS campaign on the Mediterranean macchia of Castelporziano.
imply deeper penetration of PAR into the moss carpet at highlight
conditions; a simple light penetration model slightly improved the 6.4.6. Acetone and acetaldehyde
correlation (Haapanala et al., 2006). Acetone and acetaldehyde in the atmosphere are the result of both
primary emissions from biogenic and anthropogenic sources and
6.4.3. Monoterpenes secondary formation from other gaseous precursors. Fluxes of these
European ecosystems are substantial sources of atmospheric carbonyls have been observed to be emitted from coniferous forests in
monoterpenes and many flux measurement experiments have Europe (Rinne et al., 2007) as well as in the US (Schade and Goldstein,
concentrated on these compounds. The diurnal variations of the 2001). Also the Mediterranean macchia ecosystem is observed to emit
monoterpene fluxes above boreal coniferous forests appear to be these compounds (Davison et al., 2008). No emission of these
relatively well reproduced by the temperature dependent Tingey– compounds from broadleaf deciduous forests has been reported.
Guenther emission algorithm (Guenther et al., 1993). Monoterpene Anaerobic conditions in root system have been observed to enhance
emissions from Mediterranean ecosystems are better described by the emission of acetaldehyde from plant foliage (Kreuzwieser et al.,
the light and temperature dependent isoprene emission algorithm 1999) as also mechanistically explained above. However, no ecosystem
(Seufert et al., 1997; Schween et al., 1997), with important discrep- scale flux measurements of this compound, in conditions where the
ancies likely reflecting non-specific storage and stress-dependent root systems was anaerobic, have been reported.
changes in emissions (Niinemets et al., 2002a, 2002b, see above).
At coniferous forest sites a diurnal concentration cycle with 6.4.7. Other compounds
highest concentrations at night are typical (e.g. Hakola et al., 2000; The emissions of many compounds, other than isoprene,
Steinbrecher et al., 2000; Rinne et al., 2000, 2005). This is due to the monoterpenes, methanol, acetone and acetaldehyde, have been too
emission continuing during night-time, although at a lower rate small to be measured by micrometeorological flux measurement
than during the day, and to considerably reduced turbulent mixing techniques. However, there are a few other compounds which
at night. On the contrary, in the Mediterranean region, as well as in have been observed to be emitted by micrometeorological flux
Amazonian tropics, and in European mixed broadleaf forest, where measurement techniques in certain ecosystems. Some western US
the night-time monoterpene emission is practically zero due to the pine forests have been shown to emit 2-Methyl-3-buten-2-ol (MBO)
light dependence of the monoterpene emission, daytime maxima in considerable amounts (e.g. Baker et al., 1999; Schade et al., 2000).
are typical (Zimmerman et al., 1988; Schween et al., 1997; Rinne From European Pine species, no significant emissions of MBO have
et al., 2002; Spirig et al., 2005). been observed (e.g. Hakola et al., 2006). It is noteworthy that drying
The major monoterpenes emitted by forest ecosystems in Europe hay was observed to emit (Z)-3-hexenal and (Z)-3-hezenol and
are a- and b-pinene and D3-carene. For example, monoterpene hexenyl acetates (Davidson et al., 2008a). These compounds appear
emissions from Norway spruce forest consist of a- and b-pinene to be reliable markers of membrane damage and lipoxygenation, as
(Christensen et al., 2000). However, in some Mediterranean ecosys- previously indicated (Loreto et al., 2006).
tems limonene can form a significant part of the monoterpene
emission (Schween et al., 1997) and, in the case of orange orchards, 6.5. The EU large field campaigns in the Mediterranean area:
be the dominant monoterpene emitted (Christensen et al., 2000; from BEMA to ACCENT
Darmais et al., 2000).
Pioneering campaigns around the world (e.g. the SOS 1999
6.4.4. Sesquiterpenes campaign in North America, the LBA/CLAIRE 1999 campaign over the
Due to the enhanced chemical reactivity of sesquiterpenes, atmo- Amazon, the SAFARI 2000 campaign in Southern Africa) highlighted
spheric concentrations of these compounds are very small, making the predominance of isoprene as the most commonly emitted VOC by
flux measurements extremely difficult. Enclosure measurements at vegetation, over different environments, and by different ecosystems,
D. Fowler et al. / Atmospheric Environment 43 (2009) 5193–5267 5229

spanning tropical to boreal forests. In Europe, however, a different presence of small sandy dunes, and by a gradient of soil humidity
picture emerged. The BIPHOREP 1996–1997 campaign in boreal due to the presence of freshwater sources.
Europe revealed the dominance of monoterpenes as the most abun-
dantly emitted VOC from European boreal ecosystems. In otherwise 6.6. Remote sensing of BVOC
clean northerly air masses, at least the growth, if not the formation, of
organic aerosol, seems to be strongly influenced by the presence of Over the last decade space-based instrumentation, capable of
biogenic VOCs. The most likely candidates contributing to this aerosol probing the lower troposphere, has reached the level of accuracy
growth are monoterpene oxidation products (Lee et al., 2006). The necessary to quantify surface sources and sinks of trace gases from
BEMA (Biogenic Emission in the Mediterranean Area) 1996 campaign observed variations in trace gas concentrations (Palmer, 2008). The
highlighted the remarkable peculiarity of Mediterranean coastal only non-methane BVOC measured in the lower troposphere from
vegetation that is almost uniquely characterized by monoterpene- space is formaldehyde (HCHO), a high yield product of VOC
emitting species (Loreto et al., 1998). The reason for this distinctive oxidation that is measured from clear-sky backscattered solar
trait of Mediterranean vegetation remains unknown. radiation at ultraviolet wavelengths (Chance et al., 2000). The main
A new campaign during spring 2007 in the Mediterranean area, sinks of HCHO are oxidation by OH and photolysis leading to
as part of the ACCENT and European Science Foundation programme a tropospheric lifetime of several hours. Fig. 5.3 shows the global
VOCBAS programmes took place. The campaign was deliberately distribution of HCHO columns observed by the Ozone Monitoring
held on the same site of the BEMA campaign, the large peri-urban Instrument during August 2006.
natural preserved area of Castelporziano, in the conurbation of Oxidation of methane (CH4) by OH, the largest global source
Roma. This site has two main characteristics that make it an excellent of HCHO, provides a uniform HCHO background of w100 pptv,
case of study for biosphere–atmosphere interactions. First, the reflecting the w8-year lifetime of CH4. The limit of detection of
6000 ha wide preserved area of Castelporziano is a hot spot for HCHO from current space-borne instrument is approximately
biodiversity in the Mediterranean, with more than 1000 plant 4  1015 molec cm2 (Chance et al., 2000), which is close to the
species represented in the flora of the area (Davison et al., 2009). The source of HCHO from CH4 oxidation. Over the continental boundary
main ecosystems going towards the sea are characterized by oak layer, oxidation of anthropogenic and biogenic VOCs provide an
(Quercus ilex, Q. suber, Quercus cerris) and pine (P. pinea) forests, often additional source of HCHO that can reach on a local scale up to
associated with a rich understory vegetation. The part of the Estate several ppbv, equating to columns over an order of magnitude
facing the Tyrrhenian sea is characterized by sand dunes and determined by CH4 (Fig. 6.3). Observed variations in HCHO, deter-
a humid retrodunal area, with a large and extremely well preserved mined by the oxidation of VOCs, therefore provide constraints on
area covered by Mediterranean ‘‘macchia’’ vegetation, prevalently emissions of the parent VOCs. Horizontal transport smears the
shrubs and small evergreen trees, such as Juniperus communis, Q. ilex, local relationship between VOC emissions and HCHO columns,
Phillyrea latifolia, Arbutus unedo, Rosmarinus officinale, Erica arborea, the extent of which is determined by wind speed and the time-
and Cistus incanus. Second, the preserved area is only distant 25 km dependent yield of HCHO from the VOC oxidation (Palmer et al.,
from the centre of Roma in the S-E direction. It is exposed to 2003). Over a number of global regions, variations in HCHO
a constant wind circulation that favours transport of air masses from columns are determined by isoprene (Palmer et al., 2006), due to
the city center during night-time, and from the sea during daytime. its rapid production and high molar yield of HCHO (Palmer et al.,
This periodically exposes vegetation to urban pollutants and may 2006). Other reactive biogenic VOCs, such as monoterpenes, also
trigger formation of secondary pollutants that are contributed by have short atmospheric lifetimes but they quickly produce acetone
BVOC precursors (Chameides et al., 1988; Di Carlo et al., 2004). with a high yield that has an atmospheric lifetime of weeks and
The campaign (i) provided fluxes of BVOC from Mediterranean consequently slows down the production of HCHO (Palmer et al.,
vegetation, in a season during which plants are in optimal physio- 2006). Long-lived VOCs such as CH4 and CH3OH, while being the
logical conditions prior to drought and heat stress conditions largest sources of HCHO, only contribute to the slowly varying
experienced later in the year, but during which BVOC emisions are background of HCHO.
thought to be constrained by leaf development limitations; Early work showed that the magnitude and distribution of
(ii) investigated, by coupling concentration and flux measurements, GOME-derived isoprene emissions based on HCHO measurements
the in situ extent of BVOC reactivity, and in particular whether in were more consistent with in situ measurements than either the
situ BVOC oxidation could drive formation of secondary organic GEIA or BEIS2 isoprene inventories, based on Guenther et al. (1995)
compounds in the atmosphere; (iii) identified whether BVOCs can (Palmer et al., 2003). Monthly mean distributions of HCHO, and
act as precursors of photochemical smog; (iv) provided a second inferred isoprene emissions, during summertime are dominated by
assessment, ten years after the campaign organized by the high values over the southeastern states of the USA (Chance et al.,
EC-BEMA project (Seufert et al., 1997), of BVOC emission in an 2000) due to a large density of isoprene-emitting oak trees over the
area largely affected by anthropogenic and climatic changes, thus Ozarks Plateau (Wiedinmyer et al., 2005). Examination of GOME
creating the foundation for a historical series of measurements orbital data revealed large variations in HCHO, explained by
which may be especially important in view of current and future changes in surface temperature, which led to inferred monthly
climate change factors, and, in particular, of the simultaneous and mean isoprene emissions that were significantly lower than those
strong increase of temperature, drought and pollutants in the predicted by bottom–up models (Palmer et al., 2003). Later work
Mediterranean area; and (v) fostered interdisciplinary collaboration showed that the observed seasonal and year-to-year variability was
between the communities of biologists, atmospheric chemists and consistent with the MEGAN model (Guenther et al., 2006), but
physicists, further catalyzing research on the important roles of GOME-derived isoprene emissions were 25% higher (lower) at the
BVOCs in the environment. beginning (end) of the growing season (Abbot et al., 2003; Palmer
The campaign was run in the macchia strip placed between the et al., 2006). Both MEGAN and GOME show a maximum over the
dunes and the main forested land inside the Castelporziano Estate. Southeast US but disagree in the precise location, with implications
The oak and pine forested area was extremely well characterized for modelling surface ozone (Fiore et al., 2005).
during the BEMA campaign (1997) but the macchia vegetation Isoprene emissions from tropical ecosystems have been esti-
received only limited attention (Owen et al., 1997). The macchia strip mated to contribute 75% of the global isoprene budget, but are not
is characterized by a modest roughness of the terrain due to the well quantified. Over tropical South America, the widespread
5230 D. Fowler et al. / Atmospheric Environment 43 (2009) 5193–5267

Fig. 6.3. Formaldehyde columns (1016 molec cm2) retrieved from the NASA Aura Ozone Monitoring Instrument (OMI) for August 2006, courtesy of Thomas Kurosu, Harvard
Smithsonian Center for Astrophysics.

extent of biomass burning in the dry season means that without consequently lead a greater number of useful data points; 2) spatial
high-resolution data the only practical approach is to use data disaggregation of different HCHO sources, e.g. biomass burning and
over west Amazonia, which is largely unaffected by fires (Barkley biogenic VOC emissions over tropical regions, which can lead a better
et al., 2009). There is a strong seasonal cycle of GOME HCHO description of land-surface processes; and 3) their usefulness in
columns over this region reproduced each year during 1996–2001, planning and executing measurement campaigns. For example,
characterized by large values in the wet and dry seasons, sepa- analysis of GOME-2 and OMI HCHO data, which have local overpass
rated by low values in the wet-to-dry transition period (May– times of 09:30 and 13:30, will allow us to develop a crude under-
July); this is consistent with in situ isoprene concentration standing of the diurnal cycle of BVOC emissions on continental scales.
measurements (Palmer et al., 2007). This large-scale reduction in Knowledge of the complex organic chemistry associated
isoprene emissions suggests a major temporary shift in underlying with BVOC oxidation will only improve with a concerted effort to
meteorology or phenology, but its origin remains unclear. This increase the number of laboratory and field measurements. With
study found that MEGAN and GOME were in better agreement in the rapid increase in available remotely sensed datasets that could
the dry season, when GOME isoprene emissions could not be be brought to bear on estimating BVOC emissions (e.g. HCHO, leaf
explained by changes in surface temperature. GOME isoprene phenology, land cover) there should be scope to develop new
emissions in the wet season could be not explained by changes in functional descriptions of isoprene emission that are independent
surface temperature, precipitation or soil moisture, suggesting of the assumptions made in traditional bottom–up models derived
either an unexplained process that determines isoprene during from in situ measurements.
this season or noisy data (Barkley et al., 2009). These substantial
open questions will be readdressed with data from newer sensors 7. Deposition and resuspension of aerosols onto
and should be the subject of extensive year-long ground-based and from terrestrial surfaces
measurement campaigns.
There are a number of uncertainties associated with the HCHO 7.1. Introduction
measurement and the approach used to infer surface emissions of
isoprene from these measurements (Palmer et al., 2003; Millet et al., Aerosols present a complex multi-variate and multi-scale
2006), which lead to uncertainties that total more than 100% of problem to environmental research. They have strong impacts on
estimated emissions (Palmer et al., 2006). Work related to the anal- climate indeed they represent the largest uncertainties in our ability
ysis of GOME data has suggested that HCHO production predicted by to predict climate change (IPCC, 2007), human health and provide
chemical mechanisms typically used by large-scale chemistry- a means for pollutant deposition to sensitive ecosystems (trans-
transport models are in error by more than 25% (Palmer et al., 2006). porting e.g. nitrogen, sulphur and toxic metals). A vital component of
Current studies that determine VOC emissions using HCHO the global aerosol cycle is deposition of many critical secondary
column data are already focusing on data from newer space-borne process-derived condensed phase compounds as well as primary
sensors (e.g. Aura OMI and MetOp GOME-2) that have better spatial generated aerosols, both in the ultrafine and coarse modes. The
(100 s km2) and daily temporal resolution, enabling more detailed majority arise from the consequences of anthropogenic activities
testing of current bottom–up inventories. The advances in resolution, associated with global industrialisation and urbanization where
in particular, improve: 1) the probability of cloud-free scenes and large emissions of reactive nitrogen species lead to an increase in
D. Fowler et al. / Atmospheric Environment 43 (2009) 5193–5267 5231

nitrogen aerosol formation which is eventually removed by wet or covariance (EC) measurements of size-resolved aerosol number
dry deposition. Sulphate aerosol remains an important issue in fluxes (e.g. Duan et al., 1988; Gallagher et al., 1997; Nemitz et al.,
the Eastern US and China. Over the last 150 years the atmospheric 2002a; Sievering, 1983; Vong et al., 2004). The measurements
particulate loading has changed from coal and other solid fuel usually cover the size spectrum between 0.1 and 0.5 mm, but even in
burning to modern combustion processes that liberate a greater this size range the particle statistics of these instruments have often
preponderance of submicron, ultrafine particles. This has initiated been marginal in deriving statistically significant fluxes. Several
a paradigm shift in some quarters of the relative importance of studies have attempted to extend measurements to smaller and also
coarse mode as opposed to fine mode particulates and their to larger sizes. Particles significantly <0.06 mm cannot be sized with
behaviour for mainly health related reasons. This shift will likely current optical techniques. Instead, recent studies have attempted to
move into a third phase where emissions from nano-particle measure size-segregated fluxes of smaller particles, either using the
technologies are starting to play an increasing role. In addition, it is relaxed eddy accumulation (REA) technique combined with size
becoming obvious that particulate matter provides an area of policy selection using a differential mobility analyser or interpreting total
conflict: efforts to curb PM concentrations to protect human health particle number fluxes during periods where a certain size-range
are likely to reduce global dimming and thus further accelerate dominated the flux (Grönholm et al., 2007; Pryor, 2006). Due to
climate change, although there are also components, such as soot, limited counting statistics, fluxes had to be averaged over many days
that have a negative impact on both human health and the climate to obtain robust statistics and these measurement methods are
system. The primary aims of research into the biosphere/atmosphere therefore not yet suitable to study short-term processes (Fig. 7.1).
exchange of particles are: Variability between measurements is likely to be linked to differ-
ences in turbulence, canopy morphology (including leaf area index)
(a) to improve our estimates of primary aerosol emissions from and surface roughness length (cf. Section 7.4.1).
diffuse sources and their parametrisations, Size-segregated EC fluxes of larger particles are only possible
(b) to measure directly the contribution of particle deposition to when these particles are abundant and occur as the result of specific
the deposition of compounds that are detrimental to ecosys- mechanisms, e.g. in dust storms, biomass burning or industrial
tems or may accumulate in water, soils or crops (e.g. N and S processes. High volume, closed path aerodynamic Mie scattering
compounds, heavy metals and nano particles), time of flight optical particle counters, for sizes 0.5 < Dp < 20 mm,
(c) to derive parametrisations of the deposition velocity (Vd) of and open path forward scattering optical particle counters have been
particles, for inclusion into CTMs e.g. aimed at predicting used to measure deposition rates of super-micron particles and fog
deposition, human health impacts and climate impacts, droplets (e.g. Beswick et al., 1991; Burkhard et al., 2002; Klemm and
(d) to study aerosol formation and dynamics in the atmosphere. Wrzesinski, 2007; Kowalski and Vong, 1999). As a result of current
instrument limitations, measurement evidence is sparse on depo-
If the models predict the size distribution of the aerosol explic- sition rates in the important accumulation mode (0.3–2 mm),
itly, Vd needs to be parametrised as a function of particle diameter which contains much of the mass of sulphur, nitrogen and secondary
(Dp). By contrast, where the models only deal with the bulk species, organics. Recently developed mass-based flux measurement
Vd may be derived from measurements of the individual compounds approaches based on time-of-flight aerosol mass spectrometry (see
(e.g. Ruijgrok et al., 1997) or, more commonly, from a weighting of below) may go some way towards filling this gap in the future.
Vd(Dp) with a typical size distribution of the aerosol component. It is sometimes easier to measure super-micron emission fluxes.
For example, Fratini et al. (2007) reported measurements of desert
7.2. Review of new measurement approaches and instrumentation dust resuspension, and Nemitz et al. (2000c) presented urban
EC flux measurements in the range 0.8–10 mm, made with
Progress in the quantification and parameterisation of surface/
atmosphere exchange of particles and aerosol compounds is closely
related to developments in the measurement technology. In the 3 Beech forest (mean)
1970s and 80s measurements of aerosol dry deposition were made dep. only, CPC (Pryor, 2006)
Beech forest (median), dep. only,
in the wind tunnel or using surrogate surface collectors, such as 2 CPC (Pryor, 2006)
knife-edge collectors, inverted frisbee type dust deposition gauges Scots pine, CPC nucleation event
and moss bags. Although these techniques are still being used, they (Gaman et al., 2004)
Scots pine, DMA REA
have attracted serious criticism as their aerodynamic properties are
(Gronholm et al., 2007)
usually not representative for the surface for which deposition is to 10
9
be estimated. The lack of an understanding of the relationship 8
Vd [mm/s]

7
between deposition on these collectors and deposition to the
6
surrounding landscape reduce the value of the measurement to 5
trend detection. More recently, non-intrusive micrometeorological
4
flux measurement techniques have increasingly been extended
to measure surface/atmosphere exchange particle fluxes at the 3
field scale. Measurements of particle fluxes fall into two categories:
particle number fluxes (total or size-segregated) and chemically 2
resolved aerosol fluxes.

7.2.1. Flux measurements of particle numbers (size-resolved


or total), without information on chemical composition 1
7 8 9 2 3 4 5 6 7 8 9 2
These measurements are usually used to derive (size-dependent) 10 100
deposition velocities which can be used in atmospheric transport Dp [nm]
and deposition models. Since the 1980s, fast optical particle counters
(e.g. OPCs such as ASASP-X/555X, Particle Measurement Systems; Fig. 7.1. Summary of measured size-segregated particle deposition velocities to forest
FAST, Droplet Measurement Technologies) have been used for eddy for particles with diameters < 100 nm.
5232 D. Fowler et al. / Atmospheric Environment 43 (2009) 5193–5267

aerodynamic particle sizers (APS 3320, TSI Instruments), both made and organic aerosol to urban areas and forests (Nemitz et al., 2008;
in conditions where fluxes were large. Phillips et al., in preparation; Thomas, 2007; Thomas et al., 2009,
In general, better counting statistics can be achieved when 2007). The Q-AMS monitors only one single mass/charge ratio (m/z)
integrated particle number fluxes are measured over larger size at a time. However, the quadrupole MS can be switched very rapidly
ranges. For example, condensation particle counters (CPCs) are now so that quasi-continuous time series of concentrations can be
more commonly used to measure total particle number fluxes with established at typically 10 different m/z at 10 Hz, similar to the use of
a lower cut-off between 2.5 and 20 nm over ice, seawater, short and the PTRMS for VOC measurements. Since >100 different m/z
tall vegetation canopies, as well as urban areas (e.g. Buzorius et al., contribute to the organic mass spectrum, with the Q-AMS, the total
1998; Buzorius et al., 2001; Dorsey et al., 2002; Held et al., 2006; organic aerosol mass flux has to be estimated from fast response
Martensson et al., 2002; Nemitz et al., 2002a; Nilsson and Rannik, measurements at a few m/z. The arrival of the next generation AMS
2001). CPC derived flux measurements are dominated by particles based on a high-resolution time-of-flight mass spectrometers
in the range 10–100 nm, with smaller particles dominating during (HR-ToF-AMS) (De Carlo et al., 2006) provides the prospect of
nucleation events (Buzorius et al., 2001). Depending of the model monitoring all m/z continuously at 10 Hz. This should enable a fully
used, CPCs have a response time of around 1 s, which is sufficient to quantitative flux measurement of the organic fraction and provide
measure fluxes from taller towers, but requires flux corrections for data to apply statistical approaches, currently used to deconvolve the
shorter vegetation. An alternative EC flux measurement approach organic mass concentration in different organic aerosol classes (e.g.
that integrates overall particles was implemented by Fontan et al. Ulbrich et al., 2009), to the flux measurement.
(1997) based on particle counting through a combination of corona
charging and detection by electrometers. Only very few gradient
measurements of small particle number fluxes have been reported 7.3. Area sources of particles
in the literature (Hummelshoj, 1994) due to the large errors
associated with these measurements. Work has started to extend 7.3.1. Resuspension
EC approaches to the measurement of particle fluxes from moving Resuspension of particles has been studied extensively, both theo-
platforms such as aircraft (e.g. Buzorius et al., 2006) and ships retically, in the wind tunnel or through concentration measurements
(Norris et al., 2007). (Braaten and Paw, 1992; Harrison et al., 2001; Nicholson, 1988, 1993;
Nicholson et al.,1989). A full review of the understanding of this process
7.2.2. Flux measurements of individual aerosol chemical species is beyond the scope of this paper. Instead, we here focus on a recent
Measurements of chemically resolved aerosol fluxes can be used development involving the first application of micrometeorological
to quantify deposition inputs directly, to investigate effects of flux measurement techniques to the direct measurement of resus-
aerosol composition on exchange rates and to understand apparent pension fluxes, and their potential to derive new parametrisations of
emission fluxes. The number of studies that have applied micro- the process. Nemitz et al. (2000f) measured super-micron size-segre-
meteorological approaches to measure fluxes of aerosol compounds gated aerosol fluxes measured with an Aerodynamic Particle Sizer (APS
has been surprisingly limited. Up to the 1990s, the main option 3320, TSI Inc.) from a tower, some 65 m above the city of Edinburgh,
was gradient measurement with labour intensive manual sampling Scotland. The measurements, binned according to diameter and wind
techniques based on filter packs or denuder/filter combinations speed lead to a parameterisation of the form:
(Duyzer, 1994; Rattray and Sievering, 2001; Wyers and Duyzer,      
1997), where a key challenge is to achieve the precision required to dFm Dp dlog Dp ¼ am Dp U bðDp Þ
     
resolve the very small aerosol gradients, which are often <3%. am Dp ¼ exp 68:69þ73:39 1exp 0:730Dp ½mm
Manual sampling techniques have also been used in REA approaches    
b Dp ¼ 20:12exp 0:506Dp ½mm ð1Þ
to measure fluxes, e.g. of sulphate to a maize crop (Meyers et al.,
2006) and of ions and heavy metal above a city (Nemitz et al., where U is wind speed measured at a height of 65 m.
2000d). A family of automated real-time gradient monitors, based The measurements show that above U (65 m) ¼ 6 m s1 resus-
on gas and aerosol capture by continuously flushed wet rotating pension becomes an important particle source in the urban envi-
denuders and steam jet aerosol collectors, respectively, and online ronment, with a mode centred around a Dp of 2.8 mm, which
analysis by ion chromatography and/or flow injection analysis increases with decreasing wind speed. An attempt to include traffic
(for NHþ4 ) (Thomas et al., 2009) has been used in a number of studies activity in the parametrisation failed due to the overriding effect of
to measure deposition of water-soluble inorganic aerosol compo- wind speed. This suggests that, although vehicle induced resus-
nents (Nemitz et al., 2004b, 2000e). pension may be important to lift particles off the surface at street
Eddy covariance measurements of aerosol chemical species were level, high wind speeds are nevertheless required to flush these out
first presented for SO2 4 deposition to grassland, based on an analyser of the street canyons. It should be noted that the windy periods with
with thermal conversion to SO2 (Wesely et al., 1985), with no further largest coarse particle emissions did not result in the largest
studies until the advent of aerosol mass spectrometry offered the concentrations, due to increased dispersion during these periods.
prospect for fast measurement of a number of aerosol compounds. Similarly, Fratini et al. (2007) measured coarse aerosol fluxes
Held et al. (2003) theoretically investigated the suitability of aerosol during desert storms in the Alashan desert in Nothern China, using
mass spectrometers based on single-particle analysis by laser abla- an optical particle counter. The authors found that the dependence
tion and ionisation for disjunct eddy covariance flux measurements, of the resuspension flux (in mg cm2 s1) of PM1, PM2.5 and PM10 on
dealing with the limited counting statistics and quantification issues u* (in m s1) could be described by the power relationships of
related to this instrument (De Carlo et al., 2006). An alternative F1 ¼ 469u3.11
* , F2.5 ¼ 6220u*
3.34
and F10 ¼ 47,500u3.36
* , respectively. It
instrument, the Aerodyne Aerosol Mass Spectrometer (AMS), should be noted that these two studies address different resus-
which is based on thermal vapourisation coupled with electron pension processes, reflecting road dust resuspension with vehicle
impact ionisation, averages over a much larger aerosol population contribution, and natural saltation processes, respectively.
and is quantitative for submicron aerosol components that are non-
refractory, i.e. volatilise at the vapouriser temperature of w600  C. An 7.3.2. Urban emissions of aerosols
operational EC system using the quadrupole-based AMS (Q-AMS) In recent years, flux measurement techniques have been
2
(Nemitz et al., 2008) has been used to measure fluxes of NO 3 , SO4 extended to the urban environment to quantify emission fluxes of
D. Fowler et al. / Atmospheric Environment 43 (2009) 5193–5267 5233

trace gases such as CO2, N2O, CO and VOCs (Nemitz et al., 2002b; cities (Boulder, US; Mexico City; Gothenburg, Sweden; Manchester,
Velasco et al., 2005; Vesala et al., 2007), but also of particles. Total Edinburgh and London, UK) (Grivicke et al., 2007; Nemitz et al.,
number fluxes have been measured with condensation particle 2008; Phillips et al., 2007; Thomas, 2007). The AMS measures
counters over several cities (Dorsey et al., 2002; Martensson et al., total organic aerosol mass contained in particles with 60–800 nm
2006; Nemitz et al., 2000f). Martin et al. (in press) recently compared vacuum-aerodynamic diameter and volatilizes at typically 600  C.
the pattern of particle number fluxes measured at four different More information on the organic aerosol classes can be obtained
locations, three of which are shown in Fig. 7.2. Fluxes, typically from the organic mass spectrum, with statistical techniques
covering the diameter range 0.01 (or 0.003) to 2 mm, ranged from (e.g. Ulbrich et al., 2009), which can be used to separate the organic
5000 to 70,000 # cm2 s1 and showed a clear dependence on traffic mass flux into fluxes of (primary) hydrocarbon-like organic aerosol
activity confirming the role of traffic emissions as the major source (HOA) and (secondary) oxygenated organic aerosol (OOA), where
of particles in the urban area. They derived a parametrisation for the OOA can often be divided into a more (OOA-I) and a less (OOA-
the flux (FPred in cm2 s1) over each city based on friction velocity II) oxidized component. The measurements to date show clear
(u* in m s1), sensible heat flux (H in W m2) and traffic activity (TA diurnal fluxes of HOA reflecting the pattern of the surface sources
in veh s1) of the form (Fig. 7.3). However, the flux ratios of HOA/CO and HOA/CO2 vary
h i over the day, indicating that either (a) some of the HOA evaporates
FPred ¼ C EFfriction u* þEFheat H þEFtraffic TAF0 ; (2) at rates that vary over the day (e.g. Robinson et al., 2007) or (b) that
the fuel mix contributing to the emissions of HOA and CO varies
which is shown in Fig. 6.3 for comparison. Here C is a site-specific over the night. Measurements suggest that food cooking may
factor in the range 0.12–0.55. The emission factors are contribute to HOA emissions in the evening in London. Fluxes of
EFfriction ¼ 4500 cm3, EFheat ¼ 6.53  106 W1 s1 and EFtraffic lies in OOA-I appear to be mainly downwards consistent with its
the range 9000–12,600 veh1 cm2, depending on the location of production during long-term transport, while small upward fluxes
the traffic census site. The sink flux (F0) ranged from 13,000 to of OOA-II were measured, indicating that some OOA-II formation
57,000 cm2 s1. occurs below the measurement height of typically 30–200 m. The
In the Edinburgh study of Nemitz et al. (2000f) aerosol number urban flux measurements indicate that SO2 4 is deposited to most
fluxes were dominated by traffic activity, while the aerosol mass city centres. Apparently, with the introduction of ultra low sulphur
emission fluxes were dominated by the wind-driven resuspension fuels, there are no primary sources of this compound. Fluxes of NO 3
described in the previous section This may be different for less windy were more variable: in Gothenburg, Edinburgh, London and
locations as demonstrated by Schmidt and Klemm (2008), who Boulder, net emission was observed, but fluxes were dominated by
presented flux measurements of super-micron particles made with individual, often cool or foggy days, indicating urban NO 3 forma-
a novel disjunct eddy covariance system, based on an Electronic Low tion under these conditions. By contrast, above Manchester and
Pressure Impactor (ELPI, Dakati, Finland), indicating net coarse-mode Mexico City, the average flux was downwards.
deposition to the German town of Münster. During the measurement
periods, wind speed averaged 8.0 m s1 in Edinburgh and 4.4 m s1 in 7.4. Dry deposition of particles
Münster (Klemm O., personal communication). Donateo et al. (2006)
reported flux measurements of PM2.5 above an urban area made with 7.4.1. Dry deposition rates to vegetation
an optical detector calibrated against gravimetric PM2.5 measure- Dry deposition of atmospheric particles can account for a large
ments. These measurements indicated continuous net upward fluxes fraction, sometimes more than half, of the total deposition of many
and represent a combination of Aitken, accumulation mode particles important chemical compounds in the atmosphere, (e.g. nitrate;
and super-micron particles. Lovett, 1994), contributing significantly to global biogeochemical
Chemically resolved flux measurements using the AMS EC cycling. Understanding atmospheric deposition processes in relation
technique described above have now been made over a range of to the ever increasing sources of fine particles in particular is
therefore becoming a research priority. In general, both aerosol mass
and number, and, increasingly, surface area, must be determined
to assess the environmental impacts of anthropogenic activities.
Mapping between number and mass fluxes however requires either
a detailed knowledge of the aerosol mass size distribution at rela-
tively high temporal resolution or by direct measurement of the
aerosol deposition flux as a function of both size and chemical
composition. In this short summary we will critique the present level
of understanding from an observational perspective and identify the
current gaps in our knowledge. There have been a number of recent
reviews on the subject of atmospheric aerosol deposition which
have attempted to collate the sparse experimental results available
from the previous two and half decades. These consist of many
disparate and difficult to compare or even reconcile, methodologies.
The reviews, whilst achieving this difficult process in some
respects, have succeeded mainly in highlighting the general lack of
a systematic approach towards improved understanding of mecha-
nistic deposition processes and, despite best efforts, have simply
reinforced the view of continued large uncertainties. Reviews have
been very much focused on either a modelling (Petroff et al., 2007b;
Zhang and Vet, 2006) or a measurement perspective (Petroff et al.,
Fig. 7.2. Summary of diurnal patterns of measured particle number fluxes (dotted lines)
2007a; Pryor et al., 2007).
and their parametrisation (solid lines) for three UK cities: M – Manchester; L – London; Many regional pollutant deposition models are currently
E – Edinburgh; Win06 – winter 2006 etc. struggling to correctly incorporate physico-chemical properties of
5234 D. Fowler et al. / Atmospheric Environment 43 (2009) 5193–5267

400
40 Manchester
300 Manchester
20
200
0
100
-20
0
-40
-100 200
London 1200 London 150
HOA flux [ng m-2 s-1]

HOA flux [ng m-2 s-1]

NO3- flux [ng m-2 s-1]


100

NO3- flux [ng m-2 s-1]


800
50
400
0
200 0 -50
Boulder 40 Boulder
100
20
0
0
-100 120
Gothenburg 40
Gothenburg
80
20
40
0
0
-20
0 6 12 18 24 0 6 12 18 24

Fig. 7.3. Averaged diurnal cycles of the fluxes of hydrocarbon-like organic aerosol (HOA) and nitrate (NO
3 ) over a range of cities as measured by eddy covariance using an Aerodyne
Aerosol Mass Spectrometer (data from Nemitz et al., 2008a (Boulder); Phillips et al., 2007 (Manchester, London) and Thomas, 2007 (Gothenburg)). The grey range indicates the
5th–95th percentile.

aerosols using realistic coupled sectional approaches, particularly (cf. e.g. Pryor et al., 2008b). The data must be treated with
with respect to secondary organic aerosols. However, it could be circumspection as (a) little account of particle composition is
stated that the scientific community has not delivered any signifi- provided in many of the studies reporting these data, (b) the sizes
cant improvement in the accuracy of model predictive capabilities reported are an ad hoc mixture of optical, electrical mobility, mass
for the atmospheric aerosol deposition pathway over the last two and aerodynamic diameters with little quantitative information on
decades, compare for example one of the first reviews of model the influence individual measurement techniques may have in
uncertainty, Ruijgrok et al. (1997), with e.g. Petroff et al. (2007b). altering actual depositing particle size (which will be particle
The aerosol modelling and composition community are pushing growth factor and hence composition dependent) and (c) system-
ahead with such developments whilst seemingly unaware of the atic detailed information on the morphology of the surfaces is not
poor state of knowledge of deposition processes and caution is always reported which hinders model development.
required to avoid simply wasting research effort here. Whether the
current level of understanding and uncertainty is acceptable 7.4.1.1. Friction velocity (u*). Increased turbulence increases trans-
depends on the compound of interest and its position with the port in the turbulent part of the atmosphere, decreases the effective
aerosol mass size distribution prevalent in the atmosphere. Little thickness of the quasi-laminar sub-layer and increases the drag
in the way of detailed sensitivity studies has been available since coefficient. It is therefore not surprising that Vd increases with
Ruijgrok et al. (1997). Feedback of such sensitivity studies to the increasing u*. Most modelling approaches and measurements indi-
measurement community would also appear to be an area that cate a near-proportion relationship between Vd and u* for submicron
requires improvement (Zhang and Vet, 2006). particles (Fig. 7.4).
There has also been little in the way of any new laboratory
investigation, at least for atmospherically relevant conditions, that 7.4.1.2. Surface roughness length (z0) and canopy morphology. The
can usefully inform these communities on specific gaps in knowl- effect of the surface roughness length (for closed canopies:
edge that need to be pursued. While some progress is being made z0 w 0.1  canopy height) extends beyond its effects on increasing
to highlight gaps in knowledge with respect to models (Petroff u*. Vd for forest tends to be by a factor of 5–10 larger than Vd for
et al., 2007b), these again show that different model descriptions of grass, due to the increased height of the canopy and leaf area index
atmospheric particle deposition, which rely on very limited semi- and the enhanced turbulence induced by forest canopies. Davidson
empirical data and highly tunable, collection efficiency parame- et al. (1982) showed theoretically that even for the same vegetation
terisations, are as variable or more so than the atmospheric type (grassland), Vd may change within a factor of 5, depending on
observations that do exist (e.g. Zhang and Vet, 2006). Some the exact morphology of the vegetation. Similar results were more
improvements in relating natural surface morphology descriptions recently obtained by Petroff et al. (2007b) for forests, emphasizing
to wind tunnel studies have been made. the influence of leaf dimensions and orientation on Vd. There is
In the following we explore the dependence of particle depo- measurement based evidence as well: Ould-Dada (2002) used wind
sition velocity (Vd) on key parameters, in comparison with tunnel studies to investigate the dry deposition velocity of submi-
measurements. Most of these measurements were made over cron particles to model Norway spruce (Picea abies). A total canopy
forests, where they are generally easier to obtain than over short deposition velocity of 5 mm s1 was found which is in line with
vegetated surfaces for many reasons related to the micrometeoro- previous micrometeorological measurements to real forest cano-
logical flux technique. The fluxes have been determined using pies reported in the literature. However, the deposition pattern was
mainly but not exclusively direct micrometeorological techniques found to be a highly complex function of height within the canopy.
D. Fowler et al. / Atmospheric Environment 43 (2009) 5193–5267 5235

tall vegetation:
10 forest (0.10-0.18 μm) (Gallagher et al., 1995)
forest (0.18-0.24 μm) (Gallagher et al., 1995)
forest (0.24-0.30 μm) (Gallagher et al., 1995)
forest (0.30-0.50 μm) (Gallagher et al., 1995)
forest (0.05 μm) (Gaman et al., 2004)
forest (0.05-0.75 μm) (Lamaud et al., 1994)
1
forest (0.05-0.06 μm) (Pryor et al., 2009)
forest (fog) (Vermeulen et al., 1997)
forest (sulphate) (Wesely et al., 1983)
Vd [cm s ]
-1 forest (< 0.02 μm) (Buzorius et al., 2001)

short vegetation:
0.1
grass (< 2 μm) (Allen et al., 1991)
grass (fog) (Dollard & Unsworth, 1983)
grass (sulphate) (Neumann & den Hartog, 1985)
grass (sulphate) (Nicholson & Davies, 1987)
grass (sulphate) (Wesely et al., 1985)
0.01 field/snow (0.15-0.30 μm) (Duan et al., 1988)
field/snow (0.50-1.0 μm) (Duan et al., 1988)
moorland (0.12-0.13 μm) (Nemitz et al., 2002)
moorland (0.15-0.16 μm) (Nemitz et al., 2002)
moorland (0.22-0.24 μm) (Nemitz et al., 2002)
moorland (0.4-0.5 μm) (Nemitz et al., 2002)
0.001 heathland (0.4-0.5 μm) (Nemitz et al., 2004)

0.0 0.2 0.4 0.6 0.8 1.0


-1
u* [m s ]

Fig. 7.4. Dependence of small particle deposition on friction velocity (u*) for a range of surfaces. Adapted from Pryor et al (2008b). Measurements over aerodynamically rough
vegetation are shown with bold lines and symbols, while thin lines and symbols are used for measurements over short vegetation. The greyscale of lines and symbols refers to the
measured or estimated particle size (darker shading referring to larger sizes).

More studies such as these are needed to improve model devel- Interception and impaction are most effective in the intermediate
opment and in-canopy gradient measurements are needed to size range, but less effective than the other two processes. The
validate multi-layer deposition models. resulting trough in Vd(Dp) (Fig. 7.5) is partially responsible for
the survival of the accumulation mode in the atmosphere.
7.4.1.3. Particle diameter (Dp). Both theoretical approaches and When comparing measurements of Vd(Dp) with each other or
measurements show an effect of particle size on Vd: the main with model results it needs to be borne in mind that different
deposition processes (Brownian diffusion, interception, impaction instruments measure different types of diameter (i.e. geometric,
and gravitational settling) are all size-dependent. Brownian diffu- aerodynamic, vacuum-aerodynamic, electro-mobility and optical),
sion is responsible for high Vd for small particles (Dp < 100 nm), which are often difficult to compare without exact information
while gravitational settling is the dominant process for Dp > 5 mm. on particle shape and composition. Some instruments dry the

Fig. 7.5. Evolution of the deposition velocity Vd with the particle diameter Dp on grass and grass-like canopies (lhs) and coniferous canopies (rhs) for friction velocity between 0.35 and
0.56 m s1, as given by various measurement campaigns and six existing models from the literature. Canopy characteristics used by models are hc ¼ 0.07 m, z0 ¼ 0.01 m, LAI ¼ 4,
dn ¼ 3 mm, a ¼ 1.78 for grass and h ¼ 17 m, hc ¼ 7 m, z0 ¼ 1 m, LAI ¼ 22, dn ¼ 1 mm, a ¼ 3.81 for forest. Deposition velocities are recalculated at the same reference height zR ¼ ¼ 100z0.
The parameters of Slinn’s model (1982) are fIN ¼ 0.01, dr ¼ 20 mm, cv/cd ¼ 1/3, b ¼ 2. The model of Zhang et al. (2001) is applied on Land Use Categories #6 (grass) and #1 (evergreen-
needle-leaf trees), the corresponding parameters being, respectively, fIM ¼ 1.2 and 1, and fB ¼ 0.52 and 0.56; from Petroff et al. (2007b).
5236 D. Fowler et al. / Atmospheric Environment 43 (2009) 5193–5267

particles, while others measure the wet size. The wet size at the reported with measurements (in particular on canopy structure). The
measurement height may not reflect the size at which they impact high sensitivity of the models to the canopy structure further adds
with the surface, due to water uptake or release during the depo- uncertainty to the application of simple generalized parametrisations
sition process, in a response to humidity gradients. Despite recent at the regional scale, where input parameters are limited.
advances in measurement technology, it is still not known how
particle composition, particle shape and particle hygroscopicity 7.4.3. Dry deposition rates to urban areas
influence micro-scale deposition mechanisms and collection Very little is currently known about deposition rates to urban
efficiencies onto and by different surface types with very different areas, despite their importance for estimating the contribution of
microstructures, which are known to significantly influence e.g. aerosol deposition to the soiling and weathering of buildings, and
aqueous phase aerosol contact angle and therefore the likely for the atmospheric lifetime in the atmosphere (Pesava et al., 1999).
collection efficiency. Although aerosol deposition measurements have been made in
cities, these were usually made with surrogate collectors (e.g. Yun
7.4.1.4. Atmospheric stability (z ¼ 1/L). There is strong evidence et al., 2002), which are unlikely to be representative of the uptake
from a range of studies that Vd can be greatly enhanced in by urban structures. Alternatively, the soiling of building has been
unstable conditions. Fig. 7.6 summarizes the findings from several studied, which provides information of the aerosol deposition to
field studies, by exploring the dependence of Vd, normalized by a particular receptor, but not on the net removal rate from the
u*, on the inverse of the Monin–Obukhov length (L), a standard atmosphere to the urban matrix (Horvath et al., 1996).
measure of atmospheric stability. The cause for this enhancement Application of micrometeorological flux measurement techniques
is not fully understood and therefore difficult to reproduce in to the urban environment has now been demonstrated (see Section
numerical models. 7.2.2 above). However, the net flux of particles above urban areas is
dominated by emission sources from the city and deposition rates
7.4.2. Parameterising and modelling deposition rates can only be applied if a chemical aerosol species or size-class is found
The first detailed sensitivity study of aerosol deposition models which is not emitted from the city. Nemitz et al. (2000d) presented
compared to field observations was conducted by Ruijgrok et al. initial measurements of chemically resolved aerosol fluxes at
(1997). They reported a factor of 5 uncertainty in model predictions a coastal site and showed that chloride was deposited at low wind
for submicron aerosol deposition velocities based on input uncer- speeds, presumably reflecting deposition of sea salt, while it was
tainty. Despite this there was general consensus that dry deposition emitted during windy periods, probably reflecting wind-driven
measurements (mainly by gradient filter pack and throughfall tech- resuspension of previously deposited material. Similarly, Nemitz
niques) yielded deposition rates significantly larger than analytical et al. (2008) measured SO2- 4 deposition fluxes to urban environment
models were predicting for forested surfaces compared to grasslands with the Q-AMS EC system. These measurements suggest deposition
(Erisman, 1993), which led to some improvement in aerosol collec- velocities in the range of 2–6 mm s1, but will need to be supported
tion efficiency descriptions in models. Later the first eddy covariance with data from other cities. As mentioned above, Schmidt and Klemm
measurements at the same locations tended to confirm this, (2008) detected net deposition of PM2.5 to a German town, but these
notwithstanding the sampling issues associated with aerosol growth fluxes almost certainly contain an upward component.
factors mentioned below (Erisman et al., 1996). The latest review
(Petroff et al., 2007b) suggests this uncertainty has now been 7.5. Uncertainties
reduced, to around a factor of 3. Efforts to reduce this uncertainty
further are restricted by the quality of the measurements (as dis- 7.5.1. Uncertainties in the application of micrometeorological flux
cussed in Section 6.4 below) and the completeness in the metadata measurement techniques for deriving the local flux
Technical challenges in applying micrometeorological techniques
to the measurement of aerosol fluxes go beyond those encountered
0.04 for gas flux measurements. Aerosol fluxes are often small and
grass; Wesely et al. (1985) deposition rates slow, resulting in small concentration differences
forest; Gallagher et al. (1995)
heathland; Nemitz et al. (2004) that need to be resolved for gradient and REA measurements. Simi-
Dp = 0.1 μm larly, the relative corrections, e.g. for density fluctuations (Webb et al.,
0.03 0.2 μm 1980), may become large and can easily result in reversal of the sign
0.3 μm
of the flux. A continuing challenge of particle flux measurements is
0.4 μm
0.5 μm that, due to the limited counting statistics of the measurement,
Vd / u*

standard data processing techniques, such as tests and corrections


0.02 based on co-spectral analysis, and non-stationarity tests are difficult
to implement. While most gas analysers respond to many thousands
of molecules per tenth of a second, aerosol counters may only detect
tens of particles resulting in statistical uncertainties (Fairall, 1984).
0.01 Similarly, in mass-based measurements the contribution from a few
large particles may greatly affect gradients and EC results. Different
aerosol measurement instruments respond to different parameters,
not all of which are conserved. For example, artificial fluxes may be
0.00 introduced if particles are sized according to their wet size which
-0.04 -0.02 0.00 0.02 responds to humidity fluctuations (Kowalski, 2001), and similar
-1 problems are caused by volatilisation or formation in the atmosphere
1/L [m ]
which is discussed in more detail below. The exact effect on the
unstable stable measurement depends on the setup and requires careful consider-
Fig. 7.6. Summary of the dependence of aerosol deposition velocity on the Monin–
ation: most previously reported submicron aerosol number fluxes
Obukhov length (L), indicating a sharp increase of normalized deposition velocity (Vd/u*) have relied on optical particle counting techniques that use closed
in unstable conditions; from Pryor et al. (2008b), modified. path, high power active laser cavity scattering cells. Furthermore
D. Fowler et al. / Atmospheric Environment 43 (2009) 5193–5267 5237

most of these, but not all, adopt recycling dry particle free sheath air is applied to a chemically resolved (but not size resolved) particle
flow with significantly lower RH than the ambient aerosol flow to ensemble, evaporation/condensation and/or heterogeneous chem-
minimize contamination of the instrument optics. These instruments istry in/on particle surfaces could cause flux divergence. It can also
therefore most likely provide a measure of the dry or partially dry modify fluxes if it leads to growth or shrinkage across the cut-off size
particle size and not the ambient particle size. Hence measurements of the particle probe. Apparent emission fluxes with a CPC setup above
of submicron particle fluxes reported in the literature are likely a grassland fertilized with NH4NO3, attributed these fluxes to aerosol
representative of ‘‘dry or near dry aerosol deposition velocities’’ growth due to NH3 and HNO3 uptake and used the fluxes to derive
of optical particle size and not the actual ambient deposition velocity, particle growth rates across the 11 nm cut-off of the CPC (Nemitz et al.,
whereas most large particle fluxes are likely a combination of both, 2008). This demonstrates that flux measurements can be used to infer
some being derived from closed path and other open path instru- information on S and thus on aerosol processing, if the true deposition
ments. Flux measurements using photometric techniques to deter- rate can be estimated independently.
mine total PM2.5 and PM10 mass fluxes may not be subject to these The partitioning of species between the gas and particle phase is
effects and for routine measurement of total mass fluxes e.g. for also associated with gas flux divergence (Soerensen et al., 2005) and
network applications may be the most appropriate when combined can change the net rate of surface uptake of, for example, nitrate if
with additional composition measurements. the deposition velocities of the gas and particle phase species differ
Care needs to be taken that the inlet system does not respond in substantially (Pryor and Soerensen, 2000). The likelihood of flux
a way that may be correlated with w. This could occur during non- contamination due to non-conservative behaviour can be estimated
isokinetic sampling of coarse particles or during fast switching of using time-scale analysis (De Arellano and Duynkerke, 1992), and
inlet flows in REA systems. There is also evidence that non-statio- can be quantified by deploying eddy covariance measurement
narities may affect aerosol exchange particularly often (Fontan systems at multiple heights.
et al., 1997).
(i) Horizontal advection, due to the presence of large spatial
7.5.2. Relating measured fluxes to surface exchange: flux gradients in particle number, mass and/or composition. The
divergence and the effect of chemical interactions importance of horizontal advection has been extensively
A further uncertainty of the flux estimation with micro- evaluated in the carbon dioxide flux community (Baldocchi
meterological techniques is that, although the local flux at the et al., 2001; Hong et al., 2008) and is likely to be important to,
measurement height may be correct, it may differ from the actual but with few exceptions (Vong et al., 2004) the horizontal
surface/atmosphere exchange. The most commonly applied form of advection term has generally been neglected in most particle
the scalar conservation equation is (Pryor et al., 2008b): flux studies. The potential influence of horizontal advection
can be quantified using a horizontally dispersed measurement
array.
(ii) The influence of non-local or ‘top–down’ processes in dictating
vertical exchange. Observed scalar fluxes near to the ground are
derived from two components: local surface-driven turbulence,
Here term (1) is the local change in concentration, term (2) and non-local or ‘top–down’ processes such as entrainment of
advection by the mean flow, term (3) represents the divergence of air from above the mixed-layer which can cause fluxes that are
the turbulent flux, term (4) vertical transport by diffusion, term (5) counter to local gradients (Holtslag and Moeng, 1991). The
vertical transport by sedimentation and term (6) concentration importance of non-locally induced turbulence in dictating
changes due to sources or sinks. observed fluxes has been documented in gas exchange studies
Methods of estimating particle (and other scalar) fluxes at the (Gao et al., 1989), but received less attention in the aerosol
air–surface interface have typically relied on the assumptions of community, despite evidence that it plays a substantial role in
horizontal homogeneity, steady state, the absence chemical source dictating flux magnitudes and may provide an explanation for
or sink of the scalar, that the constant flux layer assumption applies upward fluxes in environments that have traditionally been
to the lowest tens of meters above the surface (Businger et al., viewed as solely particle sinks (Pryor et al., 2008a). The potential
1971), and that the turbulence responsible for transporting the influence of ‘top-down’ processes on observed near-surface
scalar of interest is locally induced (Monin and Obukhov, 1954; fluxes can be identified using scalar correlations (Sempreviva
Monin and Zilitinkevich, 1974). However, as described in this sub- and Gryning, 2000) and quadrant analysis.
section, there are multiple causes of flux divergence (i.e. that the
flux observed at some height above the surface is not equal to that With the advent of techniques to measure compound-resolved
at the surface). Three dominant sources of particle flux divergence aerosol mass fluxes, there is growing evidence that particle depo-
are described below along with methods for their identification and sition velocities of NO þ
3 and NH4 to semi-natural vegetation tend to
quantification: exceed those derived for SO2 4 or from particle number flux
Non-conservative behaviour of the scalar under study (i.e. parti- measurements (Nemitz et al., 2004b; Thomas et al., 2009) (Fig. 7.7).
cles) due to the interaction of other particle dynamics processes with the The deposition rate of these compounds measured over heathland
vertical exchange ði:e: Ss0Þ. The degree to which S deviates from 0 (i.e. and forest greatly exceed those predicted theoretically for short and
the magnitude of the vertical flux divergence due to phase transitions) tall vegetation, respectively (Fig. 7.2). The likely cause is evapora-
is determined by; the chemical climate (Nemitz and Sutton, 2004; tion of NH4NO3 near the ground during the deposition process,
Nemitz et al., 2004b; Sutton et al., 2007), particle ensemble (Pryor and where thermodynamic equilibrium favours the gas phase, due to
Binkowski, 2004), and specific aspect of the particle ensemble being the depletion of NH3 and HNO3 by deposition of these reactive
observed. If Eq. (3) is applied to consideration of the mass of the entire gases to foliar surfaces and warm surface temperatures. Thus, the
particle ensemble, then only mass transfer (i.e. evaporation and/or flux measured well above the canopy is not limited by the physical
condensation) can result in flux divergence. While if Eq. (3) is applied interaction of the particles with the vegetation surface, but reflects
to a size-resolved number particle ensemble for any given particle the evaporation sink in the airspace above. This is supported by
diameter, could Ss0 derive from concentration changes resulting the fact that the relationship between Vd and u* does not differ
from nucleation, coagulation, and condensation/evaporation. If Eq. (3) between surfaces (Fig. 7.7), which implies that turbulent transport
5238 D. Fowler et al. / Atmospheric Environment 43 (2009) 5193–5267

40 be large (Nemitz et al., 2002a). Often important parameters (e.g.


+
NH4 to heathland (Nemitz et al., 2004) leaf dimensions) are not provided in the scientific papers to provide
-
NO3 to oak forest (Thomas, 2007) the input parameters to apply the models to the measurement
30 datasets. Thus there is an urgent need to harmonise flux measure-
ment approaches as much as possible and to provide guidance
for auxiliary parameters that should be measured and provided with
Vd [mm/s]

the measurement datasets to maximise their potential for model


20
evaluation.

7.6.1.2. Improved measurements in the accumulation mode. There


10 are still significant gaps in observations, particularly with respect to
the critical size range associated with the transition between
Brownian and turbulent impaction dominated particle capture
0 regimes. This size range is still poorly resolved by eddy covariance
0.0 0.2 0.4 0.6 0.8 1.0 techniques (w0.5 < Dp < 2 mm) as a consequence of few measure-
u* [m/s] ments being available by any suitably characterized techniques and
the large errors associated with these. The predicted minimum in Vd
Fig. 7.7. Apparent nitrate and ammonium deposition velocities derived from chemi-
as a function of size in this transition regime can vary widely
cally speciated micrometeorological flux measurements. The large values are indicative
of an additional loss of ammonium nitrate near the surface, due to evaporation. between different models (Fig. 7.5) and this will have significant
consequences for long-term integrated dry deposited mass fluxes.
This is therefore seen as a key area in need of attention by both
(which scales with u*) is the main constraint on the flux. Since NH3 models and field observations.
and HNO3 deposit to semi-natural vegation much more effectively
than NH4NO3 aerosol, this shift to the gas phase increases 7.6.1.3. Understanding the effect of stability and leaf properties on
the deposition rate of total ammonium and total nitrate. Since this deposition velocities. Another serious issue is the lack of any
evaporation only occurs close to the canopy, it represents a non- detailed testable hypothesis in models explaining the link between
resolvable (sub-grid) process in traditional transport models. increasing Vd and atmospheric stability and which most measure-
Future parameterisations of Vd should account for this additional ments have reported in the literature for particle sizes Dp < 0.5 mm,
sink for highly volatile aerosol components. most clearly seen for Aitken and small accumulation mode sizes.
So far there are no wind tunnel studies of aerosol deposition to
7.5.3. Interpretation of measurements for model verification vegetated surfaces that take account of atmospheric stability and
Deposition is effectively a number dominated process which, is these are needed to allow further model development. Studies of the
subject to large uncertainties. Mapping from number to mass space hydrophobicity and anti-adhesion of non-smooth leaf surfaces show
requires detailed knowledge of composition, size, shape and hygro- that the morphology of plant epidermal cells and the morphology
scopicity. The hygroscopicity of aerosols can potentially generate the and distribution density of epicuticular waxes significantly affect
largest uncertainty, not just in the measurement, but also in inter- their hydrophobicity and anti-adhesion properties and potentially
preting the measurements. For example, the size at which a particle the adhesion of aerosol particles following impaction and inter-
interacts with the vegetation surface often differs from its actual size ception, Ren et al. (2007). The microstructure of plant surfaces has
at the measurement height, which again may differ from the size been well documented but an interesting phenomenon which might
reported by a given instrument (dry vs. wet; geometric vs. optical have potentially serious implications for some studies of dry and wet
diameter etc.). Most previous studies of deposition have not deposition is the so-called self-cleaning mechanism of some leaf
been sufficiently complete to address any of these uncertainties structures (referred to as the ‘‘Lotus Effect’’). Some plant leaves are
with respect to a complete closure of aerosol number and mass for completely lacking in microstructures while others can have sunken
model comparison through use of growth factors. As a consequence, or raised nervatures which as a consequence cause super hydro-
size-segregated and chemically speciated eddy covariance (and phobic behaviour. This in turn leads to a remarkable self-cleaning
related) aerosol flux measurement techniques cannot currently process whereby fog droplets e.g. rolling down the leaf surface pick
provide unambiguous results of particle number or mass fluxes to up aerosols and remove them from the leaf surface. Experiments
surfaces as a consequence of fluctuations in aerosol size distributions whereby leaf surfaces have been artificially contaminated with
on timescales that can lead to sampling biases. This sampling radioactively tagged aerosols and then subjected to artificial fog
ambiguity and the potential for bi-directionality in aerosol fluxes droplets have been used to determine the retention rate of aerosols
limit the accuracy with which Vd can be determined and hence will to plant surfaces and these can range from over 90% to less than 10%
hamper improvement in model mechanistic descriptions of particle depending on the species examined and which were linked to
deposition to natural vegetated surfaces. differences in leaf microstructure and orientation (Neinhuis and
Barthlott, 1997).
Considering the many inherent uncertainties in field flux
7.6. Future research needs measurements more wind tunnel studies are needed under better
controlled conditions of stability, surface morphology and aerosol
7.6.1. Deposition measurements and reporting composition. The question is how detailed should a model be to
7.6.1.1. Standardisation of eddy covariance approaches and data describe adequately the surface interactions with aerosols (whether
analysis procedures. Comparisons between different measurement they be ‘‘dry’’ or ‘‘wet’’ aerosols)? And more importantly how can
systems are currently made difficult by the diverse approach used to their importance be measured? What information should be
measure the fluxes as well as to analyse and present the data. For reported on surface morphology for future model development in
example, some authors have derived parameterisations of Vd(Dp) order to attempt inclusion of these effects in future? These questions
averaging over the negative (deposition) fluxes only, while other are best tackled by revisiting wind tunnel studies coupled with
authors have averaged over the entire datasets. The difference can modern particle measurement techniques.
D. Fowler et al. / Atmospheric Environment 43 (2009) 5193–5267 5239

7.6.1.4. Filtering or accounting for chemical interactions and water variability however these have focussed on dust suspension and
uptake. A particular aspect of the data processing is the correction for the subsequent impact on both horizontal and vertical dust mass
aerosol dynamics due to water equilibration and/or chemical inter- fluxes, which can be considerable (Gillette and Chen, 2001).
actions. A growing number of datasets indicates that size-segregated As deposition mechanisms for submicron aerosols are
particle number fluxes are affected by chemical effects (Nemitz et al., controlled by turbulent interaction with surfaces, and are highly
2004a,b) and this is confirmed by the first results from the chemically sensitive to particle size and micro-scale structures, it is likely that
resolved mass fluxes from the Q-AMS eddy covariance system, which deposition velocities too are also affected by surface anisotropy and
indicates that often some chemical aerosol components may be the manner in which this links to the microstructure. The notion of
emitted at the same time as others are being deposited. Although a mean aerosol deposition velocity in this context has little value
some modelling studies have been successful in qualitatively (much as it is now thought to be for suspension fluxes) (Okin, 2005)
reproducing the observations both for bulk chemical fluxes and size- and a probabilistic approach must be used. Unfortunately, unlike
segregated fluxes (e.g. Nemitz and Sutton, 2004; Van Oss et al., 1998), dust emission mass fluxes, there are virtually no observations of the
standardized operational procedures for correction have not yet been impact of sub-grid scale surface anisotropy on aerosol deposition.
developed. Indeed, we do not currently have the strategies in place Initially, a theoretical model study could explore the likely impact
to test whether a particular dataset may be affected by chemical of anisotropy on effective dry deposition rates, for example
interactions. It is unclear whether correction procedures will ever be adopting the concept of a lateral cover parameter (l) as a measure
sufficiently accurate to fully correct for these effects, given the small of the vegetative canopy area intercepted by the wind and its
value of the deposition rates. contribution through surface drag to the surface roughness, from
which an effective aerodynamic roughness length for the landscape
7.6.2. Deposition models can be calculated (1993). In parallel, as a first step to improving
7.6.2.1. Migration to a probabilistic approach. Comparison of understanding in this area (which has been relatively moribund for
measured deposition velocities as a function of size with different some considerable time) (Pryor et al., 2008b) the community needs
regional scale model descriptions show large differences. Hence, to collect high quality micrometeorological aerosol flux measure-
unacceptable errors will very likely be incurred in annual cumulative ments over a number of different surfaces with very different
mass deposition values. A detailed sensitivity analysis between anisotropic variability. The observations should focus on deter-
different deposition schemes used in current regional models with mining the frequency distribution function, f(Vd (Dp)) of aerosol
observations has not yet been undertaken. Given the large apparent deposition velocities.
difference between measurements and model schemes, it may be
more appropriate to move towards a probabilistic approach in deriving 7.7. Conclusions – aerosols
deposition estimates, by exploring a range of possible solutions,
together with statements on their probability. For this purpose, prob- After little progress in the understanding of surface/atmosphere
ability density function distributions for aerosol deposition velocities exchange of aerosols in the 1980s and early 1990s, the development
need to be developed which can be used to test sensitivities to these of novel instrumentation suitable for flux measurements has led to
factors in regional transport and global climate models. new investigations into the surface exchange, extending micro-
meteorological flux measurements to the urban environment
7.6.2.2. Improvement of modelling approaches. New modelling and the sea. For example, new developments in mass spectrometry
approaches compare favourably with the available measurement have enabled the first eddy covariance flux measurements of
2
database, with the caveats on the data quality mentioned above. aerosol components (NO 3 , SO4 and organics) above urban areas
Fig. 7.5 demonstrates that the model of Petroff et al. (2007b) in and vegetation, providing new information on sources, sinks and
particular appears to be successful, similar to earlier modelling results chemical processing, together with deposition rates of the accu-
of Davidson et al. (1982). Both models have in common that they mulation mode and the potential of studying deposition rates in
include a detailed description of the canopy morphology. This intro- relation to particle composition. Furthermore, size-segregated
duces additional requirements for input parameters and further particle flux measurement approaches have now been extended to
degrees of freedom for adjustments to make the model match the the sub-100 nm size range, providing the first data for model
measurements. However, good agreement is achieved with measured evaluation. The first long-term flux measurements of total aerosol
canopy characteristics, increasing the confidence in the modelling number provide increasingly robust datasets of removal rates.
approach. However, the model of Petroff et al. (2007b) needs to be As more detailed flux measurements as a function of size and
simplified for application in operational chemical transport models composition have become available, it is becoming clear that size-
and standardized characterizations for the different vegetation classes segregated particle number flux measurements are often influenced
need to be developed. by hygroscopic growth and chemical processing. This highlights the
need to minimize or to filter/correct for these effects when the data
7.6.2.3. Impact of surface anisotropy on suspension & deposi- are used for model validation. Apparent upward fluxes have been
tion. Spatial organisation of vegetation on sub-grid scales can used to study the formation of NH4NO3 or biogenic SOA formation.
influence aerosol surface exchange properties by introducing Measured effective deposition rates of NH4NO3 to semi-natural
significant perturbations to mean wind flows by altering the vegetation greatly exceed those of other aerosol compounds, indi-
probability density functions for turbulence velocities above that cating that new sub-grid parameterisations need to be developed to
surface which in turn can alter the magnitudes of aerosol surface account of the additional deposition mediated through the evapo-
exchange fluxes. The impact of this surface anisotropy is often seen ration of volatile aerosol components (e.g. NH4NO3).
in observations of dust suspension fluxes over surfaces where Theoretical developments demonstrate that models of dry
elongated regions may occur which are free of vegetation (e.g. deposition need to account for canopy structure and small-scale
Gillette and Chen, 2001). As a result of this, neighbouring surfaces, morphology. Existing models are now capable of reproducing selected
which have the identical vegetative indexes, can produce dust measurements, but will need to be simplified for operational appli-
fluxes that differ from one another by as much as a factor of 4–8 cation in transport models and incorporate effects of atmospheric
(Okin, 2005). Recently models have been developed that capture stability. The data available for model validation are highly dispersed
and demonstrate the importance of sub-grid cell isotropic spatial in terms of quality, approaches, diameter measured and auxiliary
5240 D. Fowler et al. / Atmospheric Environment 43 (2009) 5193–5267

information provided. Harmonised approaches in data processing and Table 2


presentation are needed. More sensitivity studies and probabilistic Estimates of global N2O and CH4 budgets (Tg y1).

approaches are needed to explore the range of possible deposition N2O sourcea Tg N2O–N y1 CH4 sourceb Tg CH4 y1
estimates. In addition, the potential of modelling concepts that Natural sources
account for small-scale spatial variability (increasingly applied for Oceans 3.8 (1.8–5.8) Oceans 4 (0.2–20)
resuspension) should be explored to estimate dry deposition. Here Atmosphere 0.6 (0.3–1.2) Termites 20 (2–22)
Soils 6.6 (3.3–9) Wetlands 100 (92–232)
a new series of targeted, high-quality wind-tunnel experiments,
Othersc 21 (10.4–48.2)
coupled with the improved measurement technology would help
decrease remaining uncertainties. Anthropogenic sources
Agriculture 2.8 (1.7–4.8) Rice cultivation 60 (25–90)
Biomass burning 0.7 (0.2–1) Biomass burning 50 (27–80)
8. Ecosystem–atmosphere exchange of the radiatively Energy & industry 0.7 (0.2–1.8) Energyd 106 (46–174)
active gases – N2O and CH4 Otherse 2.5 (0.9–4.1) Ruminants 81 (65–100)
Waste disposal 61 (40–100)
8.1. Introduction Total sources 17.7 (8.5–27.7) 503 (410–660)

Sinks
Atmospheric concentrations of the three main greenhouse gases Stratosphere 12.5 (10–15)f Stratosphere 40 (32–48)
CO2, CH4 and N2O have increased since the industrial revolution in Soils 1.5–3g Soils 30 (15–45)
the 18th century due to anthropogenic activities. Increased fossil fuel Tropospheric OH 445 (360–530)
burning, land use change and the intensification of agriculture Total sinks 14 (11.5–18) 515 (430–600)
facilitated by the manufacture of synthetic nitrogen and conse- a
Sources are estimates for the 1990s as provided by IPCC (2007), Table 8.7.
quently population growth are the main causes. Increased fossil fuel b
From Wuebbles and Hayhoe (2002).
combustion is the main cause for rises in CO2, whereas microbial c
Others ¼ marine sediments, geological sources and wild fires.
d
processes in soils, sediments, and waters and rumens of animals, are Energy ¼ natural gas, coal mining and other fuel related sources.
e
responsible for the bulk of the observed increased atmospheric CH4 Atmospheric deposition, aquatic systems, sewage.
f
Hirsch et al. (2006).
and N2O concentrations. In this section, current understanding of g
Cicerone (1989).
CH4 and N2O exchange at the surface, and especially of the biological
processes, measurement methodologies and models, are reviewed.
The wider consideration of the global biogeochemical cycles of these oceans, termites) are also large and dominated global emissions until
trace gases are provided in outline for context only, as these would the 20th century (Table 2). Increased livestock production and fossil
take the review substantially beyond the focus on surface–atmo- fuel use are the main reasons for the atmospheric increase of CH4
sphere exchange. (IPCC, 2007). Soils are are a minor sink for CH4 and accounts for
approximately 6% of the global budget; the dominant removal
8.2. Global budgets of N2O and CH4 process for atmospheric CH4 is oxidation by OH, mainly in the
troposphere.
Atmospheric N2O and CH4 concentrations have risen from back-
ground levels prior industrialisation from 270 to 320 ppb N2O and 8.3. Biological sources of N2O and CH4
from 700 to 1782 ppb CH4 in 2006 (http://www.esrl.noaa.gov/gmd/
aggi/). Nitrous oxide concentration has increased at a relatively 8.3.1. The biology of production and consumption
uniform rate, with a mean annual growth rate over the last 10 years of of N2O and CH4 in soils and sediments
0.76 ppb year1 (Hirsch et al., 2006). By contrast, the growth rate in Microorganims are the dominant sources of N2O and CH4 in the
CH4 concentration has changed considerably since the early 1990s troposphere. A knowledge of the underlying processes and micro-
from a steady monotonic increase of approximately 15 ppb year1 in bial community structure is essential to improve global estimates of
the later decades of the 20th century until the early 1990s, following N2O and CH4. The main microbial reactions involved (nitrification,
which annual rates of change varied between increases of denitrification, methanogenesis and CH4 oxidation) are ubiquitous
5 ppb year1 to decreases of a few ppb year1 (IPCC, 2007). These very to all live containing ecosystems and are all sensitive to anthro-
large and inter-annual variations in CH4 concentration remain unex- pogenic activities (e.g. irrigation, drainage, fertilization) and
plained and present an important challenge to the research commu- climate (temperature and precipitation).
nity (IPCC, 2007). Nitrous oxide is a by-product of aerobic nitrification and an
The global budget of N2O is constrained by the sink strength in the obligate intermediate in the denitrification pathway, and is emitted
stratosphere and atmospheric increase (http://www.esrl.noaa.gov/ by both nitrifiers and denitrifiers. Production and consumption
gmd/aggi); hence, the global source strength is 15. 8–16 Tg N2O– of N2O is regulated by oxygen partial pressure; nitrification is
N y1 (Crutzen et al., 2008, Hirsch et al., 2006. Cicerone, 1989). additionally controlled by the concentration of NHþ 4 , while deni-
There are some indications that also soils may significantly act as sink trification is also controlled by availability of carbon and NO 3
for atmospheric N2O and that the soil N2O reduction has decreased (Conrad, 1995). Denitrification is the main biological process
within the last decades (Chapuis-Lardy et al., 2007; Conen and Neftel, responsible for returning fixed N to the atmosphere as N2, thus
2007). However, this is so far not considered in any global estimate. closing the N cycle (Philippot et al., 2009). This reduction of soluble
The atmospheric increase is largely attributed to agricultural N to gaseous N represents a loss for agriculture, since it can deplete
activity (Table 2). Natural sources of N2O, the oceans, tropical and the soil of NO
3 , an essential plant nutrient. The denitrification N2O/
temperate forests and grasslands/savannahs, are unlikely to have N2 product ratio is variable, and N2O may even be the dominant end
changed much since pre-industrial times except where land use product. However, denitrification also provides a valuable
has changed significantly. ecosystem service by mediating N removal from NO 3 polluted
The CH4 budget is constrained by measurements of the major waters in sediments and other water-saturated soils (Mosier et al.,
sources and d13C signature to a global total of 430–600 Tg y1 1998). Denitrifiers can be sinks for N2O. Sink activity appears to be
(Wuebbles and Hayhoe, 2002, IPCC, 2007). Anthropogenic sources stimulated by low availability of mineral N (Chapuis-Lardy et al.,
contribute 70% of the total budget. Natural sources (wetlands, 2007; Conen and Neftel, 2007).
D. Fowler et al. / Atmospheric Environment 43 (2009) 5193–5267 5241

Methane is produced by methanogenic archaea in anaerobic soil countries, principally northern Europe, the USA, Canada and Japan.
(Philippot et al., 2009). The organisms require low redox conditions For N2O, studies on N fertilized agricultural soils dominate and for
as well as on the fermentative production of precursors for the CH4 studies on rice paddy fields and northern wetlands. Studies
methanogens. The main terrestrial CH4 sources are wetland ecosys- from Asia, where N demand is increasing at a faster rate than
tems, where both methanogens and methanotrophs are present and elsewhere, are now emerging. There are insufficient data from
active. Methane is consumed by methanotrophs active in the aerobic agricultural systems in Central and South American and African
layers of most soils; undisturbed soils are largest CH4 sinks. countries, from new emerging cropping systems, especially biofuel
crops, and land use change in temperate as well as tropical countries
8.3.2. Distribution of active microbial populations in soils to provide the detailed understanding required for model validation
Although many different microbial species can produce and and for inclusion in emission inventories.
consume N2O and CH4, information on the microbial biodiversity can
provide useful insight into the health and functioning of the soil. The 8.3.4. Plant-mediated transport and production of N2O and CH4
development and recent automation of molecular methods have The general view is that soil-based microbial production and
made it possible to characterise the abundance and function of soil consumption of CH4 and N2O are the major processes involved in
microbial populations relatively quickly. One of these methods is the regulating biosphere–atmosphere exchange of these two green-
analysis of the Phospholipid Fatty Acid (PLFA) composition of the house gases. Based on this perception, and combined with the lack
microbial membrane (Bach et al., 2008). The method, together with of appropriate methodologies, our current knowledge about their
analysis of microbial biomass carbon, Gram staining, N mineraliza- exchange rates is almost exclusively based on observations ach-
tion rates, N2O, NO and CH4 fluxes, was applied to soils from arable, ieved using shallow, soil anchored enclosures. For many ecosystems
grassland, wetland and forest ecosystems from the main climate such enclosures may exclude the vegetation (e.g. tall crops and
zones in Europe as part of the NitroEurope Project (http://www. forests) and biases in emission estimates through omission of the
nitroeurope.eu). The PLFA composition provided an overview on pathway through tall vegetation may result from this methodology.
the distribution of functional microbial groups in soils of different It is well documented that soil–atmosphere transport of both
landuses. There was a good separation between microbial commu- CH4 and N2O is mediated by aerenchymatic wetland herbaceous
nities from wetlands and forests, but a closer similarity between species such as rice (e.g. Yan et al., 2000). Mangrove prop roots and
microbes from grasslands and croplands (Fig. 8.1). The ratio of two also wetland and flood-tolerant trees have been shown to mediate
marker PLFAs, cyclic fatty acids and precursor monounsaturated CH4 and N2O transport from the soil to the atmosphere, e.g. through
fatty acids, is an index of bacterial stress. In this study, the stress the bark of black alder or from hybrid poplar seedlings (McBain
parameter correlated with soil NO emissions and these were related et al., 2004), but only under conditions when the root zone was
to N-deposition rates and soil acidity (Pfeffer et al., personal exposed to above ambient concentrations of the gas.
communication). N2O emissions correlated positively with the The role of non-aerenchymatic plants and in particular trees in
abundance of gram-negative bacteria, potential N-mineralization the exchange of CH4 and N2O between the soil–plant system and the
rates and microbial biomass carbon. This can be explained by the fact atmosphere has only been sparsely investigated. Recent investiga-
that gram-negative bacteria contain many microbial groups impor- tions, however, have emphasized a non-negligible role of vegetation
tant for the N-cycle, such as nitrifiers, free-living N2-fixers and in the biosphere–atmosphere exchange of greenhouse gases.
several denitrifiers.
8.3.4.1. Methane from vegetation. In 2006 Keppler et al. reported
8.3.3. N2O and CH4 fluxes from the main global ecosystems a very surprising observation that higher plants had the capability to
Biosphere atmosphere exchange of N2O and CH4 has been emit CH4 under aerobic conditions with a mean emission rate of
studied for over 30 years. The data available are biased towards the 374 ng CH4 g1 dw h1. From their findings they calculated a global
large N2O and CH4 emitting ecosystems in highly developed CH4 source strength of 62–236 Tg y1 for living plants and 1–7 Tg y1
for plant litter, the sum of which equals c. 10–40% of the total global
CH4 source strength. Methyl-ester groups of pectin, an abundant
polysaccharide in cell walls of non-woody plant tissue, served as
a precursor for CH4 (Keppler et al., 2008). UV light appears to be
important in emissions of CH4 from plant material. Vigano et al. (2008)
demonstrated that in the absence of UV light CH4 was not produced
until the temperature reached 70–80  C; with UV light emissions
were significant already at room temperature with rates up to
67 ng CH4 g1 dw h1. McLeod et al. (2008) provided further evidence
that not only CH4 but also ethane, ethylene and CO2 are produced from
methyl-ester groups of pectin under UV irradiance, and that reactive
oxygen species (ROS) arising from environmental stress may have
a role in the formation of CH4 from pectin. By contrast, Dueck et al.
(2007) observed no significant CH4 emissions from photosynthesizing
or dark respiring leaves, adding evidence to speculations that
plant derived CH4 originates from abiotic processes. While processes
generating CH4 within on on the surfaces of vegetation have clearly
Fig. 8.1. Principal component analysis of microbial communites, determined as PLFAs been identified, the up-scaling to large areas and to the global atmo-
(nmol g1 soil dry weight) of 13 NitroEurope sites representing different landuses. sphere remains largely speculative and have not therefore been
Abbreviations: AM: arbuscular mycorhiza fungi; Sites: Forests: FI-Hyy ¼ Hyytiälä, FIN; shown to contribute significantly to the global inventory.
DK-Sor ¼ Sorø, DK; NL_Spe ¼ Speulder Bos, NL; DE-Hog ¼ Högelwald, DE; grasslands:
UK-Ebu ¼ Easter Bush, UK; CH-Oen ¼ Oensingen, CH; HU-Bug ¼ Bugac, HU; croplands:
DE-Geb ¼ Gebesee, DE; FR-Gri ¼ Grignon, FR; IT-Cas ¼ Castellaro, I; IT-BCi ¼ Borgo
8.3.4.2. Nitrous oxide from vegetation. In a number of experiments,
Cioffi, I; wetlands: FI_Lom: Lompolojänkkä, FIN; UK-Amo ¼ Auchencorth Moss, UK; especially crops, plant-mediated emission of N2O have been observed.
Figure provided by B. Kitzler and M. Pfeffer (BWF, Austria). Chen et al. (1999) found N2O emissions up to 2.8 mg m2 d1 from the
5242 D. Fowler et al. / Atmospheric Environment 43 (2009) 5193–5267

plants in a soil–rye grass (L. perenne) system, and plant-mediated N2O and CH4 fluxes and their control by physical, chemical and micro-
emissions from maize, soybean and wheat contributed up to 11,16 and bial processes has largely arisen from flux chamber measurements.
62% to the total sum of N2O emissions, respectively (Zou et al., 2005). In Recent development of high frequency instruments, that detect
contrast, Müller (2003) found that presence of plants in old grassland very small concentration changes, has improved our knowledge of
could induce N2O uptake. However, these studies did not identify the N2O and CH4 biosphere atmosphere exchange at the field/land-
mechanisms underlying the plant based N2O production or reduction scape scale and at a high temporal resolution.
processes. Chang et al. (1998) observed that barley (Hordeum vulgare)
and oilseed rape (Brassica napus) emitted N2O from the shoots upon 8.4.1. Flux chambers
irrigation with water containing N2O, and hypothesized that N2O was Usually closed (non-steady state) chambers are used for N2O
conveyed by the plants from the soil to the atmosphere via the tran- and CH4 flux measurements, e.g. Butterbach-Bahl et al. (1997),
spiration stream. In contrast, Smart and Bloom (2001) found that N2O Conen and Smith (1998). Advantages of chambers over microme-
emissions from wheat (Triticum aestivum) leaves was correlated with teorological techniques are that chambers are low cost and can be
leaf NO3 assimilation activity. They found that N2O was formed during used on small fields/plots. Disadvantages include limited spatial
in vitro NO2 -reductase activity of the leaves and suggested that N2O averaging of a spatially variable quantity due to small area (usually
formation during NO 2 photo-assimilation could be an important <1 m) of most of these enclosures. Recent developments in
global biogenic N2O source. Conversion of 15NO 15
3 to N2O in a range of chamber methodologies include:
aseptically grown plant species was reported by Hakata et al. (2003),
and increased N2O emission from soybean was observed concomitant 1) an inter-comparison of the main chambers types employed for
with an herbicide induced accumulation of plant NO 2 (Zhang et al., N2O and CH4 chambers used within the European community
2000) providing further evidence for in planta production of N2O. (Philatie personal communication 2008), similar to the inter-
The potential for tree species to act as conduits for N2O emis- comparison of soil respiration chambers (Pumpanen et al., 2004).
sions were demonstrated in the work by Pihlatie et al. (2005). In ACCENT has contributed towards the funding of this exercise;
a laboratory experiment with beech (F. sylvatica) seedlings it was 2) the validity of the commonly used linear regression equation to
found that fertilization with 15N-ammonium-nitrate (15NH15 4 NO3) calculate fluxes from non-steady state chambers was ques-
induced foliage 15N2O emissions and exposing the beech roots to tioned, as it may underestimate the true flux. An exponential
elevated N2O concentrations induced significant emissions of N2O approach may be more accurate (Kroon et al., 2008),
from shoots and leaves (Fig. 8.2). Pihlatie et al. also found that 3) development of the fast box method (Hensen et al., 2006)
concentrations of dissolved N2O in leaves in a beech forest canopy facilitates chamber measurements from many spots within the
exceeded ambient atmospheric concentrations, indicating a poten- field and establish a picture of the spatial heterogeneity of N2O
tial for canopy N2O emissions. and CH4 emissions very quickly. This method requires
In summary, substantial evidence exist that plants contribute combining manual chambers with sensitive fast response
directly to the emission of CH4 and N2O. Yet, most work has been analysis of N2O and CH4, for example using tunable diode laser
based on small-scale laboratory work and the scale of the fluxes techniques.
appears small. However, there is an urgent need for field-based
measurements and more detailed explanation of the underlying
processes. 8.4.2. Micrometeorological methods
The development of tunable diode lasers for CH4 and N2O provides
8.4. New developments in measurements of N2O a method of measuring N2O and CH4 biosphere atmosphere exchange
and CH4 and denitrification by micrometeorological methods at high temporal frequency (30 min)
over surfaces where fluxes are reasonably large (approx 20 ng m2 s1
N2O and CH4 fluxes are measured at scales ranging from a few of N2O or CH4). This measurement approach is particularly valuable for
grams of soil to several km. Each scale and method has contributed heterogeneous ecosystems, i.e. grazed grasslands, and soft surfaces,
to our current understanding of biosphere atmosphere exchange of where compaction by walking to a flux chamber may release gases
N2O and CH4 (Denmead, 2008). Our global understanding of N2O into the chamber, e.g. peat wetlands or dung heaps. Eddy covariance
measurements of CH4 for example were made over northern wetlands
in Finland (Rinne et al., 2007) and of N2O for example over grazed
grasslands in Scotland (Di Marco et al., 2004). For well-defined point
sources, such as manure heaps and landfill sites the Gaussian plume
method has been used to calculate the emission strength, by either
walking or driving through the emission plume (Skiba et al., 2006;
Hensen et al., 2006).

8.4.3. Comparison of eddy covariance with chamber methods


Scaling up to the field and regional scale is usually based on data
from small flux chambers. Several studies have been conducted to
establish the validity of this approach. It is interesting that for
N2O fluxes from grasslands and arable soils (Christensen et al., 1996)
chambers strategically placed within the footprint of the micro-
meteorological tower are in reasonable agreement with eddy
covariance. However for CH4 fluxes from rice paddies discrepancies
of a factor of 2–3 between chamber and micrometeorological
method, the chambers giving lower emissions, were reported
Fig. 8.2. Nitrous oxide emissions (mg N2O–N m2 h1) from beech (Fagus sylvatica)
leaves after exposing the beech roots to different concentrations of N2O in the root
by Kanemasu et al. (1995) from the Philippines and Hargreaves
compartment solution. Bars indicate average (þSE) of two different beech seedlings. (personal communication) from a rice paddy field in the Po Valley,
Modified after Pihlatie et al. (2005). Italy. These different observations suggest that more comparisons
D. Fowler et al. / Atmospheric Environment 43 (2009) 5193–5267 5243

need to be carried out and that chambers may well be suitable unfortunately microbial populations have a larger impact on the
for relatively firm surfaces, but not those of low bulk density or isotopic and isotopomer signatures of N2O than the production
complete waterlogged. pathway itself (Sutka et al., 2003). Very recently, the 15N isotopic
abundance of soil-emitted NO was determined for the first time
8.4.4. Recent methodological advances in measurements (Li and Wang, 2008). The authors found very 15N-depleted NO with
of total denitrification rates d15N values down to 50& and could identify both nitrification and
For a full understanding of the processes and accurate simulation denitrification as sources of soil-emitted NO.
of observations using models, we need to know the removal rate of It is concluded that many challenges of quantifying total deni-
N2O in the ecosystem. The only natural process of permanent trification and differentiating between N2O produced by nitrifica-
removal of excess N from ecosystems is denitrification to N2. The tion or denitrification remain.
very high natural background of atmospheric N2 hampers direct
quantification of total denitrification. A wealth of methods has been 8.5. Modelling of N2O and CH4 fluxes at site and regional scales:
developed in the past decades for quantification of total denitrifi- approaches, applications and uncertainties
cation (Groffman et al., 2006). Unfortunately, none is without
drawbacks and even today, there is no method that can be used at Signatory states to the United Nations Framework on Climate
the field scale or at high temporal resolution. The most common Change (UNFCC) are required to produce annual national inventories
method is the acetylene inhibition method (Balderston et al., 1976), of greenhouse gas emissions from all anthropogenic sources,
by which the terminal step of denitrification, i.e. the reduction of including emissions from soils. With regard to CH4 and N2O, soils are
N2O to N2 is inhibited by acetylene. Major drawbacks are that it is not the main sources in their respective global atmospheric budgets. The
easy to achieve 100% diffusion of C2H2 to the active denitrification IPCC (2006) recommends three different approaches (Tier 1–Tier 3) to
sites, that nitrification is also inhibited by C2H2 and that C2H2 provide emission inventories. Tier 1 represents the simplest way to
interacts with NO in oxic environments (Bollmann and Conrad, model or estimate GHG fluxes on site and regional scales. It is a purely
1997). To overcome these problems a completely new concept of statistical approach, relating e.g. soil N2O emissions to the amount of
replacing the background N2 during soil core incubations with applied fertilizer. In the 2007 IPPC reporting guidelines (IPCC, 2006)
a noble gas (e.g. with a He:O2 mixture) has been developed and the default emission factor for direct N2O losses from soils following
facilitates direct measurements of N2O and N2 (Scholefield et al., N fertilization is 1%. However, even if one assumes that this factor is
1997; Butterbach-Bahl et al., 2002). The major drawback is the high representative on a global scale, which has been questioned in the
capital investment in equipment and the time-consuming flushing recent past (Crutzen et al., 2008), a fixed emission factor cannot
procedure to remove N2. consider reported effects of climate, management or soil properties
The use of stable isotope analysis either in tracer studies with on the magnitude of GHG exchange. Therefore, based on a detailed
isotopically enriched tracer compounds or at the natural abundance survey on reported soil N2O emissions worldwide, Stehfest and
level offer promising alternatives, but very little progress has been Bouwman (2006) developed a more detailed statistical approach for
made in the last 5 years. Application of 15NO 3 containing fertilizer calculation of emission inventories, which also considers general
and monitoring 15N-labelled N2O and N2 provides a suitable tracer environmental factors such as climate, texture and soil organic carbon
for denitrification to N2 for agricultural N fertilized soils, but not in contents and management related factors such as fertilization rates
N-poor environments. For these 15N tracers can artificially stimulate and crop types. However, beside the fact that the demand on required
N turnover, microbial immobilisation or dissimilatory reduction of input information is much larger, this approach also has its weak-
NO þ
3 to NH4 . For N-poor environments natural abundance of N and O nesses: a) the high uncertainty of the developed statistical model,
isotopes may offer an alternative, as due to kinetic isotope frac- b) the rough classification scheme (which is due to the limited
tionation the intermediates and the end product of denitrification availability of field datasets describing GHG emissions for different
become increasingly depleted in 15N, whereas the remaining environmental conditions) and c) incomplete coverage of pulse
15 18
soil NO 3 becomes increasingly enriched in N and O. If substrate is events, which may dominate annual site budgets.
not limiting, large kinetic N isotope fractionation factors of up To account for the huge spatial and temporal variability of GHG
to 40& can be observed during denitrification (Groffman et al., fluxes on site to regional scales the development and use of process-
2006). However, if denitrification is limiting or rates are small, as in oriented models may at present be the most promising approach
the case for N-poor ecosystems, the apparent N isotope fractionation (Butterbach-Bahl et al., 2004). These models simulate the GHG
is too small to provide unambiguous interpretation of the data. exchange at a given site based on the underlying processes, i.e. by
Dual-isotope labelling with 15N and 18O-enriched NO 3 can simulating the dominant physico-chemical, plant and microbial
identify nitrification or denitrification as source of N2O; this infor- processes involved in ecosystem C and N cycling and associated GHG
mation is desirable for the models (Wrage et al., 2005). However, exchange (Li et al., 2000). As a general assumption, one defines
there are problems. At low pH NO 2 , intermediate of nitrification that the controlling factors for e.g. microbial C and N turnover such
and denitrification, rapidly undergoes O-isotope exchange with as temperature, moisture and substrate responses, are comparable
water (Casciotti et al., 2007). This O-isotope exchange might lead to across different climatic zones and landuses and that by capturing
misinterpretation of the results when stoichiometric relationships the major biogeochemical processes within an ecosystem it is
in the different N2O formation pathways are assumed, and is very possible to predict the temporal variability of fluxes. Such models
likely the cause of O-isotope exchange between N2O and water, as require a thorough process understanding of the coupled C and
reported by Kool et al. (2009). N (P) cycles, even though the level of process description may vary
The most recent approach of quantifying denitrification rates and between the models currently in use (e.g. Li et al., 2000). However,
differentiating between nitrification and denitrification as sources of these models also have the drawback, that modelling of ecosystem
N2O is the analysis of N2O isotopomers. Intramolecular physico- processes involves a huge dataset of parameters needed to describe
chemical site differences between terminal and central N atom lead heat transfer, water movement, plant and microbial growth or
to differences in N isotope ratios between the two positions during anthropogenic management. In comparison to statistical approaches
N2O formation and consumption. Differences in this so-called mechanistic models often show an improved performance with
15
N site preference have been attributed to N2O production during regard to reproducing observed differences in GHG fluxes between
nitrification and denitrification, respectively (Pérez et al., 2001), sites, seasons and management practices (e.g. Kesik et al., 2006).
5244 D. Fowler et al. / Atmospheric Environment 43 (2009) 5193–5267

However, the increasing use of mechanistic models also shows This is a shortcoming, which needs to be properly addressed in
that we still need to improve our process understanding, e.g. the future work. Nevertheless, biogeochemical models offer a great
regulation of microbial processes and its dependence on microsite opportunity to improve our understanding of ecosystem processes
variability of environmental conditions such as redox potential or and GHG exchange and they will play an important role in identi-
the feedback of temperature on organic matter decomposition. fying and predicting consequences and feedbacks of global changes
A good example of the deficiencies in understanding N2 vs. N2O (climate and land use change) for ecosystem functioning and
production during denitrification is provided by (Groffman et al., biosphere–atmosphere trace gas exchange.
2006). This gap in knowledge also precludes the parameterisation
of the denitrification process in biogeochemical models such as
8.6. Validation of models by landscape and regional
DNDC or DayCent, which in consequence leads to a systematic
scale measurements
underestimation of N2 losses using either model.
Biogeochemical process models have recently been used in
Developments in inverse modelling and direct large-scale
a number of studies for calculating regional soil GHG emission
measurements provide very powerful tools to constrain and verify
inventories (Fig. 8.3). Thereby regionalisation is achieved by
our bottom up models and inventories. For example, Bergamaschi
coupling of the models to GIS databases holding all the relevant
et al. (2005) compared inverse models with national bottom–up
information needed for initializing (soil and vegetation properties,
inventories for CH4. These developments are taken further within
management) and driving (meteorological conditions) the models
the NitroEurope project (www.nitroeurope.eu).
(Kesik et al., 2006). Such an approach partly neglects landscape
The development of instruments sensitive enough to measure
processes, such as e.g. lateral flow and transport of nutrients and
very small concentration differences has made it possible to
sediments via leaching or erosion. An increasing number of groups
directly measure CH4 and N2O concentrations from aircraft and
are working on fully coupled landscape models, which do allow
satellites. For example, in the UK aircraft based N2O and CH4
consideration of interactions between the biosphere, hydrosphere
concentrations measurements downwind of the British coast have
and atmosphere at landscape scales. National inventories for N2O
delivered unique measurements of CH4 and N2O at the country
and/or CH4 emissions from soils using DNDC or DayCent have been
scale and provided independent top-down estimates of UK emis-
calculated for US, UK, China, Germany, India or Europe. Even on
sions. Measurements were interpreted by using a simple boundary-
a global scale, the GIS coupled Forest-DNDC model was used to
layer budget approach and the dispersion model NAME. This
estimate N2O emissions from tropical rain forest soils (Werner
approach suggests that the bottom up national emission inventory
et al., 2007). Increasingly biogeochemical models have been used to
underestimates CH4 emissions by a factor of two and N2O emissions
study potential strategies for mitigating GHG emissions from
by a factor of three (Polson, submitted for publication). An under-
soils on site as well as on regional scales (e.g. Li et al., 2006) or to
estimation of the UK national CH4 inventory was also reported by
improve our understanding how future changes in climate or land
Bergamaschi’s et al. (2005) comparison of bottom up and inverse
use may feedback on biosphere–atmosphere exchange of GHG
modelling approaches.
(Parton et al., 2007). Uncertainty in such emission inventories and
Satellite-borne instruments, such as SCIAMACHY, are able to
mitigation/feedback studies is associated with the uncertainties in
provide CH4 concentration measurements at the global scale.
input parameters as well as with the uncertainties in model
SCIAMACHY can clearly detect spatial and temporal variations in
parameters. However, in present studies the uncertainties in input
CH4 concentrations in the boundary layer, a considerable achieve-
parameters have mainly been addressed using Monte Carlo tech-
ment given the small enhancements in a large background signal.
niques (Kesik et al., 2006; Werner et al., 2007), whereas model
Using these methods emissions due to coalfields, rice cultivation,
parametric uncertainty is often neglected (Van Oijen et al., 2005).
ruminants and wetlands are visible for China and India and the Po
valley in Italy (Buchwitz et al., 2005, 2006). Comparisons between
SCIAMACHY observations of CH4 concentrations and those derived
from simple emission inventories revealed large regional and
seasonal differences, especially over tropical rainforests. To some
extent these differences were caused by overestimating CH4
concentrations when water vapour concentrations were large
(Frankenberg et al., 2008). With this correction, SCIAMACHY still
estimates a larger CH4 budget for the tropics than previously esti-
mated and is a clear priority for direct measurements at the surface
using chamber and micrometeorological methods. Validation of the
global N2O budget using satellites is currently not yet possible as
sufficiently accurate and precise N2O satellite data with high
sensitivity near the earth’s surface have not yet been obtained.

8.7. Conclusions

The key developments and gaps in knowledge are:

1. New molecular tools are now available to link soil microbial


biodiversity with soil function and can provide an overview of the
distribution of functional microbial groups in soils of different
landuses, and assign trace gas emissions to the active microbial
Fig. 8.3. N2O emissions from agricultural soils in Europe using the GIS coupled
population.
DNDC model. For further details on databases and methodology see Butterbach-Bahl 2. Instrument development has facilitated CH4 and N2O flux
et al. (2009). measurements at the field and landscape scale and provides
D. Fowler et al. / Atmospheric Environment 43 (2009) 5193–5267 5245

long-term measurements at large spatial scale and high temporal


resolution at key sites.
3. New methods to study denitrification rates to N2 and isotope
studies to elucidate the microbial pathway responsible for N2O
production and removal are being developed, but none of
these methods can currently be used at the field scale or high
frequency temporal scales.
4. There are insufficient data to scale up CH4 and N2O emissions to
the global scale or to include ‘new’ crops, i.e. bioenergy crops.
5. There is a gap in knowledge of the contribution and quantifi-
cation of plants, especially trees, in producing and transporting
N2O, CH4 from soil to atmosphere.
6. Biogeochemical models have been developed and synthesize
our understanding of ecosystem processes and GHG exchange.
They play an important role in identifying and predicting
consequences and feedbacks of global changes (climate and land
use change) for ecosystem functioning and biosphere–atmo-
sphere trace gas exchange. These models also provide mecha-
nistic tools to up-scale emissions to regional and global scales.
7. Inverse modelling, tower and aircraft based boundary layer
budget studies have been developed and now provide appro-
priate tools to challenge and validate bottom up inventories at
the regional and country scale. Fig. 9.1. A schematic diagram of the processes affecting organic species at the air–sea
interface.
9. Exchange of trace gases and aerosols over the oceans

In this section, considering recent developments in surface– being ejected directly with primary aerosol from the sea surface
atmosphere exchange over the oceans, the focus has been narrowed (Fig. 9.1).
to organic trace gases and aerosols, in which there have been major Several important organic emissions from the ocean have
recent advances in understanding. been identified previously. The best known is dimethyl sulphide
(DMS) which is produced biogenically in the ocean (e.g. Keller
9.1. New trace gas interactions at the air–sea interface et al., 1989; Liss et al., 1997, and references therein), and yields
the inorganic aerosol component sulphate upon complete
Considering the size and potential importance of the air–ocean oxidation in the atmosphere (Kiene and Bates, 1990). It is also
interface, it is surprisingly poorly characterized for most organic well established that organohalogens are emitted in various
trace gases. These organic species are known to play important roles forms (e.g. methyl iodide, bromoform, methyl bromide) from
in the Earth’s atmosphere, impacting ozone chemistry and aerosol phytoplankton, bacteria, molluscs and worms (e.g. Gribble,
formation, thereby influencing the Earth’s overall oxidation capacity 1992). Following atmospheric oxidation, these can affect either
and radiative budget (Williams, 2004 and references therein). tropospheric or stratospheric ozone, depending on the lifetime
It should be noted that the net primary production (NPP) of the of the species. However, over the period of the ACCENT project
ocean is comparable in size to that of the terrestrial biosphere (2003–2008), there has come a realisation that the surface ocean
(ca 45 PgC yr1), even though there is approximately 100 times less can play an important role in the budgets of many more organic
biomass in the ocean than on land. The relative paucity of ocean trace gases. For example, the surface ocean has been shown
based data compared to terrestrial sites is due partly to accessibility recently to be a large reservoir for oxygenated organic species
and partly to the high spatial and temporal variation within the e.g. acetone (Singh et al., 2003; Williams et al., 2004). The
limited oceanic biomass. Moreover, there has been a perception from possible influence of oceanic isoprene on marine clouds has also
earlier studies of oceanic alkanes and alkenes that the global ocean been hotly debated (Meskhidze and Nenes, 2006). Finally
is a relatively minor source term. Over the period of the ACCENT a surface ocean source of methanol first speculated in mesocosm
project, this view has changed remarkably and recent studies are studies has been implemented in a global model assessment of
beginning to recognise the profound effects of the ocean–air inter- methanol, thereby generating an improved fit between model
face on global chemical budgets. For many important chemical and measurement data (Millet et al., 2008). Therefore this article
species in the atmosphere the role of the ocean remains the greatest has been focussed on the more recent discoveries related to
uncertainty in the budget. acetone (CH3COCH3), methanol (CH3OH), isoprene (C5H8),
The sunlit regions of the oceans are home to a myriad tiny monoterpenes (C10H16) and alkyl nitrates (RONO2) in order to
plant species and bacteria. These organisms photosynthesise highlight the new developments in air/ocean interactions.
carbon dioxide (CO2) from the atmosphere into biomass, and
a fraction of the carbon ‘‘leaks’’ out into the surrounding seawater 9.1.1. Case studies
in the form of organic compounds. Some small volatile species 9.1.1.1. Acetone (ocean uptake). Over the past 5–6 years our
with low Henry’s Law coefficients are known to escape directly to understanding of the role of the ocean in the global acetone budget
the atmosphere (e.g. dimethyl sulphide, DMS) while larger species has changed remarkably. Acetone is ubiquitous in the troposphere
will remain in the water phase. Subsequent photo-oxidation in and found at relatively high mixing ratios (ca 200 ppt) even in
both air and seawater phases generates a multitude of photo- the remote Pacific atmosphere. Since acetone is recognised as an
chemical breakdown products. These compounds may affect the important precursor for PAN, ozone and HOx, especially in the cold,
hygroscopicity and reflectivity of the marine boundary layer dry, upper troposphere, there has been interest in determining the
aerosol, either by condensing onto existing aerosol surfaces, or by sources and sinks worldwide. In 2002, Jacob et al. published
5246 D. Fowler et al. / Atmospheric Environment 43 (2009) 5193–5267

a global budget of acetone (Jacob et al., 2002) which differed from et al. (2005), the overall the sources and and sinks for acetone were
all previous budget attempts in that it considered the role of the still not balanced. To understand these processes better a new
ocean for the first time. Through inverse modelling, they estimated approach to investigate acetone ocean fluxes was made using
that the ocean was an important net source of acetone. Indeed, so-called ‘‘mesocosms’’ (Sinha et al., 2007). The mesocosms are
from the total global source of 95 Tg, some 25 Tg was estimated to light permeable Teflon tent-like structures which float on the
originate from the ocean in order to balance the known sources and surface ocean enclosing a volume of air near the surface, and with
sinks. This was pioneering work since at that time no seawater walls that extend some 20 m below the surface to restrict the
acetone measurements, or air–sea fluxes were available. However, advection of the water mass below. The airspace in the top of the
in the space of just two years this view changed dramatically. In mesocosm was continually flushed with ambient air to give a resi-
2004 the model developed by Jacob et al., 2002 was tested against dence time of approximately 3 h in contact with the water surface.
measurements over the remote Pacific. It was found that the model By measuring at the inlet and outlet, the flux could be calculated
consistently overpredicted the measured acetone mixing ratios in while phytoplankton in the water column were monitored. In the
the marine boundary layer and the authors concluded that the case of methanol a clear uptake flux (from the air to the ocean) was
ocean was a net global sink for 15 Tg, and that the sources and sinks observed throughout the experiment, whereas for DMS the flux
were not balanced. In 2004, the first open ocean measurements of was always from the ocean to the air. Interestingly, for acetone the
acetone in air and seawater were made (Williams et al., 2004). The flux was found to be variable but systematic. In strong daylight and
interhemispheric gradients and depth profiles shown by Williams in the presence of significant biological activity, acetone was
et al., 2004 were consistent with uptake of acetone from the air to emitted from the ocean to the air. In low light or biologically poor
the sea and a microbial sink in the seawater. regimes, however, acetone was taken up by the water. These results
In 2004, two important new publications emerged concerning are consistent with the results of Marandino et al., 2005 and the
acetone. The first was a laboratory-based study which re-deter- ocean being a net sink for acetone on a global scale, since most of
mined the photolysis quantum yield of acetone as a function of the ocean is oligotrophic. However, biologically active regions (e.g.
temperature and pressure (Blitz et al., 2004). It was found that upwelling zones, ocean fronts, or large natural phytoplankton
the accepted acetone photolysis rates were significantly over- blooms) can be strong sources in daylight and depending on their
estimated by a factor ranging from 3 to 5, particularly for the size could to some extent offset the general sink. It is therefore
cold, low pressure conditions of the upper troposphere. In the important to investigate these biological hotspots in future to
same year, a new shipborne measurement study was published in better constrain the global budget (Fig. 9.2).
which the authors directly measured the flux of acetone at the
ocean surface for the first time using an eddy correlation method 9.1.1.2. Methanol (ocean uptake). In many respects the global
(Marandino et al., 2005). Interestingly, the authors consistently methanol budget is similar to that of acetone discussed above. Plant
found uptake fluxes (from the air to the ocean) for acetone over growth accounts for most of the estimated global source (40–80%)
the oligotrophic Pacific ocean which became stronger further and again the role of the ocean is one of the largest uncertainties in the
from the equator. For comparison Marandino et al., 2005 also budget. Studies of methanol have consistently indicated an ocean
determined the flux of acetone by making separate measure- uptake of methanol (Williams et al., 2004; Lewis et al., 2005; Sinha
ments in the seawater (5 m depth) and air (18 m height). Similar et al., 2007; Carpenter et al., 2004 and references therein.) Recently,
to the results from the Tropical Atlantic (Williams et al., 2004), a global 3-D chemical transport model (GEOS-Chem) was used to
these water and air measurements led to highly variable flux integrate and interpret new aircraft, surface, and oceanic observations
results at the surface, whereas the direct flux measurement was of methanol in terms of the constraints that they place on the
more consistent. This strongly suggested that for acetone, the atmospheric methanol budget (Millet et al., 2008). It was shown that
actual air/ocean flux is being driven by processes in the upper- for methanol, although overall the ocean represents a net sink,
most layer (0–5 m). a separate light dependent oceanic source needs to be introduced in
Although these two new studies (Blitz et al., 2004; Marandino order to correctly simulate regional distributions in the atmosphere.
et al., 2005) had a strong impact on the original budget of Jacob This in-water source has the effect of tempering the uptake flux

Fig. 9.2. MODIS chlorophyll picture of the Southern Atlanic Ocean in January, inset the Research vessel Marion Dufresne.
D. Fowler et al. / Atmospheric Environment 43 (2009) 5193–5267 5247

particularly in the Tropics. It was deduced that that the ocean contains terrestrial sources. Maximum levels of 100–200 pptv total mono-
a large primary source (85 Tg y1) of methanol to the atmosphere and terpenes were encountered when the ship crossed an active phyto-
also a large sink (101 Tg y1), comparable in magnitude to atmo- plankton bloom, whereas over the oligotrophic ocean monoterpenes
spheric oxidation by OH (88 Tg y1). Thus the ocean is a net sink were mostly below detection limit. The monoterpenes/isoprene
overall, but the in-water source term must be included to match with ratio reached 21% in laboratory experiments (the ratio between the
available atmospheric measurements datasets. highest production rates of total monoterpenes and isoprene) and
ranged between 7 and 60% in the Southern Atlantic Ocean.
9.1.1.3. Isoprene (ocean emission). Isoprene, the strongest terrestrial
biogenic emission, has also been observed as an oceanic emission 9.1.1.6. Alkyl nitrates (ocean emission). Alkyl nitrates were assumed
(Bonsang et al., 1992) and in laboratory-based studies of plankton until recently to be exclusively of anthropogenic origin, being emitted
(Shaw et al., 2003 and references therein). It has been suggested directly from combustion or chemical processes (Simpson et al., 2002),
recently that marine isoprene emissions are the cause of cloud or being produced at low yield in the photo-oxidation of organic
droplet radius changes in marine clouds situated directly over compounds in the presence of NOx via the reaction of an organic
phytoplankton blooms (Meskhidze and Nenes, 2006). However, an peroxy radical (RO2) and NO (Roberts, 1990). However, measurements
impact of the isoprene on cloud properties appears unlikely given of MeONO2 and EtONO2 both in equatorial air and seawater (Chuck
that concentrations of isoprene measured over the Southern Oceans et al., 2002) have revealed positive saturation anomalies, and high
do not impact the organic carbon aerosol concentrations signifi- levels of RONO2 which correlate strongly with species of known
cantly (Arnold et al., 2004). In the aforementioned paper an aerosol marine origin such as bromoform (Blake et al., 1999). The mechanism
production efficiency of 2% was assumed for isoprene, and the of formation of marine alkyl nitrates still remains somewhat unclear.
modelled contribution of isoprene to organic carbon (OC) was found Production in seawater through aqueous phase photochemistry (Dahl
to be less than a 1%. Moreover, since time is required to oxidise et al., 2003) has been shown to occur via the reaction of ROO þ NO,
isoprene to nucleating products, a superpositioning of a cloud effect where photolysis of coloured dissolved organic matter (CDOM)
over a bloom in a region of high wind speeds again appears unlikely. generates the peroxy radicals and nitrate (NO 2 ) photolysis generates
Typical mixing ratios of isoprene over phytoplankton rich areas are the NO. Interestingly, the yield of the reaction ROO þ NO in seawater
200–300 ppt, approximately an order of magnitude less than over was found to be significantly higher than in the gas phase. Alterna-
the rain forest (Williams et al., 2001). However, since isoprene reacts tively, alkyl nitrates may be emitted directly from marine biota,
rapidly in air, terrestrial emissions will not impact the open ocean. although direct evidence for enzymatically mediated production has
Marine isoprene emissions could influence the local ozone produc- not yet been found. Using a chemical transport model Neu et al. (2008)
tion efficiency in regions where ship emissions of NOx occur. This found the maximum impact of the oceanic alkyl nitrates to be over the
may be significant for fishing fleets, as the fish, and hence the fleet, Western Pacific, where they were responsible for of increase of up to
follow the isoprene producing phytoplankton. 20% of the ozone column.

9.1.1.4. Halogenated organics (ocean emission and uptake). The


ocean acts as a huge reservoir for chlorine, bromine and iodine and 9.2. Aerosols
volatile organic halogen species (e.g. halocarbons) provide a pathway to
transport halogens from the water phase to the atmosphere. Previous 9.2.1. Primary marine aerosol (PMA) source functions
studies revealed that halocarbons like CH3Cl, CH3Br, CH3I, CHBr3 and Primary marine aerosol (PMA), or sea-spray aerosol is a major
CH2Br2 are emitted from various marine organisms, especially macro- source of global natural aerosol mass budgets and is important for
and microalgae, (Scarratt and Moore, 1999 and references therein). The global climate. Mass is dominated by the super-micron size range
global sources of CH3I, CH2Br2 and CHBr3 are dominated by marine and traditionally source functions have been derived in this size
contributions. Algal emissions of halogenated compounds vary regime. The supermicon size range also contributes significantly to
considerably, not only from species to species, but also as a function of aerosol scattering (Kleefeld et al., 2002) and optical depth (Mulcahy
age, temperature, time of day, nutrition, partial desiccation, grazing, et al., 2008) and thus the direct climate forcing effect. In terms of the
light and tidal movement (Ekdahl et al., 1998). Polybrominated species submicron size range, number concentration rather than mass
(e.g. bromoform) are primarily emitted by macroalgae which occur becomes important in terms of the indirect radiative forcing effect
only in coastal regions, whereas monohalgenated compounds can be through the production of cloud nuclei (O’Dowd et al., 1999). Only
produced from various open ocean biomes. quite recently it has become accepted that submicron sea-spray
aerosol exists, and as a result, submicron source functions are rela-
9.1.1.5. Monoterpenes (ocean emission). Recent laboratory incuba- tively new in terms of development.
tion experiments and shipboard measurements in the Southern The PMA source function describes the flux of sea-spray aerosol,
Atlantic Ocean have provided first evidence for marine production of i.e. the number of droplets produced per unit surface area and per
monoterpenes (Yassaa et al., 2008). Nine marine phytoplankton unit of time, evaluated typically at 10 m above the ocean surface.
monocultures were investigated using a GC–MS equipped with an Hence the function describes an effective flux, parameterised in
enantiomerically selective column and found to emit at rates, terms of ambient parameters such as wind speed and water
expressed as nmol C10H16 (monoterpene). g [Chl_a]1 day1, from temperature. Measurements may provide total fluxes, i.e. the total
0.3 nmol g [chl_a]1 day1 for Skeletonema costatum and Emiliania number of particles in a given size interval, or spectral fluxes.
huxleyi to 225.9 nmol g [chl_a]1 day1 for Dunaliella tertiolecta. Nine The latter are expressed in number of droplets for a range of size
monoterpenes were identified in the sample and not in the control, intervals, i.e. mm1 m2 s1. In this review, we focus on a selection of
namely; ()-/(þ)-pinene, myrcene, (þ)-camphene, ()-sabinene, recently developed or improved source functions which span both
(þ)-3-carene, ()-pinene, ()-limonene and p-ocimene. submicron and super-micron sizes. Particular emphasis is focused
The laboratory measurements are also supported by shipboard on the submicron spray flux and chemical characteristics. A
measurements of monoterpenes in air were made between January comprehensive historical review, focused primarily on sea-salt
and March 2007, while crossing the South Atlantic Ocean, see aerosol production, is provided by Lewis and Schwartz (2005) with
Fig. 9.2. Monoterpenes were detected in air over high ocean chloro- some additions and description of specific source function formu-
phyll regions sufficiently far from land as to exclude influence from lation in O’Dowd and de Leeuw (2007).
5248 D. Fowler et al. / Atmospheric Environment 43 (2009) 5193–5267

It is often assumed that the dependences on droplet size and provides a water temperature dependence that compares favourably
environmental parameters can be separated, i.e. the source func- with independent measurements (e.g. Clarke et al., 2006) for 25  C.
tion is presented as the product of a size-dependent function, g(r),
and a function that describes the parameterisation as function of of 9.2.2. Chemical composition of primary sea spray
environmental parameters, f(a,b,.), where r is the droplet size Although the dominant mass fraction of sea-spray aerosol is sea-
at a specified relative humidity (dry, RH ¼ 80%, or wet) and a, b, . salt, organic matter also contributes to the overall mass and it has
are, e.g. wind speed, water temperature, atmospheric stability, etc. long been known that marine aerosols contain organic material
Scaling arguments show that droplet production varies approxi- (i.e. Blanchard, 1964). Field measurements suggested a significant
mately with the third power of the wind speed. However, other biogenic primary source of marine organic components (O’Dowd
types of paramterization have been proposed as well. Selected et al., 2004; Cavalli et al., 2004). In particular a dominant water-
source functions are shown in Fig. 9.3 for a wind speed of 8 m s1. insoluble organic fraction in fine marine aerosol collected during
There are two main developments to report on: the first in terms of periods of phytoplankton bloom in the North Atlantic was observed
the super-micron sizes where the data in Fig. 9.3 show that the and it was hypothesized that these insoluble organic components
discrepancy between different formulations is much reduced with could have a mainly primary origin. Similar results supporting
respect to the review situation reported by Andreas (2002). a biologically driven oceanic OC source have been recently reported
With respect to Lewis and Schwartz (2004), the uncertainty has by Spracklen et al. (2008).
been reduced by a factor of 2. For small particles a clear size The most comprehensive study to date on the organic fraction
dependence emerges varying roughly as r1.5 80 . The source functions of sea-spray aerosol has been conducted by O’Dowd et al. (2004).
shown in Fig. 9.3 were obtained using different methods and They found a significant and dominating fraction of organic matter
different physical principles but leading to consistent results. in submicron sizes, while the super-micron size range was
The second main development is the extension of the source predominately inorganic sea-salt. It should be noted, however,
function well into the submicron size range. The Mårtensson et al. that the absolute magnitudes of organic mass in the sub and
(2003) laboratory-based study extended the size-resolved source super-micron size ranges were equivalent (with one-third of the
function down to r80 ¼ 20 nm and found that the production as total organic mass residing in the coarse mode), and that it was
a function of size was also dependent on temperature. Clarke et al. their relative concentrations that differed significantly. Fig. 9.4
(2006) provide a source function for particles down to 10 nm. These illustrates the chemical composition of clean marine aerosol over
studies combine experiments in the laboratory or over the surf the north east Atlantic during winter and summer periods,
zone, to determine the spectral shape of the flux, with whitecap corresponding to low and high biological activity periods (O’Dowd
coverage which in turn is paramterized as function of wind speed. et al., 2004). Also shown is the distribution of chlorophyll-
Direct and in situ measurements of sea spray total number fluxes a derived from the SeaWifs satellite. During periods of high
(D > 10 nm) are provided by the eddy covariance (EC) method that biological activity, the organic fraction ranged from 40 to 60% of
was first attempted by Nilsson et al. (2001) in the Arctic Ocean. The the submicron mass, while during low biological activity periods,
advantage of this method, as opposed to the whitecap method, is the fraction reduced to about 10–15%. O’Dowd et al. (2004) argued
that all particles within the detectable size range may be measured, that the water-insoluble organic fraction, dominating the organic
and hence there is no restriction to bubble-mediated production. composition in the fine size fraction, was likely to be derived from
The technique was also used at a coastal station over the North East bubble-mediated production.
Atlantic by Geever et al. (2005), who quantified total number Later experiments, using the gradient technique to determine
concentration over two size ranges covering the Aitken mode (10– aerosol chemical fluxes at Mace Head (Ceburnis et al., 2008)
100 nm) and the Accumulation mode (100–500 nm). showed that the water-insoluble organic carbon (WIOC) mass
Overall, the most recent schemes agree quite well (e.g. Clarke invariably had an upward mass flux associated with it and followed
et al. surf zone study compares very well to the Mårtensson et al. similar trends to sea-salt gradients, while water-soluble organic
laboratory-based parameterisation), providing an improved level of carbon (WSOC) mass possessed a downward flux profile identical
confidence in PMA source functions over sizes from 0.01 mm to to nss-suphate. They concluded from the gradients that WSOC must
w10 mm. In addition, the Mårtensson et al. (2003) parameterisation be formed from secondary aerosol formation processes while WIOC
must be formed from primary production. These conclusions were
108 further supported by Facchini et al. (2008) who conducted bubble-
bursting experiments amidst a plankton bloom over the NE Atlantic
107 during the MAP (Marine Aerosol Production) cruise in 2006
(Fig. 9.5). During these experiments, it was found that spray
106 particles exhibited a progressive increase in the organic matter
content from 3  0.4% up to 77  5% with decreasing particle
-1
dF/dLog (r) m s

105 diameter from 8 to 0.125 microns (Fig. 9.5).


-2

Submicron organic matter was almost entirely water insoluble


104
Monahan et al 1986 (WIOM) and consisted of colloids and aggregates exuded by
Monahan Extrapol phytoplankton. Facchini et al. (2008) found that the WIOC to sea-
103 Martensson et al., JGR, 2003 salt mass ratio fingerprint as a function of particle size in the bubble
Vignati et al., JGR, 2001
102 Gong, JGR, 2003 tank experiments matched that observed in atmospheric samples
Clarke et al., JGR, 2006 both at Mace Head (shown in Fig. 9.5) and on the MAP cruise over
de Leeuw et al., JGR, 2000
101 the open ocean. These results conclusively confirmed that the
de Leeuw et al., AMS, 2003
Reid et al., JGR, 2001, WIOC component observed in marine air samples relate to primary
100 aerosol production. Electron microscopy observations of individual
10-2 10-1 100 101 102
particles collected at the ocean surface in a number of sites sup-
Radius (microns)
port the hypothesis that complex exopolimers and the microgels
Fig. 9.3. Compilation of sea-spray source functions. Flux values are for a wind speed of forming from these, produced by bacteria and algae, are involved in
8 m s1. bubble-bursting processes (Bigg and Leck, 2008).
D. Fowler et al. / Atmospheric Environment 43 (2009) 5193–5267 5249

Fig. 9.4. (Left) Chemical and mass size distributions for North Atlantic marine aerosol during periods of low biological activity and high biological activity. (Right) oceanic chlo-
rophyll-a concentrations over the North Atlantic for low and high biological activity periods.

These results indicate that a sea-spray source function should et al. (2005) accumulation mode number flux, combined with
not only consider size-resolved mass, but also chemical composi- the Yoon et al. (2007) seasonal modal diameter (minus secondary
tion. The first attempt at a combined submicron organic-inorganic aerosol mass), and integrated with the seasonal trend in WIOM/sea-
sea-spray source function, implemented in a regional climate mode, salt ratios to produce a physico-chemical flux function driven by
was produced by O’Dowd et al. (2008). They combined the Geever wind speed and satellite-derived chlorophyll-a concentrations over

Fig. 9.5. (Left) Average chemical composition relative concentration from bubble-bursting tank samples. (Bottom) average ratio of WIOC to sea-salt from atmospheric samples at
Mace Head. Data for the 0.06–0.125 mm size range are not reported from the bubble tank because the total carbon analyses in this stage were below detection limit. The bars are the
standard deviation of the mean. (Right) Visible satellite image of plankton bloom off the west coast of Ireland during the MAP cruise June–July 2006. Bubble-bursting tank
experiments were conducted in and around the plankton bloom.
5250 D. Fowler et al. / Atmospheric Environment 43 (2009) 5193–5267

Fig. 9.6. (Top panel) Near surface sea-spray mass concentrations around the European regions and average wind vectors. (Bottom Panel) Percentage primary organic contribution to
sea-spray mass.

the North East Atlantic. The model predicted results compare well (MSA). Some dicarboxylic acids were associated with secondary
to seasonal observations at Mace Head and are illustrated for winter formation mechanisms in previous papers (i.e. Kawamura and
and summer seasons in Fig. 9.6. Sakaguchi, 1999) but a relevant fraction of the observed concen-
trations of oxidized organic matter in marine aerosol still remains
9.2.3. Secondary aerosol production unaccounted. Modelling studies by Meskhidze and Nenes (2006),
In recent years, significant effort has been made into the study of proposed that isoprene emissions from plankton were sufficient
new particle formation in the coastal zone in the hope that it would to produce enough water-soluble organic aerosol to significantly
elucidate key processes associated with nucleation over the open enhance CCN concentrations and cloud albedo; however, it was
ocean. Most of these studies have focused on nucleation in coastal later revealed that the isoprene fluxes were inadvertently over-
zones (e.g. at Mace Head) and revealed regular particle bursts, with estimated by a factor of 100. Nevertheless, Zorn et al. (2008) also
burst concentrations often exceeding 106 cm3. These events have confirmed the dominance of organic aerosol mass in air overlying
been linked to release of biogenic iodine vapours from coastal algae plankton blooms over the southern ocean but no detail on speci-
followed by the photochemical production of iodine oxide aerosols ation was elucidated. Recently a new secondary organic aerosol
(O’Dowd et al., 2002; McFiggans et al., 2004). A detailed review of component, produced through the reaction of gaseous amines with
studies into these processes is found in O’Dowd and Hoffmann sulphuric acid has been found in marine aerosol over the North
(2005). Further studies revealed that the nucleation mode particles Atlantic (Facchini et al., 2008). Dimethyl and diethyl ammonium
could also contain some organic aerosol mass suggesting that salts (DMAþ and DEAþ) are the most abundant organic species,
secondary organic aerosol production also occurs in marine air second only to MSA, detected in fine marine particles in North
and contributes to aerosol growth (Vaattovaara et al., 2006). The Atlantic and represent on average 11% of SOA and a dominant part
findings of O’Dowd et al. (2004) and Ceburnis et al. (2008) also (35% on average) of the aerosol water-soluble organic nitrogen
point to significant contributions of secondary organic aerosol, (WSON). Several evidences support the hypothesis that DMAþ and
manifesting itself in the WSOC component. The most relevant DEAþ have a biogenic oceanic source even if the formation mech-
organic secondary component (SOA) is methanesulphonic acid anism of these biogenic amines remains unclear.
D. Fowler et al. / Atmospheric Environment 43 (2009) 5193–5267 5251

In conclusion, apart from MSA and a few dicarboxylic acids and which can involve atmospheric organic compounds and oxidants
amine salts, the vast majority of secondary organic marine aerosol (airborne microorganisms) (Ariya and Amyot, 2004; Sun and Ariya,
remains to be identified, suggesting that other formation mecha- 2006). They also comprise either biological particles including alive,
nisms and alternative SOA components should be studied. dead cells and cell fragments, capable of nucleating cloud droplets and
ice particles via physical processes (Möhler et al., 2007) or any kind of
10. The processes of wet scavenging of aerosols organic substances deriving from biomolecules and contributing to
and trace gases from the atmosphere aerosol masses. Airborne microorganisms are incorporated into cloud
droplets and raindrops by nucleation scavenging as they have CCN or
10.1. Introduction IN potential (e.g. Bauer et al., 2003; Möhler et al., 2007) or by washout
processes. Some investigations clearly show that most of these
Precipitation or wet scavenging is an efficient cleaning mecha- microorganisms are able to develop at low temperatures (between 5
nism of the atmosphere. It combines all the in and below cloud and 5  C) encountered in clouds. Furthermore, measurements of
processes that take up trace gases and particles into liquid drops or concentrations of adenosine triphosphate (ATP) in cloud water indi-
crystals forming a cloud and deposits the material to terrestrial or cate that most microorganisms are still metabolically active (Amato
marine surfaces in rain or snow. et al., 2007a,b,c,d).
Despite the presence of ice in many cloud systems, interactions
10.2. Nucleation scavenging of drops and ice crystals between trace chemicals and ice are not well understood (Abbatt,
2003). Chemical solutes originally dissolved in a supercooled
Droplets form on a subset of the aerosol particles, the cloud drop may be retained or expelled from the drop as it freezes. Non-
condensation nuclei (CCN), present in every air mass. This mech- volatile species, such as sulphate, are efficiently retained during
anism is probably the most important to incorporate pollutants freezing but this retention process is not well characterized for
into the cloud phase. Depending on their size, chemical composi- many soluble gases found in clouds (Voisin et al., 2000). Cloud
tion and the ambient relative humidity, aerosol particles take up modelling studies have found that partitioning of solutes during
a certain amount of water (Pruppacher and Klett, 1997) and when hydrometeor freezing may significantly affect chemical distribu-
exceeding their critical size they activate to cloud droplets. In tions in the troposphere and deposition to the ground (Audiffren
the classical Köhler theory, only their composition with respect to et al., 1999; Mari et al., 2000; Crutzen and Lawrence, 2000; Yin
insoluble material and inorganic salts is considered. Recent studies et al., 2005; Kärcher and Basko, 2004). A better understanding of
(e.g. Anttila and Kerminen, 2002; Sorjamaa et al., 2004; Romak- the partitioning of volatile chemical solutes during freezing is
kaniemi et al., 2005; Kokkola et al., 2006; Topping et al., 2007) have needed to quantify their effects on tropospheric gas-phase and
highlighted the importance of soluble trace gases and partly soluble precipitation chemistry.
organic substances which often coat the surface of the particles for Bacteria which have entered the liquid phase find therein
the activation properties. a solution of organic compounds which may serve as nutrients.
Even though our knowledge of the formation of droplets is now Recent studies show that living and active microorganisms,
reasonably satisfactory, the nucleation of ice crystals as a subject is still including bacteria, yeasts and fungi, are present in the atmospheric
quite poorly understood. In the atmosphere, significant numbers of ice water phase (Fuzzi et al., 1997; Bauer et al., 2003; Amato et al., 2005,
particles start to form only below 5  C co-existing still with liquid 2007a). These microorganisms could play an active role in chemistry
drops. Homogeneous freezing of liquid droplets depends on the size; and microphysics of clouds as discussed by a growing number of
large droplets can freeze homogeneously at temperatures of around scientists (Ariya and Amyot, 2004; Amato et al., 2005, 2007b; Möhler
33  C, whereas at 40  C even the smallest droplets freeze homo- et al., 2007; Deguillaume et al., 2008). Indeed, living microorganisms
geneously. New insight into the homogeneous nucleation of ice crystals are clearly biocatalysts which could transform organic compounds
under these conditions which correspond to cirrus clouds was as an alternative route to photochemistry. Many unresolved ques-
obtained in the AIDA chamber (Benz et al., 2005; Möhler et al., 2006). In tions remain on this topic and long-term observations can be used
the temperature range between 5 and 40  C, the presence of to evaluate the diurnal and seasonal variations of structure and
insoluble nuclei is necessary to initiate the formation of an ice crystal. activity of microorganisms as a function of environmental conditions
These ice nuclei (IN) are aerosol particles that can act in four main ways: (i.e. humidity, light, temperature, pH.).

– Deposition mode: water is adsorbed directly from the vapour 10.3. Impaction scavenging of aerosol particles
phase onto the surface of an IN where it is transformed into ice
– Condensation–freezing mode: this is a hybrid process that Inside the cloud unactivated aerosol particles remain between
requires supersaturation with respect to water. Here, the CCN the nucleated drops as interstitial aerosol. These particles can
that has formed the drop acts now as an IN. This process seems collide with the hydrometeors and become incorporated into the
far more effective than the deposition mode. cloud particles. However, due to the fact that already the main part
– Freezing mode: the IN, scavenged by the drop, initiates the ice of the particle mass was scavenged by nucleation, inside cloud
phase from within a supercooled water droplet this process does not contribute significantly to the pollution mass
– Contact mode: the IN initiates the ice phase at the moment of in precipitation (Flossmann, 1998a,b; Flossmann and Wobrock,
contact with the supercooled drop 1996). An importance can be attributed to this process in combi-
nation with the contact mode freezing of the previous section.
The number of IN depends on the chemical properties of the Once the hydrometeors fall and leave cloud base, on their way to
aerosol particles. It has been found that there exists a dependency the earth’s surface they meet an unperturbed aerosol particles
on supersaturation and also on temperature. In contrast to CCN, population. Here, the collision with aerosol particles can contribute
a good IN should be insoluble and have a crystalline-type structure a significant portion to the aerosol particle loading of the rain on
to facilitate the formation of the ice lattice (e.g. silicate). the ground, depending also on the height of cloud base.
Recently, the role of primary biological aerosols for nucleation of During the cloud lifetime, chemical processes lead to the
drops and ice crystals has been highlighted (Deguillaume et al., 2008). formation of new chemical species with relatively low volatility
These particles can be viable organisms capable of metabolic reactions such as inorganic and organic acids, which can modify the physico-
5252 D. Fowler et al. / Atmospheric Environment 43 (2009) 5193–5267

chemical properties of aerosol particles after the cloud dissipates


(Feingold and Kreidenweis, 2002; Yin et al., 2005) and lead to
secondary organic aerosols formation. For some chemical species,
aerosol particle dissolution is the only source in cloud droplets; for
instance, transition metal ions, and in particular iron, which is well
known to play a major role in the oxidizing capacity of clouds
(Deguillaume et al., 2005). Study of such complex interactions
needs process modelling efforts integrating in situ measurements.

10.4. Scavenging of gases

In addition to the particles, numerous trace gases are present


in the atmosphere. Gases are taken up into drops according to
their solubility. The maximum amount of a gas that can be taken
up into water is a function of the Henry’s law coefficient. A Fig. 10.1. Schematic of the microphysical and scavenging processes in liquid phase
comprehensive compilation of updated Henry’s law coefficients is clouds.
available at http://www2.mpch-mainz.mpg.de/~sander/res/henry.
html. Henry’s law describes the equilibrium between the then transport them over larger distances, processing the material
concentration in the air and the liquid, however, once in the liquid during transport.
phase most gases are destroyed by chemical reactions and, thus, The problem of correctly describing this process is coupled to the
an equilibrium will never be achieved. Consequently, more and problem of scales. As shown below, the nucleation of hydrometeors
more gas can be taken up into the cloud drops. Only the droplet and all subsequent reactions take place on the scale of the individual
lifetime (max. 30 min) will limit the gas scavenging. Recently, our drop or ice crystal. The formation, transport and dissipation mech-
knowledge of the uptake and reaction coefficients of the ambient anisms of clouds, however, act over a much larger region and require
trace gases has significantly increased and quite complex aqueous description on a synoptic or even hemispheric scale. In the past,
phase reaction schemes have become available (Herrmann et al., this fact imposed severe constraints in the accuracy of the modelling
2005), including an extended reaction mechanism for atmo- of these processes and resulted either in highly parameterised
spherically important hydrocarbons containing more than two dynamical models with detailed treatments of the chemistry (Wolke
and up to six carbon atoms. et al., 2005; Sander et al., 2005) or in highly simplified chemical
The complexities of the cloud processes involved in pollutant schemes in sophisticated meteorological models (Mari et al., 2000).
scavenging have discouraged investigators from simultaneously Only recently have 3-D dynamic codes with detailed microphysical
treating all aspects of multiphase chemistry and microphysics with treatment and aerosol particles (Leroy et al., 2008) and chemistry
equal rigor. However, efforts made to develop sophisticated cloud (Tost et al., 2007) become available due to developments in
models with complex multiphase chemistry allow more detailed computers. These models obviously are restricted to rather limited
studies on the interaction between microphysical and chemical modelling domains, however, they highlight e.g. the importance of
multiphase processes (Leriche et al., 2007; Ervens et al., 2004a,b). the background aerosol population for the development of the cloud
One important feature lies in a detailed representation of the (Leroy et al., 2008).
microphysical as well as multiphase chemical processes. These Fig. 10.2 displays the results of a sensitivity test for the CRYSTAL-
developments really distance themselves from the other attempts FACE cloud (Leroy et al., 2008). The simulation shows a clean
of coupling multiphase chemistry in 3D models, which are often boundary layer in which rain develops readily while precipitation
restricted to the study of inorganic species and basic organic species formation is suppressed in the polluted case by the large population
wet deposition (Tost et al., 2007). of aerosols derived from air pollutants. Not only has the precipita-
tion been suppressed but the horizontal and vertical structure of the
10.5. Clouds cloud is substantially modified.
By including as many as possible of these processes in larger
Clouds form when air ascends and following expansion and scale models parameterisations have been developed to yield the
cooling the water vapour condenses. The droplets grow further by first reliable maps on critical loads (e.g. Hoose et al., 2008; Pozzoli
condensation, then, collide and coalesce with each other, until they et al., 2008). One emphasis of these models has been the aspect of
become sufficiently heavy to fall against the updraft velocity that topography.
has suspended them until now. Depending on the height of cloud
base and the temperature conditions they might reach the ground 10.6. Orographic precipitation
as rain. If the temperature in the clouds reaches temperatures
sufficiently below zero, then ice crystals develop. They also grow by At mid-latitudes, mountainous terrain is commonly associated
water vapour deposition, and collide with each other. If they with high annual precipitation due to the forced ascent of air
become sufficiently heavy, they fall to the ground in solid or liquid resulting in cloud formation and precipitation. At a sub-grid scale,
form, as a function of the below-cloud temperature. During their however, there can be significant variations in pollutant deposition
entire lifetime, these cloud hydrometeors (¼drops or ice particles) due to local emissions and variation in topography and vegetation.
take up pollution in particulate and gaseous form and deposit it on A need, therefore, arises for fine scale process models to investigate
the ground together with precipitation. A schematic display for pollutant deposition at the kilometer scale (Dore et al., 2006).
liquid clouds is shown in Fig. 10.1. The aerosol population (size distribution and composition) has
Only few clouds form locally due to convection and, thus, have a major influence on the dynamics and microphysics of orographic
only a limited geographical impact. Most clouds are embedded in cloud development. The cloud condensation nuclei (CCN) population
large-scale system and cover areas of several thousand km2. They entering cloud base determines the extent and onset of warm rain
incorporate the local pollution when the droplets nucleate and produced by collision coalescence. These, along with the presence of
D. Fowler et al. / Atmospheric Environment 43 (2009) 5193–5267 5253

Fig. 10.2. Sensitivity study concerning the number concentration of boundary layer aerosol particles (Leroy et al., 2008) after 40 min of cloud development; the displayed domain is
a 2-D cross section of the 3-D domain restricted to 30 km in the horizontal and 15 km in the vertical; the envelopes of the different hydrometeors are specified for each panel.

heterogeneous ice nuclei affect the onset of the glaciation process and chemistry have highlighted our lack of knowledge of the organic
the efficiency of secondary ice processes such as the Hallett–Mossop nitrogen constituents (both gaseous and particulate) in the
process of ice splintering. These in turn determine the release of latent atmosphere. Recent reviews (Cornell et al., 2003; Neff et al., 2002)
heat of fusion in the cloud, which has a major influence on the vigour have indicated that the contribution of water soluble organic
and structure of the cloud dynamics. The initiation and development nitrogen (WSON) in precipitation to wet deposition may be up to
of the ice phase is crucial to the precipitation formation and its loca- one-third of the total, yet little is known about the chemical
tion within the cloud. More detailed process studies are needed to composition, form or sources of this material. Initial scepticism
understand such complex feedbacks. In the case of orographic clouds, about the nature of WSON has to some extent been dispelled
it is shown that aerosol–cloud interactions may cause a displacement (Cape et al., 2001), but the broad range of possible composition
of precipitation from the upslope side of a hill towards the downslope and emission sources means that the transfer pathways are still
side when the number of aerosols is increased (Muhlbauer and Loh- somewhat uncertain. It is known, for example, that biological
mann, 2008). Inverse relations between air pollution and orographic processes interconvert inorganic and organic nitrogen in forest
precipitation could be of major interest for weather prediction and canopies (Fang et al., 2008), but it is not clear how much biological
hydrological budget evaluation. activity may occur in the atmosphere or on the surfaces of
The initial physical and chemical state of aerosol entering the sampling equipment. The presence of both gaseous and particulate
clouds is strongly influenced by the prevailing oxidant climate and WSON in the atmosphere implies that dry deposition is an
air mass history. The degree of in-particle oxidation and resultant important but unquantified pathway for transfer of organic
hygroscopic properties has been seen previously to be closely tied to nitrogen to the earth’s surface.
the degree of gas phase photochemical ageing (Cubison et al., 2006).
The specificity, in terms of time since surface emission and oxidative
exposure, which can be made from direct aerosol measurements, is 10.8. Conclusions and some priority areas of future research
however relatively poor. This may be inferred much more accurately
however by making coincident measurements of gas phase volatile The processing of atmospheric contaminants as gases and
organic tracers (VOCs). particles by clouds and precipitation form part of the biogeo-
chemical cycling within the atmosphere. The gases and particles
10.7. Organic N in air and rain are taken up into cloud hydrometeors during the lifetime of a cloud,
processed and either released during evaporation or deposited onto
Better understanding of the processes that link the chemical the ground with the liquid or solid precipitation. Knowledge of the
and biological properties of aerosols with cloud formation and underlying processes has greatly advanced concerning the liquid
droplet growth has indicated a need for better knowledge of the phase. Here, the gap concerning the role of the organic material is
organic components. Although transport or organic C, and depo- now almost closed. The greatest uncertainty nowadays lies with the
sition to the earth’s surface, has not been regarded as quantita- ice phase. Generally, the ice phase is chemically less active than
tively important for ecosystem health, organic N has the potential the liquid phase. Thus, the uptake and processing is reduced.
to add to the known effects of inorganic N wet deposited from the Furthermore, the role of the ice phase in the precipitation forma-
atmosphere especially in remote areas. Studies of precipitation tion and deposition is not completely known and these weaknesses
5254 D. Fowler et al. / Atmospheric Environment 43 (2009) 5193–5267

in understanding give rise to large uncertainty in values are interactions of the different components are not yet studied mainly
obtained by current models. because of instrument and resource limitations. In the 1950s the
The second challenge in the modelling of wet deposition is work started with single component fluxes to a few ecosystems.
linked to role that bacteria and other living organic matter (bio- Currently, as presented in this section, the state of knowledge is
aerosols) can play in the microphysics of a cloud and in atmospheric grouped into families of components (VOC, reactive nitrogen, GHG,
aqueous phase chemistry, which is largely unknown. particles), in which our understanding provides regional budgets
as well as deposition fluxes of chemical species of interest for
11. Ecosystem–atmosphere exchange – concluding remarks effects assessment and forms a part of continental scale integrated
assessment. Clearly, progress has been rapid in some areas, moti-
This paper has reviewed the state of knowledge of atmosphere– vated by pressure for control measures. The next steps in the wider
surface interactions of a broad range of trace gases and particles. integration will require similar pressure to deliver the necessary
Given the wide range of chemical species reviewed and the information to support policy development.
reasonably self contained sections within the paper, this concluding Within the area of reactive nitrogen it has been shown that there is
section does not attempt to provide a summary of the sections. The a dynamic exchange between the atmosphere and surface, regulated
following commentary reflects on the overall direction of the by stomatal and chemical interactions; deposition, re-emission and
science which, like the subject material, is becoming ever more re-deposition processes and by the exchange of different forms of
global and searching for integrating mechanisms. There has been nitrogen in interaction with the status of the system (saturation,
substantial progress over the last decade in process understanding, carbon, phosphorus, water-filled pore space, etc) (Sutton et al., 2008a;
field measurement and in modelling. Models have been developed Pilegaard et al., 1995). Sulphur has a large impact on the uptake and
incorporating the process understanding to generalize the release of ammonia at the surface. This is all is part of the nitrogen
ecosystem–atmosphere exchange over regions and are currently cascade, where one molecule of reactive nitrogen that enters a system
able to describe the fluxes with uncertainties of the order of 30% in in oxidized or reduced state is used and transformed in that system,
wet deposition and 50% in dry deposition for the main chemical can be leached to the groundwater as nitrate, entering the river and in
species. However, there is a lack of measurements to evaluate the the estuaries where it can be emitted as N2O contributing to climate
models. For the future, extensive measuring campaigns and long- change and into the stratosphere, where eventually it is broken down
term monitoring of fluxes are necessary to further develop this depleting the ozone layer (Galloway et al., 2003). There is evidence
important field. The development of super sites in the European that increased nitrogen deposition leads to NO emissions from the soil.
EMEP network with a full spectrum of gas and aerosol phase trace This links nitrogen deposition with the oxidant surface–atmosphere
atmospheric constituents and continuous measurements of exchange. Oxidants in their turn affect the ecosystem health and
surface–atmospheric fluxes represents an important development therewith the nitrogen uptake and use efficiency. These are examples
in this direction. The development of long-term micrometeorolog- of the strong interaction between the components in the surface–
ical CO2 flux monitoring sites, initially within CarboEurope and atmosphere exchange.
followed by AmeriFlux, but now extending to a global network
provides much of the infrastructure for extension to trace gases. 11.3. Future developments
Such long-term flux measurements of reactive pollutants to test,
develop and validate models represent an important development, Agriculture is a major source of emissions to the atmosphere,
which, in turn will be expanded regionally. which relative to industry has been regulated substantially less (Aneja
et al., 2008). Until now the policy requirements for food security from
11.1. Policy needs agriculture has moderated the willingness to limit emissions of trace
gases from this sector. More and more, however, it is recognised that
In the policy development there is a need to address environ- the production should be within the limits of sustainability, limiting
mental priority issues, among which none are currently greater than pollutant emissions to surface or groundwater and to the atmosphere
climate change. However, human health, ecosystem quality and the of reactive nitrogen compounds, greenhouse gases, persistent organic
sustainable use of natural resources is growing in importance. pollutants, phosphorus and odour. Agriculture is a collection of diffuse
There is a current tendency to group the environmental impacts sources with many uncertainties in the emissions. Greenhouse gas
because they are linked through the pollutants and the receptor emissions and ammonia are uncertain because the emissions depend
which follows historical tradition. New, much broader and more on farm management practices, soil type, fertilizer use, type of crop
integrated directions include global change, reactive nitrogen, or animal breeding, size and location of the farm, etc. This makes
air quality and climate change, bringing together a range of related quantification of the sources and successful targeted measures and
issues in the search for more sustainable solutions to the under- policies for control very difficult.
lying problems. This is however far from being implemented in The next step will include quantifying the strong interactions
policies because it involves a larger scale and therewith global between land use, food supply, biodiversity and biogeochemical
political developments. However, the process has begun within cycling of key nutrients. Studying these landscape interactions is
initiatives such as those described in the Millenium Ecosystem a new topic, where just a few exercises have been attempted. The
Assessment (2005). Ultimately, political developments will be landscape scale, such as agricultural areas, complex terrain, urban
required to allow such integrated environmental policies to be areas, etc. have their own dynamic and interactions which differ
developed, but evidence of the time it is taking to develop appro- from the normally studied stationary conditions. It is necessary to
priate controls to limit greenhouse gas emissions suggest that study this scale because of the need to determine the contribution
it will be a slow process. of the individual sources which need to be controlled. In rural,
agriculture dominated areas there is a large contribution of
11.2. Current understanding different reactive nitrogen sources, such as animals in- or outside
housing systems, application of fertilizers or manure, storage of
It is questionable whether our current understanding of the manure, traffic and small industries. Within such an area deposi-
atmosphere–surface exchange is keeping pace with these more tion and re-emission takes place with a high spatial and temporal
integrated needs. The processes are very variable and the dynamic. Especially for nitrogen the efficiency of use can be
D. Fowler et al. / Atmospheric Environment 43 (2009) 5193–5267 5255

improved if the individual contribution of the different sources and Dôme: major groups and growth abilities at low temperatures. FEMS Micro-
biology Ecology 59 (2), 242–254.
resulting losses can be quantified.
Amato, P., Hennebelle, R., Magand, O., Sancelme, M., Delort, A.-M., Barbante, C.,
The majority of surface–atmosphere exchange measurements Boutron, C., Ferrari, C., 2007b. Bacterial characterization of the snow cover at
have been made in rural landscapes, yet a large fraction of emissions Spitzberg, Svalbard. FEMS Microbiology Ecology 59 (2), 255–264.
occur in urban areas, where most of the global human population Amato, P., Demeer, F., Melaouhi, A., Martin-Biesse, A.-S., Sancelme, M., Laj, P.,
Delort, A.-M., 2007c. A fate for organic acids, formaldehyde and methanol in
now resides. In urban areas there is a concentration of sources cloud water: their biotransformation by micro-organisms. Atmospheric
of gases and pollutants from industry, traffic and households. The Chemistry and Physics 7, 4159–4169.
emissions to the atmosphere and the resulting exposure in these Amato, P., Parazols, M., Sancelme, M., Mailhot, G., Laj, P., Delort, A.-M., 2007d. An
important oceanic source of micro-organisms for cloud water at the puy de
areas are therefore a growing concern globally and our knowledge D̂ome (France). Atmospheric Environment 41, 8253–8263.
base for many of the processes in these areas is very limited. This Ammann, C., Brunner, A., Spirig, C., Neftel, A., 2006. Technical note: water vapour
paper reports some of the first measurements of fluxes of aerosols concentration and flux measurements with PTR-MS. Atmospheric Chemistry
and Physics 6, 4643–4651.
and trace gases by micrometeorological methods over urban areas. Andreas, E.L., 2002. A review of the sea spray generation function for the open
The limitations until recently have been partly instrument avail- ocean. In: Perrie, W.A. (Ed.), Atmosphere–Ocean Interactions, vol. 1. WIT Press,
ability, but there has also been a reluctance to work in urban areas Southampton, UK, pp. 1–46.
Aneja, V.P., Schlesinger, W.H., Erisman, J.W., 2008. Farming pollution. Nature
simply due to the complexity of the surface and inertia within Geoscience 1, 409–411.
the measurement community. It is necessary to understand the Anttila, T., Kerminen, V.-M., 2002. Influence of organic compounds on the cloud
dynamics in emissions and deposition of primary and secondary droplet activation: a model investigation considering the volatility, water solu-
bility, and surface activity of organic matter. Journal of Geophysical Research D
particles. Trees or treated surfaces might act as depositing surfaces
107 (D22), 4662. doi:10.1029/2001JD001482.
improving air quality in cities (McDonald et al., 2007) or decreasing Ariya, P.A., Amyot, M., 2004. New directions: the role of bioaerosols in atmospheric
the net ammonia emissions around farms. In many cases the chemistry and physics. Atmospheric Environment 38, 1231–1232.
scientific challenge is clear and the measurement and modelling Arneth, A., Monson, R., Schurgers, G., Niinemets, Ü., Palmer, P.I. Why are estimates of
global isoprene emissions so similar (and why is this not so for monoterpenes)?
needs are well within the capability of the community. There are Atmospheric Chemistry and Physics, in press.
therefore many opportunities to advance the science which will find Arnold, S.R., Chipperfield, M.P., Blitz, M.A., Heard, D.E., Pilling, M.J., 2004. Photodisso-
rapid application in the policy arena. ciation of acetone: atmospheric implications of temperature-dependent quantum
yields. Geophysical Research Letters 31, L07110. doi:10.1029/2003GL019099.
The emerging subject of bioaerosols has been very briefly Ashmore, M., Fuhrer, J., 2000. Use and abuse of the AOT40 concept. Atmospheric
mentioned in this review, and this represents a developing field Environment 34, 1157–1158.
with application in cloud physics, air quality and human health as Ashmore, M.R., 2005. Assessing the future global impacts of ozone vegetation. Plant
Cell and Environment 28, 949–964.
well as the advance of basic understanding. A diverse range of Asman, W.A.H., Sutton, M.A., Schjoerring, J.K., 1998. Ammonia: emission, atmo-
natural and anthropogenic particles are capable of initiating the ice spheric transport and deposition. New Phytologist 139, 27–48.
phase, but some of the most active naturally occurring ice nuclei (IN) Audiffren, N., Cautenet, S., Chaumerliac, N., 1999. A modeling study of the influence of
ice scavenging on the chemical composition of liquid-phase precipitation of
are biological in origin and have the capacity to catalyze freezing at a cumulonimbus cloud. Journal of Applied Meteorology 38 (8), 1148–1160 (AMS).
temperatures near to 2  C. Based on the ubiquitous distribution of Bach, L.H., Frostegård, Å., Ohlson, M., 2008. Variation in soil microbial communities
biological IN in snow and rain from global locations, they are likely across a boreal spruce forest landscape. Canadian Journal of Forest Research 38,
1504–1516.
to encounter the appropriate conditions to affect atmospheric
Baker, B., Guenther, A., Greenberg, J., Goldstein, A., Fall, R., 1999. Canopy fluxes of
processes leading to precipitation. The subject of organic aerosols 2-methyl-3-buten-2-ol over a ponderosa pine forest by relaxed eddy accumu-
and cloud formation has been recently reviewed by Sun and Ariya lation: field data and model comparison. Journal of Geophysical Research 104,
(2006). The cloud nucleation processes are currently among the 26107–26114.
Balderston, W.L., Sherr, B., Payne, W.J., 1976. Blockage by acetylene of nitrous oxide
most uncertain in the global climate change modelling and reduction in Pseudomonas perfectomarinus. Applied and Environmental Micro-
predictions. Surface fluxes of bioaerosols are a rapidly developing biology 31, 504–508.
field, and have wide application within current priority areas in Baldocchi, D., Falge, E., Gu, L., Olson, R., Hollinger, D., Running, S., Anthoni, P.,
Bernhofer, C., Davis, K., Evans, R., Fuentes, J., Goldstein, A., Katul, G., Law, B.,
atmospheric composition, air quality and climate. Lee, X., Malhi, Y., Meyers, T., Munger, W., Oechel, W., Paw, U.K.T., Pilegaard, K.,
Schmid, H.P., Valentini, R., Verma, S., Vesala, T., Wilson, K., Wofsy, S., 2001.
FLUXNET: a new tool to study the temporal and spatial variability of ecosystem-
Acknowledgements scale carbon dioxide, water vapour, and energy flux densities. Bulletin of the
American Meteorological Society 82, 2415–2434.
Barkley, M.P., Palmer, P.I., De Smedt, I., Karl, T., Guenther, A., Van Roozendael, M., 2009.
We gratefully acknowledge financial support from the European
Regulated large-scale annual shutdown of Amazonian isoprene emissions?
Commission for the ACCENT, NitroEurope and GRAMINAE projects, Geophysical Research Letters 36, L04803. doi:10.1029/2008GL036843. 2009.
from COST for ACTION 729 and the European Science Foundation for Bassin, S., Calanca, P., Weidinger, T., Gerosa, G., Fuhrer, J., 2004. Modelling seasonal
the Nitrogen in Europe (NinE) program. This work was also supported ozone fluxes to grassland and wheat: model improvement, testing and appli-
cation. Atmospheric Environment 38, 2349–2359.
by the UK Department for Environment Food and Rural Affairs. Bauer, H., Giebl, H., Hitzenberger, R., Kasper-Giebl, A., Reischl, G., Zibuschka, F.,
Puxbaum, H., 2003. Airborne bacteria as cloud condensation nuclei. Journal of
Geophysical Research 108 (D21), 4658. doi:10.1029/2003JD003545.
References Beauchamp, J., Wisthaler, A., Hansel, A., Kleist, E., Miebach, M., Niinemets, Ü.,
Schurr, U., Wildt, J., 2005. Ozone induced emissions of biogenic VOC from
Abbatt, J.P.D., 2003. Interactions of atmospheric trace gases with ice surfaces: tobacco: relations between ozone uptake and emission of LOX products. Plant
adsorption and reaction. Chemical Reviews, ACS Publications 103, 4783. Cell and Environment 28, 1334–1343.
Abbot, D.S., Palmer, P.I., Martin, R.V., Chance, K.V., Jacob, D.J., Guenther, A., 2003. Benz, S., Megahed, K., Möhler, O., Saathoff, H., Wagner, R., Schurath, U., 2005.
Seasonal and interannual variability of isoprene emissions as determined by T-dependent rate measurements of homogeneous ice nucleation in cloud
formaldehyde column measurements from space. Geophysical Research Letters. droplets using a large atmospheric simulation chamber. Journal of Photo-
doi:10.1029/2003GL017336. chemistry and Photobiology A: Chemistry 176, 1–3 (special issue, 208–217).
Allen, A.G., Harrison, R.M., Nicholson, K.W., 1991. Dry deposition of fine aerosol to Bergamaschi, P., Krol, M., Dentener, F., Vermeulen, A., Meinhardt, F., Graul, R.,
a short grass surface. Atmospheric Environment 25A, 2671–2676. Ramonet, M., Peters, W., Dlugokencky, E.J., 2005. Onverse modelling of national
Altimir, N., Kolari, P., Tuovinen, J.-P., Vesala, T., Bäck, J., Suni, T., Kulmala, M., Hari, P., and European CH4 emissions using the atmospheric zoom model TM5. Atmo-
2006. Foliage surface ozone deposition: a role for surface moisture? Bio- spheric Chemistry and Physics 5, 2431–2460.
geosciences 3, 209–228. Beswick, K.M., Hargreaves, K.J., Gallagher, M.W., Choularton, T.W., Fowler, D., 1991.
Amato, P., Ménager, M., Sancelme, M., Laj, P., Mailhot, G., Delort A-, M., 2005. Size-resolved measurements of cloud droplet deposition velocity to a forest
Microbial population in cloud water the Puy de Dôme: implications for the canopy using an eddy-correlation technique. Quarterly Journal of the Royal
chemistry of clouds. Atmospheric Environment 39, 4143–4153. Meteorological Society 117 (499), 623–645.
Amato, P., Parazols, M., Sancelme, M., Laj, P., Mailhot, G., Delort, A.-M., 2007a. Micro- Bigg, E.K., Leck, C., 2008. The composition of fragments of bubble bursting at ocean
organisms isolated from the water phase of tropospheric clouds at the Puy de surface. Journal of Geophysical Research 113, D11209. doi:10.1029/2007JD009078.
5256 D. Fowler et al. / Atmospheric Environment 43 (2009) 5193–5267

Blake, N.J., Blake, D.R., Wingenter, O.W., Sive, B.C., Thornton, D.C., Bandy, A.R., Buzorius, G., Kalogiros, J., Varutbangkul, V., 2006. Airborne aerosol flux measure-
Atlas, E., Flocke, F., Rowland, F.S., 1999. Aircraft measurements of latitudinal, ments with eddy correlation above the ocean in a coastal environment. Journal
vertical, and seasonal variations of NMHC, methyl nitrate, and selected halo- of Aerosol Science 37, 1267–1286.
carbons during the first aerosol characterization experiment (ACE-1). Journal of Cape, J.N., Sheppard, L.J., Binnie, J., Dickinson, A.L., Dollard, G.F., Fowler, D., 1998.
Geophysical Research 104, 21803–21817. Enhancement of the dry deposition of sulphur dioxide to a forest in the
Blanchard, D.C., 1964. Sea to air transport of surface active material. Science 146, presence of ammonia. Atmospheric Environment 32, 519–524.
396–397. Cape, J.N., Kirika, A., Rowland, A.P., Wilson, D.R., Jickells, T.D., Cornell, S., 2001.
Blande, J.D., Tiiva, P., Freiwald, V., Oksanen, E., Holopainen, J.K., 2005. The effects of Organic nitrogen in precipitation: real problem or sampling artefact? The
herbivore damage and elevated ozone concentration on the volatile terpenoids Scientific World 1, 230–237.
produced by two hybrid aspen (Populus tremula x tremuloides) clones. In: Cape, J.N., 2008. Surface ozone concentrations and ecosystem health: past trends and
Abstracts. XVII International Botanical Congress, Vienna, Austria, 17–23 July a guide to future projections. Science of the Total Environment 400, 257–269.
2005, Vienna, 544 pp. Cape, J.N., Hamilton, R., Heal, M.R., 2009. Reactive uptake of ozone at simulated leaf
Blatter, A., Neftel, A., Dasgupta, P.K., Simon, P.K., 1993. A combined wet effluent surfaces: implications for ‘non-stomatal’ ozone deposition. Atmospheric Envi-
denuder and mist chamber system for deposition measurements of NH3, NHþ 4, ronment 43, 1116–1123.
HNO3 and NO 3 , 6th European Symposium of Physico-Chemical behavior of Carpenter, L.J., Lewis, A.C., Hopkins, J.R., Read, K.A., Longley, I.A., Gallagher, M.W.,
atmospheric Pollutants, Varese, Italy. 2004. Uptake of methanol to the North Atlantic Ocean surface. Global Biogeo-
Bleeker, A., Reinds, G.J., Vermeulen, A.T., de Vries, W., Erisman, J.W., 2004. Critical chemical Cycles 18 (4), GB4027.
loads and present deposition thresholds of nitrogen and acidity and their Casciotti, K.L., Böhlke, J.K., McIlvin, M.R., Mroczkowski, S.J., Hannon, J.E., 2007.
exceedances at the level II and level I monitoring plots in Europe. ECN report Oxygen isotopes in nitrite: analysis, calibration, and equilibration. Analytical
ECN-C-04–117, December 2004. Chemistry 79, 2427–2436.
Bleeker, A., Sutton, M.A., Acherman, B., Alebic-Juretic, A., Aneja, V.P., Ellermann, T., Cavalli, F., Facchini, M.C., Decesari, S., Mircea, M., Emblico, L., Fuzzi, S., Ceburnis, D.,
Erisman, J.W., Fowler, D., Fagerli, H., Gauger, T., Harlen, K.S., Hole, L.R., Horváth, L., Yoon, Y.J., O’Dowd, C.D., Putaud, J.-P., Dell’Acqua, A., 2004. Advances in char-
Mitosinkova, M., Smith, R.I., Tang, Y.S., van Pul, A., 2006. Linking ammonia acterization of size resolved organic matter in marine aerosol over the North
emission trends to measured concentrations and deposition of reduced nitrogen Atlantic. Journal of Geophysical Research. doi:10.1029/2004JD0051377.
at different scales. In: Proceedings of the UNECE NH3 Worshop, Edinburgh. Ceburnis, D., O’Dowd, C.D., Jennings, S.G., Facchini, M.C., Emblico, L., Decesari, S.,
Blitz, M.A., Heard, D.E., Pilling, M.J., Arnold, S.R., Chipperfield, M.P., 2004. Pressure Fuzzi, S., Sakalys, J., 2008. Elucidation of production mechanisms of water
and temperature-dependent quantum yields for the photodissociation of soluble and insoluble organic carbon in marine air and associated gradient
acetone between 279 and 327.5 nm. Geophysical Research Letters 31, L06111. fluxes over the Northeast Atlantic. Geophysical Research Letters 35, L07804.
doi:10.1029/2003GL018793. doi:10.1029/2008GL033462.
Bollmann, A., Conrad, R., 1997. Enhancement by acetylene of the decomposition of Chameides, W.L., Lindsay, R.W., Richardson, J., Kiang, C.S., 1988. The role of biogenic
nitric oxide in soil. Soil Biology and Biochemistry 29, 1057–1066. hydrocarbons in urban photochemical smog: Atlanta as a case study. Science
Bonn, B., Hirsikko, A., Hakola, H., Kurtén, T., Laakso, L., Boy, M., Dal Maso, M., 241, 1473–1475.
Mäkelä, J.M., Kulmala, M., 2007. Ambient sesquiterpene concentration and its link Chance, K., Palmer, P.I., Martin, R.V., Spurr, R.J.D., Kurosu, T.P., Jacob, D.J., 2000.
to air ion measurements. Atmospheric Chemistry and Physics 7, 2893–2916. Satellite observations of formaldehyde over North America from GOME.
Bonsang, B., Polle, C., Lambert, G., 1992. Evidence for marine production of isoprene. Geophysical Research Letters 27, 3461–3464.
Geophysical Research Letters 19, 1129–1132. Chang, C., Janzen, H.H., Cho, C.M., Nakonechny, E.M., 1998. Nitrous oxide emission
Braaten, D.A., Paw, U.K.T., 1992. A stochastic particle resuspension and deposition through plants. Soil Science Society of America Journal 62, 35–38.
model. In: Fifth International Conference on Precipitation Scavenging and Chang, W., Heikes, B.G., Lee, M., 2004. Ozone deposition to the sea surface: chemical
Atmosphere–Surface Exchange Processes, Richland, Washington, 15–19 July enhancement and wind speed dependence. Atmospheric Environment 38,
1991, pp. 1143–1152. 1053–1059.
Brilli, F., Barta, C., Fortunati, A., Lerdau, M., Loreto, F., Centritto, M., 2007. Response of Chapuis-Lardy, L.E., Wrage, N., Metay, A., Chotte, J.L., Bernoux, M., 2007. Soils, a sink
isoprene emission and carbon metabolism to drought in white poplar (Populus for N2O? A review. Global Change Biology 13, 1–17.
alba) saplings. New Phytologist 175, 244–254. Chen, X., Boeckx, P., Shen, S., Van Cleemput, O., 1999. Emission of N2O from rye grass
Brost, R.A., Delany, A.C., Huebert, B.J., 1988. Numerical modeling of concentrations (Lolium perenne L.). Biology and Fertility of Soils 8, 393–396.
and fluxes of HNO3, NH3, and NH4NO3 near the surface. Journal of Geophysical Christensen, S., Ambus, P., Arah, J.R.M., Clayton, H., Galle, B., Griffith, D.W.T.,
Research 93, 7137–7152. Hargreaves, K.J., Klemedtsson, L., Lind, A.-M., Maag, M., Scott, A., Skiba, U.,
Brunner, A., Ammann, C., Neftel, A., Spirig, C., 2007. Methanol exchange between Smith, K.A., Welling, M., Wienhold, F.G., 1996. Nitrous oxide emission from an
grassland and the atmosphere. Biogeosciences 4, 395–410. agricultural field: comparison between measurements by flux chamber and
Buchwitz, M., de Beek, R., Burrows, J.P., Bovensmann, H., Warneke, T., Notholt, J., micrometerological techniques. Atmospheric Environment 30, 4183–4190.
Meirink, J.F., Goede, A.P.H., Bergamaschi, P., Körner, S., Heimann, M., Schulz, A., Christensen, C.S., Hummelshøj, P., Jensen, N.O., Larsen, B., Lohse, C., Pilegaard, K.,
2005. Atmospheric methane and carbon dioxide from SCIAMACHY satellite Skov, H., 2000. Determination of the terpene flux from orange species and
data: initial comparison with chemistry and transport models. Atmospheric Norway spruce by relaxed eddy accumulation. Atmospheric Environment 34,
Chemistry and Physics 5, 941–962. 3057–3067.
Buchwitz, M., de Beek, R., Noel, S., 2006. Atmospheric carbon gases retrieved Chuck, A.L., Turner, S.M., Liss, P.S., 2002. Direct evidence of a marine source of C1
from SCIAMACHY by WFM-DOAS: version 0.5 CO and CH4 and impact of and C2 alkyl nitrates. Science 297, 1151–1154.
calibration improvements on CO2 retrieval. Atmoshperic Chemistry and Ciccioli, P., Fabozzi, C., Brancaleoni, E., Cecinato, A., Frattoni, M., Cieslik, S.,
Physics 6, 2727–2751. Kotzias, D., Seufert, G., Foster, P., Steinbrecher, R., 1997. Biogenic emission from
Burkhard, R., Eugster, W., Wrzesinski, T., Klemm, O., 2002. Vertical divergence of the Mediterranean pseudosteppe ecosystem present in Castelporziano. Atmo-
fogwater fluxes above a spruce forest. Atmospheric Research 64, 133–145. spheric Environment 31, 167–175.
Burkhardt, J., Flechard, C.R., Gresens, F., Mattsson, M.E., Jongejan, P.A.C., Erisman, J.W., Ciccioli, P., Brancaleoni, E., Frattoni, M., Di Palo, V., Valentini, R., Tirone, G.,
Weidinger, T., Meszaros, R., Nemitz, E., Sutton, M.A., 2008. Modelling the chemical Seufert, G., Bertin, N., Hansen, U., Csiky, O., Lenz, R., Sharma, M., 1999.
interactions of atmospheric ammonia and other trace gases fluxes with measured Emission of reactive terpene compounds from orange orchards and their
leaf surface wetness in a managed grassland canopy. Biogeosciences Discussions removal by within-canopy processes. Journal of Geophysical Research 104,
5, 2505–2539. 8077–8094.
Businger, J.A., Wyngaard, J., Izumi, Y., Bradley, E.F., 1971. Flux–profile relationships Cicerone, R., 1989. Analysis of sources and sinks of atmospheric nitrous oxide (N2O).
in the atmospheric surface layer. Journal of the Atmospheric Sciences 28, Journal of Geophysical Research 94, 18265–18271.
181–189. Cieslik, S.A., 2004. Ozone uptake by various surface types: a comparison between
Butterbach-Bahl, K., Gasche, R., Breuer, L., Papen, H., 1997. Fluxes of NO and N2O dose and exposure. Atmospheric Environment 38, 2409–2420.
from temperate forest soils: impact of forest type, N deposition and of liming on Cieslik, S. 2009. Ozone fluxes over various plant ecosystems in Italy: a review.
the NO and N2O emissions. Nutrient Cycling in Agroecosystems 48, 1314–1385. Ozone and Mediterranean Ecology: Plants, People, Problems. Environmental
Butterbach-Bahl, K., Willibald, G., Papen, H., 2002. Soil core method for direct Pollution, 157, 1487–1496 (Special Issue Section)
simultaneous determination of N2 and N2O emissions from forest soils. Plant Clarke, A.D., Owens, S.R., Zhou, J., 2006. An ultrafine sea-salt flux from breaking
and Soil 240, 105–116. waves: implications for cloud condensation nuclei in the remote marine
Butterbach-Bahl, K., Kesik, M., Miehle, P., Papen, H., Li, C., 2004. Quantifying the atmosphere. Journal of Geophysical Research 111, D06202. doi:10.1029/
regional source strength of N-trace gases across agricultural and forest 2005JD006565. 14.
ecosystems with process based models. Plant and Soil 260, 311–329. Clifford, D., Donaldson, D.J., Brigante, M., D’Anna, B., George, D., 2008. Reactive
Butterbach-Bahl, K., Mykhayliv, L., Werner, C., Kiese, R., Li, C., 2009. A European uptake of ozone by chlorophyll at aqueous surfaces. Environmental Science &
wide inventory of soil NO emissions using the biogeochemical models DNDC/ Technology 42, 1138–1143.
Forest DNDC. Atm. Environm 43, 1392–1402. Colbeck, I., Harrison, R.M., 1985. The photochemical pollution episode of 5–16 July
Buzorius, G., Rannik, U., Makela, J.M., Vesala, T., Kulmala, M., 1998. Vertical aerosol 1983 in North-West England. Atmospheric Environment 19, 1921–1929.
particle fluxes measured by eddy covariance technique using condensational Conen, F., Smith, K.A., 1998. A re-examination of closed flux chamber methods for
particle counter. Journal of Aerosol Science 29, 157–171. the measurement of trace gas emissions from soils to the atmosphere. Euro-
Buzorius, G., Rannik, U., Nilsson, D., Kulmala, M., 2001. Vertical fluxes and micro- pean Journal of Soil Science 49, 701–707.
meteorology during aerosol particle formation events. Tellus Series B-Chemical Conen, F., Neftel, A., 2007. Do increasingly depleted d15N values of atmospheric N2O
and Physical Meteorology 53, 394–405. indicate a decline in soil N2O reduction? Biogeochemistry 82, 321–326.
D. Fowler et al. / Atmospheric Environment 43 (2009) 5193–5267 5257

Conrad, R., 1995. Soil microbial processes and the cyclig of atmospheric trace gases. Missing OH reactivity in a forest: evidence for unknown reactive biogenic VOCs.
Philosophical Transactions of the Royal Society of London 351, 219–230. Science 304, 722–724.
Cornell, S.E., Jickells, T.D., Cape, J.N., Rowland, A.P., Duce, R.A., 2003. Organic Di Marco, C., Skiba, U., Weston, K., Hargreaves, K., 2004. Field scale N2O flux
nitrogen deposition on land and coastal environments: a review of methods measurements from grassland using eddy covariance. Water, Air and Soil
and data. Atmospheric Environment 37, 2173–2191. Pollution: Focus 4, 143–149.
Coyle, M., Smith, R., Fowler, D., 2003. An ozone budget for the UK: using Dollard, G.J., Unsworth, M.H., 1983. Field measurements of turbulent fluxes of wind-
measurements from the national ozone monitoring network; measured and driven fog drops to a grass surface. Atmospheric Environment 17, 775–780.
modelled meteorological data, and a ‘big-leaf’ resistance analogy model for dry Donateo, A., Contini, D., Belosi, F., 2006. Real time measurements of PM2.5
deposition. Environmental Pollution 123, 115–123. concentrations and vertical turbulent fluxes using an optical detector. Atmo-
Coyle, M., 2005. The Gaseous Exchange of Ozone at Terrestrial Surfaces: Non- spheric Environment 40, 1346–1360.
stomatal Deposition to Grassland, School of Geosciences, Faculty of Science and Dore, A.J., Mousavi-Baygi, M., Smith, R.I., Hall, J., Fowler, D., Choularton, T.W., 2006.
Engineering. PHD thesis. The University of Edinburgh, Edinburgh, 270 pp. A model of annual orographic precipitation and acid deposition and its appli-
Coyle, M., Fowler, D., Nemitz, E., Philips, G., Storeton-West, R., Thomas, R., 2006. cation to Snowdonia. Atmospheric Environment 40, 3316–3326.
Field measurements of the ozone flux to vegetation. In: Ozone Umbrella: Effects Dorsey, J.R., Nemitz, E., Gallagher, M.W., Fowler, D., Williams, P.I., Bower, K.N.,
of Ground-level Ozone on (Upland) Vegetation in the UK. Centre of Ecology and Beswick, K.M., 2002. Direct measurements and parameterisation of aerosol flux,
Hydrology, UK, pp. 68–104. CEH C02158 Report No. AS 06/02. concentration and emission velocity above a city. Atmospheric Environment 36,
Coyle, M., Altimir, N., Grunhage, L., Gerosa, G., Tuovinen, J.P., Loubet, B., Ammann, C., 791–800.
Cieslik, S., Paoletti, E., Fowler, D., Mikkleson, T., Ro-Poulsen, H., Cellier, P., 2008a. Dorsey, J.R., Duyzer, J.H., Gallagher, M.W., Coe, H., Pilegaard, K., Weststrate, J.H.,
Non-stomatal Ozone Deposition to Vegetation: New Insights and Models. EGU Jensen, N.O., Walton, S., 2004. Oxidized nitrogen and ozone interaction with
General Assembly 2008. European Geophysical Union, Vienna, Austria. forest. I: experimental observations and analysis of exchange with Douglas fir.
Coyle, M., Nemitz, E., Storeton-West, R., Fowler, D., Cape, J.N. 2009. Measurements of Quarterly Journal of the Royal Meteorological Society 130, 1941–1955.
ozone deposition to a potato canopy. Agricultural and Forest Meteorology, 149, Duan, B., Fairall, C.W., Thomson, D.W.,1988. Eddy-covariance measurements of the dry
656–666. deposition of particles in wintertime. Journal of Applied Meteorology 28, 642–652.
Crutzen, P.J., Lawrence, M.G., 2000. The impact of precipitation scavenging on the Dueck, T.A., de Visser, R., Poorter, H., Persijn, S., Gorissen, A., de Visser, W.,
transport of race gases: a 3-D dimensional model sensitivity study. Journal of Schapendonk, A., Verhagen, J., Snel, J., Harren, F.J.M., Ngai, A.K.Y., Verstappen, F.,
Atmospheric Chemistry 37, 81–112. Bouwmeester, H., Voesenek, L.A.C.J., van der Werf, A., 2007. No evidence for
Crutzen, P.J., Mosier, A.R., Smith, K.A., Winiwarter, W., 2008. N2O release from agro- substantial aerobic methane emission by terrestrial plants: a 13C-labelling
biofuel production negates global warming reduction by replacing fossil fuels. approach. New Phytologist 175, 29–35.
Atmospheric Chemistry and Physics 8, 389–395. Duyzer, J.H., Meyer, G.M., van Aalst, R.M., 1983. Measurement of dry deposition
Cubison, M.J., Alfarra, M.R., Allan, J., Bower, K.N., Coe, H., McFiggans, G.B., velocities of NO, N and O3 and the influence of chemical reactions. Atmospheric
Whitehead, J.D., Williams, P.I., Zhang, Q., Jimenez, J.L., Hopkins, J., Lee, J., 2006. The Environment 17 (10), 2117–2120 (1967).
characterisation of pollution aerosol in a changing photochemical environment. Duyzer, J., 1994. Dry deposition of ammonia and ammonium aerosols over heath-
ACP 6, 1–15. land. Journal of Geophysical Research – Atmospheres 99, 18757–18763.
Dahl, E.E., Saltzman, E.S., de Bruyn, W.J., 2003. The aqueous phase yield of alkyl Duyzer, J., Weststrate, H., Walton, S., 1995. Exchange of ozone and nitrogen oxides
nitrates from ROO þ NO: implications for photochemical production in between the atmosphere and coniferous forest. Journal Water, Air, & Soil
seawater. Geophysical Research Letters 30 (6), 1271. doi:10.1029/2002GL016811. Pollution 85 (4), 2065–2070. Springer, Netherlands.
Darmais, S., Dutaur, L., Larsen, B., Cieslick, S., Luchetta, L., Simon, V., Torres, L., 2000. Duyzer, J.H., Dorsey, J.R., Gallagher, M.W., Pilegaard, K., Walton, S., 2004. Oxidized
Emission flux of VOC by orange trees determined by both eddy accumulation nitrogen and ozone interaction with forests. II: multi-layer process-oriented
and vertical gradient approaches. Chemosphere 2, 47–56. modelling results and a sensitivity study for Douglas fir. Quarterly Journal of the
Davidson, C.I., Miller, J.M., Pleskow, M.A., 1982. The influence of surface structure on Royal Meteorological Society 130 (600 PART A), 1957–1971.
predicted particle dry deposition to natural grass canopies. Water, Air and Soil Duyzer, J., 2005a. Interactive comment on ‘‘A simple model to estimate exchange
Pollution 18, 25–43. rates of nitrogen dioxide between the atmosphere and forests’’. Biogeosciences
Davidson, E.A., 1991. Fluxes of nitrous oxide and nitric oxide from terrestrial Discussions 2 (Suppl.), S905–S911.
ecosystems. In: Rogers, J.E., Whitman, W.B. (Eds.), Microbial Production and Duyzer, J., 2005b. A simple model to estimate exchange rates of nitrogen dioxide between
Consumption of Greenhouse Gases: Methane, Nitrogen Oxides, and Halo- the atmosphere and forests. Biogeosciences Discussions 2, 1033–1065. 2005.
methanes. American Society for Microbiology, Washington, pp. 219–235. Eastman, J.A., Stedman, D.H., 1977. A fast sensor for ozone eddy-correlation flux
Davison, B., Brunner, A., Ammann, C., Spirig, C., Jocher, M., Neftel, A., 2008. measurements. Atmospheric Environment 11, 1209–1211.
Cut-induced VOC emissions from agricultural grasslands. Plant Biology 10, Ekdahl, A., Pedersén, M., Abrahamsson, K., 1998. A study of the diurnal variation of
76–85. biogenic volatile halocarbons. Marine Chemistry 63, 1–8.
Davison, B., Taipale, R., Langford, B., Misztal, P., Fares, S., Matteucci, G., Loreto, F., Emberson, L.D., Ashmore, M.R., Cambridge, H.M., Simpson, D., Tuovinen, J.-P., 2000.
Cape, J.N., Rinne, J., Hewitt, C.N., 2009. Concentrations and fluxes of biogenic Modelling stomatal ozone flux across Europe. Environmental Pollution 109,
volatile organic compounds above a Mediterranean macchia ecosystem in 403–413.
Western Italy. Biogeosciences Discussions 6, 2183–2216. Emberson, L.D., Büker, P., Ashmore, M.R., 2007. Assessing the risk caused by ground
De Arellano, J., Duynkerke, P., 1992. Influence of chemistry on flux–gradient rela- level ozone to European forest trees: a case study in pine, beech and oak across
tionships in the NO–O3–NO2 system. Boundary-Layer Meteorology 28, 181–189. different climate regions. Environmental Pollution 147, 454–466.
De Carlo, P.F., Kimmel, J.R., Trimborn, A., Northway, M.J., Jayne, J.T., Aiken, A.C., Erisman, J.W., 1993. Acid deposition onto nature areas in The Netherlands. Part II:
Gonin, M., Fuhrer, K., Horvath, T., Docherty, K., Worsnop, D.R., Jimenez, J.L., throughfall measurements compared to deposition estimates. Water Air Soil
2006. Field-deployable, high-resolution, time-of-flight aerosol mass spec- Pollut. 71, 81–99.
trometer. Analytical Chemistry 78, 8281–8289. Erisman, J.W., Wyers, G.P., 1993. Continuous measurements of surface exchange of
Deguillaume, L., Leriche, M., Desboeufs, K., Mailhot, G., George, C., Chaumerliac, N., SO2 and NH3: implications for their possible interaction in the deposition
2005. Transition metals in atmospheric liquid phases: sources, reactivity, and process. Atmospheric Environment 27A, 1937–1949.
sensitive parameters. Chemical Reviews 105 (9), 3388–3431. Erisman, J.W., 1994. Evaluation of a surface resistance parametrization of sulphur
Deguillaume, L., Leriche, M., Amato, P., Ariya, P.A., Delort, A.-M., Pöschl, U., dioxide. Atmospheric Environment 28, 2583–2594.
Chaumerliac, N., Bauer, H., Flossmann, A.I., Morris, C.E., 2008. Microbiology and Erisman, J.W., Draaijers, G., Duyzer, J., Hofschreuder, P., Van Leeuwen, N., Roemer, F.,
atmospheric processes: chemical interactions of primary biological aerosols. Ruijgrok, W., Wyers, P., Gallagher, M., 1996. Particle deposition to forests –
Biogeosciences 5, 1073–1084. summary of results and application. Atmospheric Environment 31, 321–332.
Denmead, O.T., Freney, J.R., Simpson, J.R., 1976. A closed ammonia cycle within Erisman, J.W., Hensen, A., Fowler, D., Flechard, C., Grüner, A., Spindler, G., Duyzer, J.,
a plant canopy. Soil Biology and Biochemistry 8, 161–164. Weststrate, H., Römer, F., Vonk, A., van Jaarsveld, H., 2001. Dry deposition
Denmead, O.T., 2008. Approaches to measuring fluxes of methane and nitrous oxide monitoring in Europe. WASP Focus 1, 17–27.
between landscapes and the atmosphere. Plant and Soil 309, 5–24. Erisman, J.W., Sutton, M.A., Galloway, J., Klimont, Z., Winiwarter, W., 2008. How
Dentener, F.J., Crutzen, P.J., 1994. Atmospheric ammonia from undisturbed land. a century of ammonia synthesis changed the world. Nature Geoscience 1, 636–
Journal of Geophysics Research 82, 3125–3133. 639. doi:10.1038/ngeo325.
Dentener, F., Drevet, J., Lamarque, J.F., Bey, I., Eickhaut, B., Fiore, A.M., Ervens, B., George, C., Williams, J.E., Boxton, G.V., Salmon, G.A., Bydder, M.,
Hauglustaine, D., Horowitz, L.W., Krok, M., Kulshrestha, U.C., Lawrence, M., Wilkinsons, F., Dentener, F., Mirabel, P., Wolke, R., Herrmann, H., 2004a. CAP-
Galy-Lacaux, C., Rast, S., Shindell, D., Stevenson, D., van Noije, T., Atherton, C., RAM 2.4 (MODAC mechanism): an extended and condensed tropospheric
Bell, N., Bergman, D., Butler, T., Cofala, J., Collins, B., Docherty, R., Ellingsen, K., aqueous phase mechanism and its application. Journal of Geophysical Research
Galloway, J., Gauss, M., Montanaro, V., Müller, J.F., Pitari, G., Rodriguez, J., 108. doi:10.1029/2002JD002202.
Sanderson, M., Solmon, F., Strahan, S., Schulz, M., Sudo, K., Szopa, S., Wild, O., Ervens, B., Feingold, G., Frost, G.J., Kreidenweis, S.M., 2004b. A modeling study of
2006. Nitrogen and sulfur deposition on regional and global scales: a multi- aqueous production of dicarboxylic acids: 1. Chemical pathways and speciated
model evaluation. Global Biogeochemistry Cycles 20 (4), GB4003. organic mass production. Journal of Geophysical Research 109, D15205.
Derome, J., Nieminen, T., Saarsalmi, A., 2004. Sulphur dioxide adsorption in Scots doi:10.1029/2003JD004387.
pine canopies exposed to high ammonia emissions near a Cu–Ni smelter in SW Facchini, M.C., Decesari, S., Rinaldi, M., Carbone, C., Finessi, E., Mircea, M., Fuzzi, S.,
Finland. Environmental Pollution 129, 79–88. Moretti, F., Tagliavini, E., Ceburnis, D., O’Dowd, C.D., 2008. An important source
Di Carlo, P., Brune, W.H., Martinez, M., Harder, H., Lesher, R., Ren, X., Thornberry, T., of marine secondary organic aerosol from biogenic amines. Environmental
Carroll, M.A., Young, V., Shepson, P.B., Riemer, D., Apel, E., Campbell, C., 2004. Science & Technology. doi:10.1021/es8018385.
5258 D. Fowler et al. / Atmospheric Environment 43 (2009) 5193–5267

Fairall, C.W., 1984. Interpretation of eddy-correlation measurements of particulate for the measurement and interpretation of gas and particle nitrogen fluxes.
deposition and aerosol flux. Atmospheric Environment 18, 1329–1337. Plant and Soil 228, 117–129.
Fairall, C.W., Helmig, D., Ganzeveld, L., Hare, J., 2007. Water-side turbulence Fowler, D., Sutton, M.A., Flechard, C., Cape, J.N., Storeton-West, R., Coyle, M.,
enhancement of ozone deposition to the ocean. Atmospheric Chemistry and Smith, R.I., 2001c. The control of SO2 dry deposition on to natural surfaces and
Physics 7, 443–451. its effects on regional deposition. Water, Air and Soil Pollution: Focus 1, 39–48.
Fall, R., 2003. Abundant oxygenates in the atmosphere: a biochemical perspective. Fowler, D., Smith, R.I., Muller, J.B.A., Hayman, G., Vincent, K.J., 2005. Changes in the
Chemical Reviews 103, 4941–4951. atmospheric deposition of acidifying compounds in the UK between 1986 and
Fall, R., Karl, T., Hansel, A., Jordan, A., Lindinger, W., 1999. Volatile organic 2001. Environmental Pollution 137, 15–25.
compounds emitted after leaf wounding: online analysis by proton transfer- Fowler, D., Smith, R.I., Muller, J., Cape, J.N., Sutton, M., Erisman, J.W., Fagerli, H., 2007.
reaction mass spectrometry. Journal of Geophysical Research 104, 15963–15974. Long term trends in sulphur and nitrogen deposition in Europe and the cause of
Famulari, D., Fowler, D., Hargreaves, K., Milford, C., Nemitz, E., Sutton, M., non-linearities. Water, Soil and Air Pollution: Focus 7, 41–47.
Weston, K., 2004. Measuring eddy covariance fluxes of ammonia using tunable Frankenberg, C., Bergamaschi, P., Butz, A., Houweling, S., Meirink, J.F., Notholt, J.,
diode laser absorption spectroscopy. WASP Focus 4 (6), 151–158. Petersen, A.K., Schrijver, H., Warneke, T., Aben, I., 2008. Tropical methane
Famulari, D., Fowler, D., Storeton-West, R.L., Nemitz, E., Hargreave, K., Rutherford, G., emissions: a revised view from SCIAMACHY onboard ENVISAT. Geophysical
Tang, Y.S, Sutton, M.A., Weston, K.J. Development of a low-cost system for Research Letters 35, LI15811.
measuring conditional time-averaged gradients of SO2 and NH3. Environmental Fratini, G., Ciccioli, P., Febo, A., Forgione, A., Valentini, R., 2007. Size-segregated
Monitoring and Assessment, in press, doi:10.1007/s10661-008-0723-6. fluxes of mineral dust from a desert area of northern China by eddy covariance.
Fang, Y.T., Gundersen, P., Mo, J.M., Zhu, W.X., 2008. Input and output of dissolved Atmospheric Chemistry and Physics 7, 2839–2854.
organic and inorganic nitrogen in subtropical forests of South China under high Fuentes, J.D., 1992. Effect of Foliage Surface Wetness on the Deposition of Ozone.
air pollution. Biogeosciences 5, 339–352. Ph.D. thesis. University of Guelph, Guelph, Ontario, Canada.
Farmer, D.K., Woolridge, P.J., Cohen, R.C., 2006. Application of thermal-dissociation Fuhrer, J., Skärby, L., Ashmore, M.R., 1997. Critical levels for ozone effects on vege-
laser induced fluorescence (TD-LIF) to measurement of HNO3, aklyl nitrates, tation in Europe. Environmental Pollution 97, 91–106.
peroxy nitrates and NO2 fluxes using eddy covariance. Atmospheric Chemistry Fuzzi, S., Mandrioli, P., Perfetto, A., 1997. Fog droplets – an atmospheric source of
and Physics 6, 3471–3486. secondary biological aerosol particles. Atmospheric Environment 31, 287–290.
Farquhar, G.D., Firth, P.M., Wetselaar, R., Wier, B., 1980. On the gaseous exchange of Galbally, I.E., 1971. Ozone profiles and fluxes in the atmospheric surface layer.
ammonia between leaves and the environment: determination of the ammonia Quarterly Journal of the Royal Meteorological Society 97, 18–29.
compensation point. Plant Physiology 66, 710–714. Galbally, I.E., Roy, C.R., 1980. Destruction of ozone at the earth’s surface. Quarterly
Feingold, G., Kreidenweis, S.M., 2002. Cloud processing of aerosol as modeled by Journal of the Royal Meteorological Society 106, 599–620.
a large eddy simulation with coupled microphysics and aqueous chemistry. Galbally, I.E., Kirstine, W., 2002. The production of methanol by flowering plants
Journal of Geophysical Research 107, 4687. doi:10.1029/2002JD002054. and the global cycle of methanol. Journal of Atmospheric Chemistry 43,
Feliciano, M.S., Pio, C.A., Vermeulen, A.T., 2001. Evaluation of SO2 dry deposition 195–229.
over short vegetation in Portugal. Atmospheric Environment 35, 3633–3643. Gallagher, M.W., Beswick, K.M., Duyzer, J., Westrate, H., Choularton, T.W.,
Felzer, B., Reilly, J., Melillo, J., Kicklighter, D., Sarofim, M., Wang, C., Prinn, R., Zhuang, Q., Hummelshoj, P., 1997. Measurements of aerosol fluxes to Speulder forest using
2005. Future effects of ozone on carbon sequestration and climate change policy a micrometeorological technique. Atmospheric Environment 31, 359–373.
using a global biogeochemical model. Climatic Change 73, 345–373. Gallagher, M.W., Beswick, K.M., Coe, H., 2001. Ozone dry deposition to coastal
Finkelstein, P.L., Ellestad, T.G., Clarke, J.F., Meyers, T.P., Schwede, D., Hebert, E.O., waters. Quarterly Journal of the Royal Meteorological Society 127, 539–558.
Neal, J.F., 2000. Ozone and sulfur dioxide dry deposition to forests: observations Galloway, J.N., Aber, J.D., Erisman, J.W., Seitzinger, S.P., Howarth, R.W., Cowling, E.B.,
and model evaluation. Journal of Geophysical Research 105, 15365–15377. Cosby, B.J., 2003. The nitrogen cascade. BioScience 53, 341–356.
Fiore, A.M., Horowitz, L.W., Purves, D.W., Levy II, H., Evans, M.J., Wang, Y., Li, Q., Galloway, J.N., Dentener, F.J., Capone, D.G., Boyer, E.W., Howarth, R.W.,
Yantosca, R.M., 2005. Evaluating the contribution of changes in isoprene Seitzinger, S.P., Asner, G.P., Cleveland, C.C., Green, P.A., Holland, E.A., Karl, D.M.,
emissions to surface ozone trends over the eastern United States. Journal of Michaels, A.F., Porter, J.H., Townsend, A.R., Vorosmarty, C.J., 2004. Nitrogen
Geophysical Research 110, D12303. doi:10.1029/2004JD005485. cycles: past, present, and future. Biogeochemistry 70, 153–226.
Fischbach, R.J., Staudt, M., Zimmer, I., Rambal, S., Schnitzler, J.P., 2002. Seasonal Gaman, A., Rannik, U., Aalto, P., Pohja, T., Siivola, E., Kumala, M., Vesala, T., 2004.
pattern of monoterpene synthase activities in leaves of the evergreen tree Relaxed eddy accumulation system for size resolved aerosol particle flux
Quercus ilex. Physiologia Plantarum 114, 354–360. measurements. Journal of Atmospheric and Oceanic Technology 21, 933–943.
Fisher, B.E.A., 1978. The calculation of long term sulphur deposition in Europe. Gao, W., Shaw, R.H., Paw, U.K.T., 1989. Observation of organized structures in
Atmospheric Environment 12, 489–501. turbulent flow within and above a forest canopy. Boundary-Layer Meteorology
Flechard, C.R., Fowler, D., 1998. Atmospheric ammonia at a moorland site. II: long- 47, 349–377.
term surface–atmosphere micrometeorological flux measurements. Quarterly Garland, J.A., Derwent, R.G., 1979. Destruction at the ground and the diurnal cycle of
Journal of the Royal Meteorological Society 124, 759–791. concentration of ozone and other gases. Quarterly Journal of the Royal Mete-
Flechard, C., Fowler, D., Sutton, M.A., Cape, J.N., 1999. A dynamic chemical model orological Society 105, 169–183.
of bi-directional ammonia exchange between semi-natural vegetation and Gasche, R., Papen, H., 1999. A 3-year continuous record of nitrogen trace gas fluxes
the atmosphere. Quarterly Journal of the Royal Meteorological Society 125, from untreated and limed soil of a N-saturated spruce and beech forest
2611–2641. ecosystem in Germany. Part 2: NO and NO2 fluxes. Journal of Geophysical
Flechard, C.R., Neftel, A., Jocher, M., Ammann, C., Leifeld, J., 2007. Temporal changes Research D, Atmospheres 104 (15), 18505–18520.
in soil pore space CO2 concentration and storage under permanent grassland. Geever, M., O’Dowd, C., van Ekeren, S., Flanagan, R., Nilsson, D., de Leeuw, G.,
Agricultural and Forest Meteorology 142, 66–84. Rannik, U., 2005. Sub-micron sea-spray fluxes. GRL 32 (2005), L15810.
Flossmann, A.I., Wobrock, W., 1996. Venting of gases by convective clouds. Journal Gerosa, G., Cieslik, S., Ballarin-Denti, A., 2003. Micrometeorological determination
of Geophysical Research 101, 18639–18649. of time-integrated stomatal ozone fluxes over wheat: a case study in Northern
Flossmann, A., 1998a. Clouds and pollution. Pure & Applied Chemistry 70, 1345–1352. Italy. Atmospheric Environment 37, 777–788.
Flossmann, A.I., 1998b. Interaction of aerosol particles and clouds. Journal of Gerosa, G., Marzuoli, R., Cieslik, S., Ballarin-Denti, A., 2004. Stomatal ozone
Atmospheric Science 55, 879–887. uptake by barley in Italy. ‘‘Effective exposure’’ as a possible link between
Fontan, J., Lopez, A., Lamaud, E., Druilhet, A., 1997. Vertical flux measurements of the concentration- and flux-based approaches. Atmospheric Environment 38,
submicronic aerosol particles and parameterization of the dry deposition 2421–2432.
velocity. In: Slanina, J. (Ed.), Biosphere–Atmosphere Exchange of Pollutants and Gerosa, G., Vitale, M., Finco, A., Manes, F., Ballarin Denti, A., Cieslik, S., 2005. Ozone
Trace Substances. Springer Verlag, Heidelberg, Germany, pp. 381–390. uptake by an evergreen Mediterranean forest (Quercus ilex) in Italy. Part I:
Fowler, D., Unsworth, M.H., 1974. Dry deposition of sulphur dioxide on wheat. micrometeorological flux measurements and flux partitioning. Atmospheric
Nature, London 249, 389–390. Environment 39, 3255–3266.
Fowler, D., Unsworth, M.H., 1979. Turbulent transfer of sulphur dioxide to a wheat Gerosa, G., Finco, A., Mereu, S., Vitale, M., Manes, F., Ballarin Denti, A., 2008. Comparison
crop. Quarterly Journal of the Royal Meteorological Society 105, 767–783. of seasonal variations of ozone exposure and fluxes in a Mediterranean Holm oak
Fowler, D., Duyzer, J.H., 1989. Micrometeorological techniques for the measurement forest between the exceptionally dry 2003 and the following year. Environmental
of trace gas exchange. In: Andreae, M.O., Schimel, D.S. (Eds.), Exchange of Trace Pollution 157 (15), 1737–1744. doi:10.1016/j.envpol.2007.11.025.
Gases Between Terrestrial Ecosystems and the Atmosphere. Wiley, Chichester, Gershenzon, J., Dudareva, N., 2007. The function of terpene natural products in the
UK, pp. 189–207. natural world. Nature Chemical Biology 3, 408–414.
Fowler, D., Flechard, C.R., Sutton, M.A., Storeton-West, R.L., 1998. Long term Gillette, D.A., Chen, W., 2001. Particle production and aeolian transport from
measurements of land–atmosphere exchange of ammonia over moorland. a ‘‘supply-limited’’ source area in the Chihuahuan desert, New Mexico, United
Atmospheric Environment (Ammonia Special Issue) 32, 453–459. States. Journal of Geophysical Research – Atmospheres 106, 5267–5278.
Fowler, D., Cape, J.N., Coyle, M., Flechard, C., Kuylenstierna, J., Hicks, K., Gödde, M., Conrad, R., 1998. Simultaneous measurement of nitric oxide production
Derwent, R.G., Johnson, C., Stevenson, D., 1999. The global exposure of forests to and consumption in soil using a simple static incubation system, and the effect
air pollutants. Water, Air and Soil Pollution 116, 5–32. of soil water content on the contribution of nitrification. Soil Biology and
Fowler, D., Flechard, C., Cape, J.N., Storeton-West, R.L., Coyle, M., 2001a. Measure- Biochemistry 30, 433–442.
ments of ozone deposition to vegetation quantifying the flux, the stomatal and Goldstein, A.H., McKay, M., Kurpius, M.R., Schade, G.W., Lee, A., Holzinger, R.,
non-stomatal components. Water, Air and Soil Pollution 130, 63–74. Rasmussen, R.A., 2004. Forest thinning experiment confirms ozone deposition
Fowler, D., Coyle, M., Flechard, C., Hargreaves, K.J., Nemitz, E., Storeton-West, R., to forest canopy is dominated by reaction with biogenic VOCs. Geophysical
Sutton, M.A., Erisman, J.W., 2001b. Advances in micrometeorological methods Research Letters 31, L22106. doi:10.1029/2004GL021259.
D. Fowler et al. / Atmospheric Environment 43 (2009) 5193–5267 5259

Gribble, G.W., 1992. Sources of halogenated alkanes. Chemical & Engineering News Held, A., Nowak, A., Wiedensohler, A., Klemm, O., 2006. Field measurements and
70 3–3. size-resolved model simulations of turbulent particle transport to a forest
Grivicke, R., Pressley, S., Allwine, G., Jobson, T., Westberg, H., Lamb, B., Jimenez, J.L., canopy. Journal of Aerosol Science 37, 786–798.
Nemitz, E., Alexander, L., Velasco, E., Molina, L., Ramos, R., 2007. Eddy covari- Hellén, H., Hakola, H., Pystynen, K.-H., Rinne, J., Haapanala, S., 2006. C2–C10
ance flux measurements of urban aerosols and related urban gaseous pollutants hydrocarbon emissions from a boreal wetland and forest floor. Biogeosciences
during the MILAGRO Mexico City Field Campaign. In: American Association for 3, 167–174.
Aerosol Research Annual Meeting 2007, Reno, 24–28 Sept 2007. American Helmig, D., Ganzeveld, L., Butler, T., Oltmans, S.J., 2007. The role of ozone
Association for Aerosol Research. atmosphere–snow gas exchange on polar, boundary-layer tropospheric
Groffman, P.M., Altabet, M.A., Böhlke, J.K., Butterbach-Bahl, K., David, M.B., ozone – a review and sensitivity analysis. Atmospheric Chemistry and
Firestone, M.K., Giblin, A.E., Kana, T.M., Nielsen, L.P., Voytek, M.A., 2006. Physics 7, 15–30.
Methods for measuring denitrification: diverse approaches to a difficult Hensen, A., Groot, T.T., van den Bulk, W.C.M., Vermeulen, A.T., Olesen, J.E.,
problem. Ecological Applications 16, 2091–2122. Schelde, K., 2006. Dairy farm CH4 and N2O emissions, from one square metre to
Grulke, N.E., Alonso, R., Nguyen, T., Cascio, C., Dobrowolski, W., 2004. Stomata open the full farm scale. Agriculture, Ecosystem & Environment 112, 146–152.
at night in pole sized and mature ponderosa pine: implications for O3 exposure Hensen, A., Nemitz, E., Flynn, M.J., Blatter, A., Jones, S.K., Sørensen, L.L., Hensen, B.,
metrics. Tree Physiology 24, 1001–1010. Pryor, S., Jensen, B., Otjes, R.P., Cobussen, J., Loubet, B., Erisman, J.W.,
Grönholm, T., Aalto, P.P., Hiltunen, V., Rannik, U., Rinne, J., Laakso, L., Hyvonen, S., Gallagher, M.W., Neftel, A., Sutton, M.A., 2008. Inter-comparison of ammonia
Vesala, T., Kulmala, M., 2007. Measurement of aerosol particle dry deposition fluxes obtained using the Relaxed Eddy Accumulation Technique. Biogeosciences
velocities using relaxed eddy accumulation technique. Tellus 59B, 381–386. Discussions 5, 3965–4000.
Grünhage, L., Jäger, H.J., 1994. Influence of the atmospheric conductivity on the Herrmann, B., Mattsson, M., Jones, S.K., Cellier, P., Milford, C., Sutton, M.A.,
ozone exposure of plants under ambient conditions: considerations for Schjoerring, J.K., Neftel, A., 2008. Vertical structure and diurnal variability of
establishing ozone standards to protect vegetation. Environmental Pollution ammonia emission potential within an intensively managed grass canopy.
85, 125–131. Biogeosciences Discussions 5, 2897–2921.
Grünhage, L., Haenel, H.-D., 2008. Detailed documentation of the PLATIN (PLant– Herrmann, H., Tilgner, A., Barzaghi, P., Majdik, Z., Gligorovski, S., Poulain, L.,
ATmosphere INteraction) model. vTI Agriculture and Forestry Reseach 319, 85 Monod, A., 2005. Towards a more detailed description of tropospheric
(Special Issue). aqueous phase organic chemistry: CAPRAM 3.0. Atmospheric Environment 39,
Guenther, A.B., Zimmerman, P.R., Harley, P.C., Monson, R.K., Fall, R., 1993. Isoprene 4351–4363.
and monoterpene emission rate variability: model evaluations and sensitivity Hertel, O., Skjøth, C.A., Lofstrøm, P., Geels, C., Frohn, L.M., Ellermann, T., Madsen, P.V.,
analyses. Journal of Geophysical Research 98, 12609–12617. 2006. Modelling nitrogen deposition on a local scale – a review of the current
Guenther, A., Hewitt, C., Erickson, D., Fall, R., Geron, C., Graedel, T., Harley, P., state of the art. Environmental Chemistry 3, 317–337.
Klinger, L., Lerdau, M., Mckay, W., Pierce, T., Scholes, R., Steinbrecher, R., Hirsch, A.I., Michalak, A.M., Bruhwiler, L.M., Peters, W., Dlugokencky, E.J., Tans, P.P.,
Tallamraju, R., Taylor, J., Zimmerman, P., 1995. A global model of natural volatile 2006. Inverse modelling estimates of the global nitrous oxide surface flux from
organic compound emissions. Journal of Geophysical Research 100, 8873–8892. 1998–2001. Global Biogeochemistry Cycles 20, GB1008.
Guenther, A., 1999. Modeling biogenic volatile organic compound emissions to the Hogg, A., Uddling, J., Ellsworth, D., Carroll, M.A., Pressley, S., Lamb, B., Vogel, C., 2007.
atmosphere. In: Hewitt, C.N. (Ed.), Reactive Hydrocarbons in the Atmosphere. Stomatal and non-stomatal fluxes of ozone to a northern mixed hardwood
Academic Press, San Diego, pp. 41–94. forest. Tellus B 59, 514–525.
Guenther, A.B., Hills, A.J., 1998. Eddy covariance measurement of isoprene fluxes. Hole, L.R., Brunner, S.H., Hanssen, J.E., Zhang, L., 2008. Low cost measurements of
Journal of Geophysical Research 103, 13145–13152. nitrogen and sulphur dry deposition velocities at a semi-alpine site: gradient
Guenther, A., Geron, C., Pierce, T., Lamb, B., Harley, P., Fall, R., 2000. Natural measurements and a comparison with deposition model estimates. Environ-
emissions of non-methane volatile organic compounds, carbon monoxide, mental Pollution 154, 473–481.
and oxides of nitrogen from North America. Atmospheric Environment 34, Holtslag, A.A.M., Moeng, C.H., 1991. Eddy diffusivity and countergradient transport
2205–2230. in the convective atmospheric boundary-layer. Journal of the Atmospheric
Güsten, H., Heinrich, G., Schmidt, R.W.H., Schurath, U., 1992. A novel sensor for Sciences 48, 1690–1698.
direct eddy flux measurements. Journal of Atmospheric Chemistry 14, 73–84. Holzke, C., Dindorf, T., Kesselmeier, J., Kuhn, U., Koppmann, R., 2006. Terpene
Guenther, A., Karl, T., Harley, P., Wiedinmyer, C., Palmer, P.I., Geron, C., 2006. emissions from European beech (Fagus sylvatica L.): pattern and emission
Estimates of global terrestrial isoprene emissions using MEGAN (Model of behaviour over two vegetation periods. Journal of Atmospheric Chemistry 55,
Emissions of Gases and Aerosols from Nature). Atmospheric Chemistry and 81–102.
Physics 6, 107–173. Hong, J., Kim, J., Lee, D., Lim, J.H. 2008. Estimation of the storage and advection
Haapanala, S., Rinne, J., Pystynen, H.-K., Hellén, H., Hakola, H., Riutta, T., 2006. effects on H2O and CO2 exchanges in a hill KoFlux forest catchment. Water
Measurements of biogenic hydrocarbon emissions from a boreal fen in the Resources Research 44, 1.
southern Finland with a REA system. Biogeosciences 3, 103–112. Hoose, C., Lohmann, U., Stier, P., Verheggen, B., Weingartner, E., 2008. Aerosol
Hakata, M., Takahashi, M., Zumft, W., Sakamoto, A., Morikawa, H., 2003. Conversion processing in mixed-phase clouds in ECHAM5-HAM: model description and
of the nitrate nitrogen and nitrogen dioxide to nitrous oxides in plants. Acta comparison to observations. Journal of Geophysical Research D: Atmospheres
Biotechnologica 23, 249–257. 113 (7), D07210.
Hakola, H., Laurila, T., Rinne, J., Puhto, K., 2000. The ambient concentrations of Horii, C.V., Munger, J.W., Wofsky, S.C., Zahniser, M., Nelson, D., McManus, J.B., 2005.
biogenic hydrocarbons at a North-European site. Atmospheric Environment 34, Atmospheric reactive nitrogen concentration and flux budgets at a North-
4971–4982. eastern U.S. forest site. Agricultural Forest Meteorology 133, 210–225.
Hakola, H., Tarvainen, V., Bäck, J., Ranta, H., Bonn, B., Rinne, J., Kulmala, M., 2006. Horvath, H., Pesava, P., Toprak, S., Aksu, R., 1996. Technique for measuring the
Seasonal variation of mono- and sesquiterpene emission rates of Scots pine. deposition velocity of particulate matter to building surfaces. Science of the
Biogeosciences 3, 93–101. Total Environment 189–190, 255–258.
Hari, P., Raivonen, M., Vesala, T., Munger, J.W., Pilegaard, K., Kulmala, M., 2003. Horváth, L., 1983. Concentration and near surface vertical flux of ammonia in the
Ultraviolet light and leaf emission of NOx. Nature 422, 134. air in Hungary. Idöjárás. Journal of the Hungarian Meteorological Service 87,
Harley, P., Greenberg, J., Niinemets, Ü., Guenther, A., 2007. Environmental controls 65–70.
over methanol emission from leaves. Biogeosciences 4, 1083–1099. Horváth, L., Asztalos, M., Führer, E., Mészáros, R., Weidinger, T., 2005. Measurement
Harmens, H., Mills, G., Emberson, L.D., Ashmore, M.R., 2007. Implications of climate of ammonia exchange over grassland in the Hungarian Great Plain. Agricultural
change for the stomatal flux of ozone: a case study for winter wheat. Envi- and Forest Meteorology 130, 282–298.
ronmental Pollution 146, 763–770. Huebert, B., Robert, C., 1985. The dry deposition of nitric acid to grass. Journal of
Harrison, R.M., Yin, J.X., Mark, D., Stedman, J., Appleby, R.S., Booker, J., Moorcroft, S., Geophysical Research 90, 2085–2090.
2001. Studies of the coarse particle (2.5–10 mm) component in UK urban Hummelshoj, P., 1994. Dry Deposition of Particles and Gases. Riso, R-658(EN). RISO
atmospheres. Atmospheric Environment 35, 3667–3679. National Laboratory, Roskilde, ISBN 87-550-1864-5.
Hatanaka, A., 1993. The biogeneration of green odour by green leaves. Phyto- Husar, R.B., Lodge, J.P., Moore, D.J. (Eds.), 1978. Sulfur in the Atmosphere. Pergamon
chemistry 34, 1201–1218. Press, New York.
Hayes, F., Mills, G., Harmens, H., Norris, D., 2007. Evidence of Widespread Ozone Husted, S., Hebbern, C.A., Mattsson, M., Schjoerring, J.K., 2000. A critical experi-
Damage to Vegetation in Europe (1990–2006). Programme Coordinating Centre mental evaluation of methods for determination of NHþ 4 in plant tissue, xylem
for the ICP Vegetation, Centre for Ecology and Hydrology, Bangor, UK. Available sap and apoplastic fluid. Physiologia Plantarum 109 (2), 167–179.
from: <icpvegetation.ceh.ac.uk>. Hüve, K., Christ, M.M., Kleist, E., Uerlings, R., Niinemets, Ü., Walter, A., Wildt, J., 2007.
Heiden, A.C., Kobel, K., Wildt, J., 1999. Einfluß verschiedener Streßfaktoren auf die Simultaneous growth and emission measurements demonstrate an interactive
Emission pflanzlicher flüchtiger organischer Verbindungen. Institut für Chemie control of methanol release by leaf expansion and stomata. Journal of Experi-
und Dynamik der Geosphäre 2: Chemie der Belasteten Atmosphäre, For- mental Botany 58, 1783–1793.
schungszentrum Jülich, Jülich. IPCC, 2006. IPCC Guidelines for National Greenhouse Gas Inventories, Prepared by
Heiden, A.C., Kobel, K., Langebartels, C., Schuh-Thomas, G., Wildt, J., 2003. Emissions the National Greenhouse Gas Inventories Programme. In: Eggleston, H.S.,
of oxygenated volatile organic compounds from plants. Part I: emissions from Buendia, L., Miwa, K., Ngara, T., Tanabe, K. (Eds.), Agriculture, Forestry and Other
lipoxygenase activity. Journal of Atmospheric Chemistry 45, 143–172. Land Use, vol. 4. IGES, Japan.
Held, A., Hinz, K.-P., Trimborn, A., Spengler, B., Klemm, O., 2003. Towards direct IPCC, 2007. Climate Change 2007 – the Physical Science Basis; Working Group I
measurement of turbulent vertical fluxes of compounds in atmospheric aerosol Contribution to the Fourth Assessment Report of the IPCC, Intergovernmental
particles. Geophysical Research Letters 30. doi:10.1029/2003GL017854. Panel on Climate Change.
5260 D. Fowler et al. / Atmospheric Environment 43 (2009) 5193–5267

Jacob, D.J., Field, B.D., Jin, E.M., Bey, I., Li, Q.B., Logan, J.A., Yantosca, R.M., Singh, H.B., Kleefeld, Ch., O’Dowd, C.D., O’Riely, S., Jennings, S.G., Aalto, P., Becker, E., Kunz, G., de
2002. Atmospheric budget of acetone. Journal of Geophysical Research – Leeuw, G., 2002. The relative scattering of sub and super micron particles to
Atmospheres 107, D10. 10.1029. aerosol light scattering in the marine boundary layer (MBL). Journal of
Jitto, P., Vinitnantarat, S., Khummongkol, P., 2007. Dry deposition velocity of sulfur Geophysical Research 107. doi:10.1029/2000JD000262.
dioxide over rice paddy in the tropical region. Atmospheric Research 85, 140–147. Kleffmann, J., Kurtenbach, R., Loerzer, J., Wiesen, P., Kalthoff, N., Vogel, B., Vogel, H.,
Jones, A.E., Weller, R., Wolff, E.W., Jacobi, H.-W., 2000. Speciation and rate of 2003. Measured and simulated vertical profiles of nitrous acid – Part I: field
photochemical NO and NO2 production in Antarctic snow. Geophysical measurements. Atmospheric Environment 37, 2949–2955.
Research Letters 27, 345–348. Kleffmann, J., Gavriloaiei, T., Hofzumahaus, A., Holland, F., Koppmann, R., Rupp, L.,
Jones, A.E., Weller, R., Anderson, P.S., Jacobi, H.W., Wolff, E.W., Schrems, O., Miller, H., Schlosser, E., Siese, M., Wahner, A., 2005. Daytime formation of nitrous acid:
2001. Measurements of NOx emissions from the Antarctic snowpack. a major source of OH radicals in a forest. Geophysical Research Letters 32,
Geophysical Research Letters 28, 1499–1502. L05818. doi: 05810.01029/02005GL022524.
Jones, M.R., Leith, I.D., Fowler, D., Raven, J.A., Sutton, M.A., Nemitz, E., Cape, J.N., Klemm, O., Wrzesinski, T., 2007. Fog deposition fluxes of water and ions to
Sheppard, L.J., Smith, R.I., Theobald, M.R., 2007. Concentration-dependent NH3 a mountainous site in Central Europe. Tellus B 59B, 705.
deposition processes for mixed moorland semi-natural vegetation. Atmospheric Kokkola, H., Sorjamaa, R., Peräniemi, A., Raatikainen, T., Laaksonen, A., 2006. Cloud
Environment 41, 2049–2060. doi:10.1016/j.atmosenv.2006.11.003. formation of particles containing humic-like substances. Geophysical Research
Junge, C.E., 1963. Air Chemistry and Radioactivity. Academic Press, New York, Letters 33, 10.
London. Kool, D.M., Wrage, N., Oenema, O., Harris, D., Van Groenigen, J.W., 2009. The 18O
Kanemasu, E.T., Flitcroft, I.D., Shah, T.D.H., Kidd, G., Nie, D., Simpson, I., Lin, M., signature of biogenic nitrous oxide is determined by O exchange with water.
Neue, H.-U., Bronson, K., 1995. Comparison of micrometeorological and Rapid communications in Mass Spectrometry 23, 104–108.
chamber measurements of methane emisisns from rice fields. In: Peng, S., Kowalski, A.S., Vong, R.J., 1999. Near-surface fluxes of cloud water evolve vertically.
Ingram, K.J., Lampe, K.J., Neue, H.-U. (Eds.), Climate Change and Rice. IRRI, Quarterly Journal of the Royal Meteorological Society 125, 2663–2684.
pp. 91–101. Kowalski, A.S., 2001. Deliquescence induces eddy covariance and estimable dry
Kärcher, B., Basko, M.M., 2004. Trapping of trace gases in growing ice crystals. deposition errors. Atmospheric Environment 35, 4843–4851.
Journal of Geophysical Research 109, D22204. doi:10.1029/2004JD005254. Kreuzwieser, J., Schnitzler, J.P., Steinbrecher, R., 1999. Biosynthesis of organic
Karl, T., Fall, R., Jordan, A., Lindinger, W., 2001a. On-line analysis of reactive VOCs compounds emitted by plants. Plant Biology 1, 149–159.
from urban lawn mowing. Environmental Science & Technology 35, 2926–2931. Kreuzwieser, J., Graus, M., Wisthaler, A., Hansel, A., Rennenberg, H., Schnitzler, J.P.,
Karl, T., Guenther, A., Jordan, A., Fall, R., Lindinger, W., 2001b. Eddy covariance 2002. Xylem-transported glucose as additional carbon source for leaf isoprene
measurement of biogenic oxygenated VOC emissions from hay harvesting. formation in Quercus robur. New Phytologist 156, 171–178.
Atmospheric Environment 35, 491–495. Kroon, P.S., Hensen, A., van den Bulk, W.C.M., Jongejan, P.A.C., Vermeulen, A.T., 2008.
Karl, T., Guenther, A., Lindinger, C., Jordan, A., Fall, R., Lindinger, W., 2001c. Eddy The importance of reducing the systematic error due to non-linearity in N2O
covariance measurements of oxygenated volatile organic compound fluxes flux measurements by static chambers. Nutrient Cycling in Agroecosystems 82,
from crop harvesting using a redesigned proton-transfer-reaction mass spec- 175–186.
trometer. Journal of Geophysical Research 106, 24157–24167. Lagzi, I., Mészáros, R., Horváth, L., Tomlin, A., Weidinger, T., Turányi, T., Ács, F.,
Karl, T., Fall, R., Rosenstiel, T.N., Prazeller, P., Larsen, B., Seufert, G., Lindinger, W., Haszpra, L., 2004. Modelling ozone fluxes over Hungary. Atmospheric Envi-
2002a. On-line analysis of the 13CO2 labeling of leaf isoprene suggests multiple ronment 38, 6211–6222.
subcellular origins of isoprene precursors. Planta 215, 894–905. Lamaud, E., Carrara, A., Brunet, Y., Lopez, A., Druilhet, A., 2002. Ozone fluxes above
Karl, T., Curtis, A.J., Rosenstiel, T.N., Monson, R.K., Fall, R., 2002b. Transient releases and within a pine forest canopy in dry and wet conditions. Atmospheric
of acetaldehyde from tree leaves – products of a pyruvate overflow mecha- Environment 36, 77–88.
nism? Plant Cell and Environment 25, 1121–1131. Lathiere, J., Hauglustaine, D.A., Friend, A.D., De Noblet-Ducoudre’, N., Viovy, N.,
Karlsson, P.E., Tang, L., Sundberg, J., Chen, D., Lindskog, A., Pleijel, H., 2007. Folberth, G.A., 2006. Impact of climate variability and land use changes on
Increasing risk for negative ozone impacts on vegetation in northern Sweden. global biogenic volatile organic compound emissions. Atmospheric Chemistry
Environmental Pollution 150, 96–106. and Physics 6, 2129–2146.
Karlsson, P.G., Karlsson, P.E., Soja, G., Vandermeiren, K., Pleijel, H., 2004. Test of the Lee, A., Goldstein, A.H., Kroll, J.H., Ng, N.L., Varutbangkul, V., Flagan, R.C.,
short-term critical levels for acute ozone injury on plants – improvements by Seinfeld, J.H., 2006. Gas-phase products and secondary aerosol yields from the
ozone uptake modelling and the use of an effect threshold. Atmospheric photooxidation of 16 different terpenes. Journal of Geophysical Research 111,
Environment 38, 2237–2245. D17305. doi:10.1029/2006JD007050.
Karnosky, D.F., Skelly, J.M., Percy, K.E., Chappelka, A.H., 2007. Perspectives regarding Lehning, A., Zimmer, W., Zimmer, I., Schnitzler, J.P., 2001. Modelling of annual
50 years of research on effects of tropospheric ozone air pollution on US forests. variations of oak (Quercus robur L.) isoprene synthase activity to predict
Environmental Pollution 147, 489–506. isoprene emission rates. Journal of Geophysical Research 106, 3157–3166.
Kawamura, K., Sakaguchi, F., 1999. Molecular distributions of water-soluble dicar- Lemon, E., Van Houtte, R., 1980. Ammonia exchange at the land surface. Agronomy
boxylic acids in marine aerosols over the Pacific Ocean including tropics. Journal 72, 876–883.
Journal of Geophysical Research 104, 3501–3509. Lerdau, M., 2007. A positive feedback with negative consequences. Science 316,
Keller, M.D., Bellows, W.K., Guillard, R.R.L., 1989. Dimethyl sulfide production in 212–213.
marine-phytoplankton. In: Saltzmann, E.S., Cooper, W.J. (Eds.), Biogenic Sulfur Leriche, M., Curier, R.L., Deguillaume, L., Caro, D., Sellegri, K., Chaumerliac, N., 2007.
in the Environment. American Chemical Society, pp. 167–182. Numerical quantification of sources and phase partitioning of chemical species
Keller, F., Bassin, S., Ammann, C., Fuhrer, J., 2007. High-resolution modelling of in cloud: application to wintertime anthropogenic air masses at the Puy de
AOT40 and stomatal ozone uptake in wheat and grassland: a comparison Dôme station. Journal of Atmospheric Chemistry 57-3, 281–297.
between 2000 and the hot summer of 2003 in Switzerland. Environmental Leroy, D., Wobrock, W., Flossmann, A.I., 2008. The role of boundary layer aerosol
Pollution 146, 671–677. particles for the development of deep convective clouds: a high-resolution 3D
Keppler, F., Hamilton, J.T.G., Braß, M., Röckmann, T., 2006. Methane emissions from model with detailed (bin) microphysics applied to CRYSTAL-FACE. Atmospheric
terrestrial plants under aerobic conditions. Nature 439, 187–191. Research. doi:10.1016/j.atmosres.2008.06.001.
Keppler, F., Hamilton, J.T.G., McRoberts, W.C., Vigano, I., Brass, M., Röckmann, T., Lewis, A.C., Hopkins, J.R., Carpenter, L.J., Stanton, J., Read, K.A., Pilling, M., 2005.
2008. Methoxyl groups of plant pectin as a precursor of atmospheric methane: Sources and sinks of acetone, methanol, and acetaldehyde in North Atlantic
evidence from deuterium labelling studies. New Phytologist 178, 808–814. marine air. Atmospheric Chemistry and Physics 5, 1963–1974.
Kerkeweg, A., Buchholz, J., Ganzeveld, L., Pozzer, A., Tost, H., Jöckel, P., 2006. Tech- Lewis, E.R., Schwartz, S.E., 2004. Sea Salt Aerosol Production: Mechanisms,
nical note: an implementation of the dry removal processes DRY DEPosition Methods, Measurements, and Models: A Critical Review. American Geophysical
and SEDImentation in the Modular Earth Submodel System (MESSy). Atmo- Union, Washington, DC.
spheric Chemistry and Physics 6, 4617–4632. Lewis, E.R., Schwartz, S.E., 2005. Sea Salt Aerosol Production. Mechanisms, Methods,
Keronen, P., Reissell, A., Rannik, Ü., Pohja, T., Siivola, E., Hiltunen, V., Hari, P., Measurements, and Models. American Geophysical Union, Washington, D.C. USA.
Kulmala, M., Vesala, T., 2003. Ozone flux measurements over a Scots pine forest Li, C., Aber, J., Stange, F., Butterbach-Bahl, K., Papen, H., 2000. A process oriented
using eddy covariance method: performance evaluation and comparison with model of N2O and NO emissions from forest soils: 1, model development.
flux–profile method. Boreal Environmental Research 8, 425–443. Journal of Geophysical Research 105, 4369–4384.
Kesik, M., Brüggemann, N., Forkel, R., Knoche, R., Li, C., Seufert, G., Simpson, D., Li, C., Salas, W., DeAngelo, B., Rose, S., 2006. Assessing alternatives for mitigating net
Butterbach-Bahl, K., 2006. Future scenarios of N2O and NO emissions from greenhouse gas emissions and increasing yields from rice production in China
European forest soils. Journal of Geophysical Research 111, G02018. doi:10.1029/ over the next twenty years. Journal of Environmental Quality 35, 1554–1565.
2005JG000115. Li, D., Wang, X., 2008. Nitrogen isotopic signature of soil-released nitric oxide (NO)
Kiene, R.P., Bates, T.S., 1990. Biological removal of dimethyl sulphide from seawater. after fertilizer application. Atmospheric Environment 42, 4747–4754.
Nature 345, 702–705. Lichtenthaler, H.K., 1999. The 1-deoxy-D-xylulose-5-phosphate pathway of iso-
King, J.S., Kubiske, M.E., Pregitzer, K.S., Hendry, G.R., Giardina, C.P., Quinn, V.S., prenoid biosynthesis in plants. Annual Review of Plant Physiology and Plant
Karnosky, D.F., 2005. Tropospheric O3 compromises net primary production in Molecular Biology 50, 47–65.
young stands of trembling aspen, paper birch and sugar maple in response to Lindinger, W., Hansel, A., Jordan, A., 1998. Review: on-line monitoring of volatile
elated CO2. New Phytologist 168, 623–636. organic compounds at pptv levels by means of Proton-Transfer-Reaction Mass
Kitzler, B., Zechmeister Boltenstern, S., Holtermann, C., Skiba, U., Butterbach- Spectrometry (PTR-MS), Medical applications, food control and environmental
Bahl, K., 2006. Controls over N2O, NOx and CO2 fluxes in a calcareous mountain research. International Journal of Mass Spectrometry and Ion Processes 173,
forest soil. Biogeosciences 3, 383–395. 191–241.
D. Fowler et al. / Atmospheric Environment 43 (2009) 5193–5267 5261

Liss, P.S., Balls, P.W., Martinelli, F.N., Coantic, M., 1981. The effect of evaporation and McDonald, A.G., Bealey, W.J., Fowler, D., Dragosits, U., Skiba, U., Smith, R.I.,
condensation on gas transfer across an air–water interface. Oceanologica Acta 4,129. Donovan, R.G., Brett, H.E., Hewitt, C.N., Nemitz, E., 2007. Quantifying the effect
Liss, P.S., Hatton, A.D., Malin, G., Nightingale, P.D., Turner, S.M., 1997. Marine of urban tree planting on concentrations and depositions of PM10 in two UK
sulphur emissions. Philosophical Transactions of the Royal Society of London conurbations. Atmospheric Environment 41 (38), 8455–8467.
Series B – Biological Sciences 352, 159–168. McFiggans, G., Coe, H., Burgess, R., Allan, J., Cubison, M., Alfarra, M.R., Saunders, R.,
Loreto, F., Ciccioli, P., Brancaleoni, E., Valentini, R., De Lillis, M., Csiky, O., Seufert, G., 1998. Saiz-Lopez, A., Plane, J.M.C., Wevill, D.J., Carpenter, L.J., Rickard, A.R., Monks, P.S.,
A hypothesis on the evolution of isoprenoid emission by oaks based on the corre- 2004. Atmospheric Chemistry and Physics 4, 701–713.
lation between emission type and Quercus taxonomy. Oecologia 115, 302–305. McLaughlin, S.B., Wullschleger, S.D., Sun, G., Nosal, M., 2007. Interactive effects of
Loreto, F., Ciccioli, P., Brancaleoni, E., Frattoni, M., Delfine, S., 2000. Incomplete 13C ozone and climate on water use, soil moisture content and streamflow in
labeling of alpha-pinene content in Quercus ilex leaves and appearance of unlabeled a southern Appalachian forest in the USA. New Phytologist 174, 125–136.
C in alpha-pinene emission in the dark. Plant Cell and Environment 23, 229–234. McLeod, A.R., Fry, S.C., Loake, G.J., Messenger, D.J., Reay, D.S., Smith, K.A., Yun, B.-W.,
Loreto, F., Fischbach, R.J., Schnitzler, J.-P., Ciccioli, P., Brancaleoni, E., Calfapietra, C., 2008. Ultraviolet radiation drives methane emissions from terrestrial plant
Seufert, G., 2001. Monoterpene emission and monoterpene synthase activities pectins. New Phytologist 9999 (9999).
in the Mediterranean evergreen oak Quercus ilex L. grown at elevated CO2 Meixner, F.X., Ammann, C., Arenz, R., Nathaus, F.J., Neftel, A., Blatter, A., Gut, A.,
concentrations. Global Change Biology 7, 709–717. Hoegger, D., Wyers, G.P., Jongejan, P., Loesch, R., Busch, J., Guesten, H., Sprung, D.,
Loreto, F., Pinelli, P., Brancaleoni, E., Ciccioli, P., 2004. 13C labelling reveals chloro- Baumann, M., 1996. The Bellheim ’95 experiment: bi-directional fluxes of
plastic and extra-chloroplastic pools of dimethylallyl pyrophosphate and their ammonia. Annales Geophysicae 14 (Suppl. II), C467.
contribution to isoprene formation. Plant Physiology 135, 1903–1907. Meskhidze, N., Nenes, A., 2006. Phytoplankton and cloudiness in the Southern
Loreto, F., Barta, C., Brilli, F., Nogues, I., 2006. On the induction of volatile organic Ocean Science. 314 (5804), 1419–1423.
compound emissions by plants as consequence of wounding or fluctuations of Meyers, T.P., Finkelstein, P.L., Clarke, J., Ellestad, T.G., Sims, P., 1998. A multilayer
light and temperature. Plant Cell and Environment 29, 1820–1828. model for inferring dry deposition using standard meteorological measure-
Loubet, B., Milford, C., Sutton, M.A., Cellier, P., 2001. Investigation of the interaction ments. Journal of Geophysical Research 103, 22645–22661.
between sources and sinks of atmospheric ammonia in an upland landscape Meyers, T.P., Luke, W.T., Meisinger, J.J., 2006. Fluxes of ammonia and sulfate over
using a simplified dispersion-exchange model. Journal of Geophysical Research maize using relaxed eddy accummulation. Agricultural Forest Meteorology 136,
106 (D20), 24183–24196. 203–213.
Loubet, B., Cellier, P., Milford, C., Sutton, M.A., 2006. A coupled dispersion and Michou, M., Laville, P., Serça, D., Fotiadi, A., Bouchou, P., Peuch, V.-H., 2005.
exchange model for short-range dry deposition of atmospheric ammonia. Measured and modeled dry deposition velocities over the ESCOMPTE area.
Quarterly Journal of the Royal Meteorological Society 132, 1733–1763. Atmospheric Research 74, 89–116.
Loubet, B., Asman, W.A.H., Theobald, M.R., Hertel, O., Tang, Y.S., Robin, P., Mikkelsen, T.N., Ro-Poulsen, H., Hovmand, M.F., Jensen, N.O., Pilegaard, K.,
Hassouna, M., Daemmgen, U., Genermont, S., Cellier, P., Sutton, M.A., 2008. Egeløv, A.H., 2004. Five-year measurements of ozone fluxes to a Danish Norway
Ammonia deposition near hot spots: processes, models and monitoring spruce canopy. Atmospheric Environment 38, 2361–2371.
methods. In: Sutton, M.A., Baker, S., Reis, S. (Eds.), Atmospheric Ammonia: Milford, C., Theobald, M.R., Nemitz, E., Sutton, M.A., 2001a. Dynamics of ammonia
Detecting Emission Changes and Environmental Impacts. Springer. exchange in response to cutting and fertilizing in an intensively-managed
Loubet, B., Milford, C., Hensen, A., Dämmgen, U., Cellier, P., Sutton, M.A., 2009. grassland. Water, Air and Soil Pollution: Focus 1, 167–176.
Advection of ammonia over a pasture field, and its effect on gradient flux Milford, C., Hargreaves, K.J., Sutton, M.A., Loubet, B., Cellier, P., 2001b. Fluxes of NH3
measurements. Biogeosciences Discussions 6, 163–196. and CO2 over upland moorland in the vicinity of agricultural land. Journal of
Lovett, G.M., 1994. Atmospheric deposition of nutrients and pollutants to North Geophysical Research (Atmospheres) 106, 24169–24181.
America: an ecological perspective. Ecological Applications 4, 629–650. Milford, C., Theobald, M.R., Nemitz, E., Hargreaves, K.J., Horvath, L., Raso, J.,
Loya, W.M., Pregitzer, K.S., Karberg, N.J., King, J.S., Giardina, C.P., 2003. Reduction of Dämmgen, U., Neftel, A., Jones, S.K., Hensen, A., Loubet, B., Sutton, M.A., 2009.
soil carbon formation by tropospheric ozone under increased carbon dioxide Ammonia fluxes in relation to cutting and fertilization of an intensively
levels. Nature 425, 705–707. managed grassland derived from an inter-comparison of gradient measure-
Marandino, C.A., De Bruyn, W.J., Miller, S.D., Prather, M.J., Saltzman, E.S., 2005. ments. Biogeosciences Discussions, 6, 819–834.
Oceanic uptake and the global atmospheric acetone budget. Geophysical Millenium Ecosystem Assessment, 2005. Ecosystems and Human Well-being:
Research Letters 32 (15), L15806. Current State and Trends. Island Press.
Mari, C., Jacob, D.J., Bechtold, P., 2000. Transport and scavenging of soluble gases in Millet, D.B., Jacob, D.J., Turquety, S., Hudman, R.C., Wu, S., Fried, A., Walega, J.,
a deep convective cloud. Journal of Geophysical Research 105, 22255–22267. Heikes, B.G., Blake, D.R., Singh, H.B., Anderson, B.E., Clarke, A.D., 2006. Form-
Martensson, E.M., Nilsson, E.D., Johansson, C., 2002. Emissions of aerosol particles in aldehyde distribution over North America: implications for satellite retrievals of
the urban environment measured by eddy correlation. In: Proceedings of the formaldehyde columns and isoprene emission. Journal of Geophysical Research
6th International Aerosol Conference, Taiwan, 9–13 September 2002. 111, D24S02. doi:10.1029/2005JD006853.
Mårtensson, M., Nilsson, E.D., de Leeuw, G., Cohen, L.H., Hansson, H.-C., 2003. Millet, D.B., Jacob, D.J., Custer, T.G., de Gouw, J.A., Goldstein, A.H., Karl, T., Singh, H.B.,
Laboratory simulations of the primary marine aerosol generated by bubble Sive, B.C., Talbot, R.W., Warneke, C., Williams, J., 2008. New constraints on
bursting. JGR-Atmospheres 108 (D9). doi:10.1029/2002JD002263. terrestrial and oceanic sources of atmospheric methanol. Atmospheric Chem-
Martensson, E.M., Nilsson, E.D., Buzorius, G., Johansson, C., 2006. Eddy covariance istry and Physics Discussions 8, 7609–7655.
measurements and parameterisaton of traffic related emissions in an urban Möhler, O., Bunz, H., Stetzer, O., 2006. Homogeneous nucleation rates of nitric acid
environment. Atmospheric Chemistry and Physics 6, 769–785. dihydrate (NAD) at simulated stratospheric conditions – Part II: modelling.
Martin, C.L., Longley, I.D., Dorsey, J.R., Thomas, R.M., Gallagher, M.W., Nemitz, E. Atmospheric Chemistry and Physics 6 (10), 3035–3047.
Ultrafine particle fluxes above four major European cities. Atmospheric Envi- Möhler, O., DeMott, P.J., Vali, G., Levin, Z., 2007. Microbiology and atmospheric
ronment, in press. processes: the role of biological particles in cloud physics. Biogeosciences
Massad, R.S., Loubet, B., Tuzet, A., Cellier, P., 2008. Relationship between ammonia Discuss. 4, 2559–2591. Available at: http://www.biogeosciences-discuss.net/4/
stomatal compensation point and nitrogen metabolism in arable crops: current 2559/2007/.
status of knowledge and potential modelling approaches. Environmental Monin, A.S., Obukhov, A.M., 1954. ‘Basic laws of turbulent mixing in the ground
Pollution 154, 390–403. layer of the atmosphere. Transactions of the Geophysical Institute. Akademy
Massman, W.J., 2004. Toward an ozone standard to protect vegetation based on Nauk USSR 151, 163–187.
effective dose: a review of deposition resistances and a possible metric. Monin, A.S., Zilitinkevich, S., 1974. Similarity theory and resistance laws for the
Atmospheric Environment 38, 2323–2337. planetary layer. Boundary-Layer Meteorology 7, 391–397.
Matsuda, K., Aoki, M., Zhang, S., Kominami, T., Fukuyama, T., Fukuzaki, N., Totsuka, T., Monks, P.S., Granier, C., Fuzzi, S., Stohl, A., Williams, M., Akimoto, H., Amman, M.,
2002. Dry deposition velocity of sulfur dioxide on a red pine forest in Nagano, Baklanov, A., Baltensperger, U., Bey, I., Blake, N., Blake, R.S., Carslaw, K., Cooper,
Japan. Journal of Japan Society for Atmospheric Environment 37 (6), 387–392. O.R., Dentener, F., Fowler, D., Fragkou, E., Frost, G., Generoso, S., Ginoux, P.,
Matsuda, K., Watanabe, I., Wingpud, V., Theramongkol, P., Ohizumi, T., 2006. Grewe, V., Guenther, A., Hansson, H.C., Henne, S., Hjorth, J., Hofzumahaus, A.,
Deposition velocity of O3 and SO2 in the dry and wet season above a tropical Huntrieser, H., Isaksen, I.S.A., Jenkin, M.E., Kaiser, J., Kanakidou, M., Klimont, Z.,
forest in Northern Thailand. Atmospheric Environment 40, 7557–7564. Kulmala, M., Laj, P., Lawrence, M.G., Lee, J.D., Liousse, C., Maione, M., McFiggans,
Mattsson, M., Herrmann, B., David, M., Loubet, B., Riedo, M., Theobald, M.R., G., Metzger, A., Mieville, A., Moussiopoulos, N., Orlando, J.J., O’Dowd, C., Palmer,
Sutton, M.A., Bruhn, D., Neftel, A., Schjoerring, J.K., 2008a. Temporal variability in P.I., Parrish, D.D., Petzold, A., Platt, U., Poeschl, U., Prévôt, A.S.H., Reeves, C.E.,
bioassays of ammonia emission potential in relation to plant and soil N parame- Reimann, S., Rudich, Y., Sellegri, K., Steinbrecher, R., Simpson, D., ten Brink, H.,
ters in intensively managed grassland. Biogeosciences Discussions 5, 2749–2772. Theloke, J., van der Werf, G.R., Vautard, R., Vestreng, V, Vlachokostas, Ch.,
Mattsson, M., Herrmann, B., Jones, S.K., Neftel, A., Sutton, M.A., Schjoerring, J.K., 2008b. vonGlasow, R. Atmospheric composition change – global and regional air
Contribution of different grass species to plant–atmosphere ammonia exchange quality (Accent Special Issue) 43 (33), 5264–5344.
in intensively managed grassland. Biogeosciences Discussions 5, 2583–2605. Monson, R.K., Harley, P.C., Litvak, M.E., Wildermuth, M., Guenther, A.B., Zimmerman, P.R.,
Matyssek, R., Sandermann, H., Wieser, G., Booker, F., Cieslik, S., Musselman, R., Fall, R., 1994. Environmental and developmental controls over the seasonal pattern
Ernst, D., 2008. The challenge of making ozone risk assessment for forest trees of isoprene emission from aspen leaves. Oecologia 99, 260–270.
more mechanistic. Environmental Pollution 156 (3), 567–582. doi:10.1016/ Monteith, J.L., Unsworth, M.H., 2007. Principles of Environmental Physics. Academic
j.envpol.2008.04.017. Press.
McBain, M.C., Warland, J.S., McBride, R.A., Wagner-Riddle, C., 2004. Laboratory-scale Mosier, A.R., Duxbury, J.M., Freney, J.R., Heinemyer, O., Minami, K., 1998. Assessing
measurements of N2O and CH4 emissions from hybrid poplars (Populus deltoides and mitigating N2O emissions from agricultural soils. Climate Change 40,
x Populus nigra). Waste Management Research 22, 454–465. 1573–1580.
5262 D. Fowler et al. / Atmospheric Environment 43 (2009) 5193–5267

Mosquera, J., Hensen, A., van den Bulk, W.C.M., Vermeulen, A.T., Erisman, J.W., 2001. Nemitz, E., Sutton, M.A., 2004. Gas-particle conversions above a Dutch heathland:
Long term NH3 flux measurements at two locations in The Netherlands. Water, III. Modelling of size-dependent NHþ 4 fluxes as modified by the NH3–HNO3–
Air and Soil Pollution: Focus 1, 203–212. NH4NO3 equilibrium. Atmospheric Chemistry and Physics 4, 1025–1045.
Moukhtar, S., Bessagnet, B., Rouil, L., Simon, V., 2005. Monoterpene emissions from Nemitz, E., Sutton, M.A., Wyers, G.P., Jongejan, P.A.C., 2004a. Gas-particle conver-
beech (Fagus sylvatica) in a French forest and impact on secondary pollutants sions above a Dutch heathland: I. Surface exchange fluxes of NH3, SO2, HNO3
formation at regional scale. Atmospheric Environment 39, 3535–3547. and HCl. Atmospheric Chemistry and Physics 4, 989–1005.
Muhlbauer, A., Lohmann, U., 2008. Sensitivity studies of the role of aerosols in Nemitz, E., Sutton, M.A., Wyers, G.P., Otjes, R.P., Mennen, M.G., van Putten, E.M.,
warm-phase orographic precipitation in different dynamical flow regimes. Gallagher, M.W., 2004b. Gas-particle interactions above a Dutch heathland: II.
Journal of Atmospheric Science 65, 2522–2542. Concentrations and surface exchange fluxes of atmospheric particles. Atmo-
Mukammal, E.I., 1965. Ozone as a cause of tobacco injury. Agricultural Meteorology spheric Chemistry and Physics 4, 1007–1024.
2, 145–165. Nemitz, E., Jimenez, J.L., Huffman, A., Canagaratna, M., Worsnop, D.R., Guenther, A.B.,
Mulcahy, J.P., O’Dowd, C.D., Jennings, S.G., Ceburnis, D., 2008. Significant enhance- 2008. An eddy covariance system for the measurement of surface/atmosphere
ment of aerosol optical depth in marine air under high wind conditions. exchange fluxes of submicron aerosol chemical species – first application above an
Geophysical Research Letters 35, L16810. doi:10.1029/2008GL034303. urban area. Aerosol Science Technology 42 (8), 636–657.
Müller, C., 2003. Plants affect the in situ N2O emissions of a temperate grassland Nemitz, E., Dorsey, J.R., Flynn, M., Gallagher, M.W., Hensen, A., Erisman, J.W.,
ecosystem. Journal of Plant Nutrition and Soil Science – Zeitschrift Für Pflan- Owen, S., Daemmgen, U., Sutton, M.A., 2009b. Aerosol fluxes and particle
zenernährung Und Bodenkunde 166, 771–773. growth above grassland following the application of NH4NO3 fertilizers. Bio-
Neff, J.C., Holland, E.A., Dentener, F.J., McDowell, W.H., Russell, K.M., 2002. The geosciences Discussions 6 (1), 341–389.
origin, composition and rates of organic nitrogen deposition: a missing piece of Neu, J.L., Lawler, M.J., Prather, M.J., Saltzman, E.S., 2008. Oceanic alkyl nitrates as
the nitrogen cycle? Biogeochemistry 57, 99–136. a natural source of tropospheric ozone. Geophysical Research Letters 35,
Neftel, A., Blatter, A., Gut, A., Hoegger, D., Meixner, F.X., Ammann, C., Nathaus, F.J., L13814. doi:10.1029/2008GL034189.
1998. NH3 soil and soil surface gas measurements in a triticale wheat field. Neumann, H.H., den Hartog, G., 1985. Eddy correlation measurements of atmo-
Atmospheric Environment (Ammonia Special Issue) 32 (3), 499–506. spheric fluxes of ozone, sulphur, and particulates during the champaign inter-
Neftel, A., Blatter, A., Otjes, R., Erisman, J.W., Hansen, A., 1999. State of the art REA comparison study. Journal of Geophysical Research 90, 2097–2110.
NH3 flux measurements. In: Proc. 10th Nitrogen Workshop, Copenhagen, Nicholson, K.W., Davies, T.D., 1987. Field measurements of the dry deposition of
August 1999. II.49. Royal Veterinary and Agricultural University, Copenhagen. particulate sulphate. Atmospheric Environment 21, 1561–1571.
Neftel, A., Ammann, C., Brunner, A., Jocher, M., Spirig, M., Davidson, B., Rinne, J., Nicholson, K.W., 1988. A review of particle resuspension. Atmospheric Environment
Dommen, J., 2007. Comparison of different approaches to measure VOC exchange 22, 2639–2651.
over grassland by PTR-MS. In: Hansel, A., Märk, T. (Eds.), 3rd International Nicholson, K.W., Branson, J.R., Giess, P., Cannell, R.J., 1989. The effect of vehicle
Conference on Proton Transfer Reaction Mass Spectrometry and Its Applications – activity on particle resuspension. Journal of Aerosol Science 20, 1425–1428.
Contributions. January 27th–February 1st, 2007, Obergurgl, Austria. Conference Nicholson, K.W., 1993. Wind-tunnel experiments on the resuspension of particulate
Series, Innsbruck University Press, pp. 80–84. material. Atmospheric Environment Part A – General Topics 27, 181–188.
Neinhuis, C., Barthlott, W., 1997. Characterization and distribution of water-repel- Niinemets, Ü., Tenhunen, J.D., Harley, P.C., Steinbrecher, R., 1999. A model of
lent, self-cleaning plant surfaces. Annals of Botany 79, 667–677. isoprene emission based on energetic requirements for isoprene synthesis and
Neirynck, J., Kowalski, A.S., Carrara, A., Ceulemans, R., 2005. Driving forces for leaf photosynthetic properties for Liquidambar and Quercus. Plant Cell and
ammonia fluxes over mixed forest subjected to high deposition loads. Atmo- Environment 22, 1319–1336.
spheric Environment 39, 5013–5024. Niinemets, Ü., Reichstein, M., 2002. A model analysis of the effects of nonspecific
Neirynck, J., Ceulemans, R. 2008. Bidirectional ammonia exchange above a mixed monoterpenoid storage in leaf tissues on emission kinetics and composition in
coniferous forest. Reduced Nitrogen in Ecology and the Environment 154, 424–438. Mediterranean sclerophyllous Quercus species. Global Biogeochemical Cycles 16
Nemecek-Marshall, M., MacDonald, M.C., Franzen, J.J., Wojciechowski, C.L., Fall, R., (1110) DOI: 1110.1029/2002GB001927.
1995. Methanol emission from leaves: enzymatic detection of gas-phase Niinemets, Ü., Seufert, G., Steinbrecher, R., Tenhunen, J.D., 2002a. A model coupling
methanol and relation of methanol fluxes to stomatal conductance and leaf foliar monoterpene emissions to leaf photosynthetic characteristics in Medi-
development. Plant Physiology 108, 1359–1368. terranean evergreen Quercus species. New Phytologist 153, 257–276.
Nemitz, E., Sutton, M.A., Fowler, D., Choularton, T., 1996. Application of a NH3 gas-to- Niinemets, Ü., Reichstein, M., Staudt, M., Seufert, G., Tenhunen, J.D., 2002b. Stomatal
particle conversion model to measurement data. In: Sutton, M.A., Lee, D.S., constraints may affect emission of oxygenated monoterpenoids from the foliage
Dollard, G.J., Fowler, D. (Eds.), Poster Proceedings of Atmospheric Ammonia: of Pinus pinea. Plant Physiology 130, 1371–1385.
Emission, Deposition and Environmental Impacts. Institute of Terrestrial Ecology, Niinemets, Ü., Peñuelas, J., 2008. Gardening and urban landscaping: significant
Edinburgh, pp. 98–103. players in global change. Trends in Plant Science 13, 60–65.
Nemitz, E., Sutton, M.A., Gut, A., San José, R., Husted, S., Schjoerring, J.K., 2000a. Niinemets, Ü, 2008. Getting hold of terpene emissions from vegetation. ILEAPS
Sources and sinks of ammonia within an oilseed rape canopy. Agricultural and Newsletter 5, 40–42.
Forest Meteorology (Ammonia Special Issue) 105 (4), 385–404. Nilsson, E.D., Rannik, U., 2001. Turbulent aerosol fluxes over the Arctic Ocean 1. Dry
Nemitz, E., Theobald, M.R., McDonald, A.G., Fowler, D., Dorsey, J.R., Bower, K.N., deposition over sea and pack ice. Journal of Geophysical Research – Atmo-
Gallagher, M.W., 2000b. Direct micrometeorological eddy-correlation measure- spheres 106 (D23), 32125–32137.
ments of size-dependent particle emission above a city. In: Midgley, P.M., Reu- Nilsson, E.D., Rannik, U., Swietlicki, E., Leck, C., Aalto, P.P., Zhou, J., Norman, M., 2001.
ther, M., Williams, M. (Eds.), EUROTRAC Symposium 2000. Springer Verlag, Journal of Geophysical Research 106 (D23), 32139–32154.
Heidelberg, Garmisch-Parenkirchen. Noe, S.M., Ciccioli, P., Brancaleoni, E., Loreto, F., Niinemets, Ü., 2006. Emissions of
Nemitz, E., Sutton, M.A., Schjoerring, J.K., Husted, S., Wyers, G.P., 2000c. Resistance monoterpenes linalool and ocimene respond differently to environmental
modelling of ammonia exchange over oilseed rape. Agriculture and Forest changes due to differences in physico-chemical characteristics. Atmospheric
Meteorology (Ammonia Special Issue) 105 (4), 405–425. Environment 40, 4649–4662.
Nemitz, E., Williams, P.I., Theobald, M.R., McDonald, A.G., Fowler, D., Gallagher, M.W., Noe, S.M., Copolovici, L., Niinemets, Ü., Vaino, E., 2008a. Foliar limonene uptake
2000d. Application of two micrometeorological techniques to derive fluxes of scales positively with leaf lipid content: ‘‘non-emitting’’ species absorb and
aerosol components above a city. In: Midgley, P.M., Reuther, M., Williams, M. release monoterpenes. Plant Biology 10, 129–137.
(Eds.), EUROTRAC-2 Symposium 2000. Springer Verlag, Berlin, Garmisch- Noe, S.M., Peñuelas, J., Niinemets, Ü, 2008b. Monoterpene emissions from orna-
Partenkirchen, Germany. mental trees in urban areas: a case study of Barcelona, Spain. Plant Biology 10,
Nemitz, E., Sutton, M.A., Wyers, G.P., Otjes, R.P., Schjoerring, J.K., Gallagher, M.W., 163–169.
Parrington, J., Fowler, D., Choularton, T.W., 2000e. Surface/atmosphere Norris, S., Brooks, I., de Leeuw, G., Smith, M.H., Moeman, M., Lingard, J., 2007. Eddy
exchange and chemical interaction of gases and aerosols over oilseed rape. covariance measurements of sea spray particles over the Atlantic Ocean.
Agricultural and Forest Meteorology 105, 427–445. Atmospheric Chemistry and Physics Discussion 7, 13243–13269.
Nemitz, E., Fowler, D., Dorsey, J.R., Theobald, M.R., McDonald, A.D., Bower, K.N., Novak, J., Pierce, T., 1993. Natural emissions of oxidant precursors, Water. Air, and
Beswick, K.M., Williams, P.I., Gallagher, M.W., 2000f. Direct measurements of Soil Pollution 67, 57–77.
size-segregated particle fluxes above a city. Journal of Aerosol Science 31, O’Dowd, C.D., Lowe, J., Smith, M.H., Kaye, A.D., 1999. The relative importance of sea-
116–117. salt and nss-sulphate aerosol to the marine CCN population: an improved
Nemitz, E., Flynn, M., Williams, P.I., Milford, C., Theobald, M.R., Blatter, A., multi-component aerosol-droplet parameterisation. Quarterly Journal of the
Gallagher, M.W., Sutton, M.A., 2001a. A relaxed eddy accumulation system for Royal Meteorological Society 125, 1295–1313.
the automated measurement of atmospheric ammonia fluxes. Water, Air and O’Dowd, C.D., Jimenez, J.L., Bahreini, R., Flagan, R.C., Seinfeld, J.H., Pirjola, L.,
Soil Pollution: Focus 1, 189–202. Kulmala, M., Jennings, S.F.G., Hoffmann, T., 2002. Marine particle formation
Nemitz, E., Milford, C., Sutton, M.A., 2001b. A two-layer canopy compensation point from biogenic iodine emissions. Nature 417, 632–636.
model for describing bi-directional biosphere–atmosphere exchange of O’Dowd, C.D., Facchini, M.C., Cavalli, F., Ceburnis, D., Mircea, M., Decesari, S.,
ammonia. Quarterly Journal of the Royal Meteorological Society 127, 815–833. Fuzzi, S., Yoon, Y.J., Putaud, J.P., 2004. Biogenically-driven organic contribution
Nemitz, E., Gallagher, M.W., Duyzer, J.H., Fowler, D., 2002a. Micrometeorological to marine aerosol. Nature. doi:10.1038/nature02959.
measurements of particle deposition velocities to moorland vegetation. Quar- O’Dowd, C.D., Hoffmann, T., 2005. Coastal new particle formation: a review of the current
terly Journal of the Royal Meteorological Society 128, 2281–2300. state-of-the-art. Environmental Chemistry 2, 245–255. doi:10.1071/EN05077.
Nemitz, E., Hargreaves, K.J., McDonald, A.G., Dorsey, J.R., Fowler, D., 2002b. Micro- O’Dowd, C.D., de Leeuw, G., 2007. Marine aerosol production: a review of the
meteorological measurements of the urban heat budget and CO2 emissions on current knowledge. Philosophical Transactions of the Royal Society.
a city scale. Environmental Science & Technology 36, 3139–3146. doi:10.1098/rsta.2007.2043.
D. Fowler et al. / Atmospheric Environment 43 (2009) 5193–5267 5263

O’Dowd, C.D., Langmann, B., Varghese, S., Scannell, C., Ceburnis, D., Facchini, M.C., flux and flux–response relationships for European wheat and potato. Atmo-
2008. A combined organic-inorganic sea-spray source function. Geophysical spheric Environment 41, 3022–3040.
Research Letters 35, L01801. doi:10.1029/2007GL030331. Polson, D. Boundary-layer-budget measurements of greenhouse gas emissions at
Okin, G.S., 2005. Dependence of wind erosion and dust emission on surface the country scale indicate national inventories underestimate nitrous oxide and
heterogeneity: stochastic modelling. Journal of Geophysical Research – Atmo- methane. Quarterly Journal of Royal Meteorological Society, submitted for
spheres 110. doi:10.1029/2004JD005288. publication.
Ould-Dada, Z., 2002. Dry deposition profile of small particle within a model spruce Pozzoli, L., Bey, I., Rast, S., Schultz, M.G., Stier, P., Feichter, J., 2008. Trace gas and
canopy. Science of the Total Environment 286, 83–96. aerosol interactions in the fully coupled model of aerosol-chemistry-climate
Owen, S., Boissard, S., Street, R.A., Duckham, S.C., Csiky, O., Hewitt, N., 1997. ECHAM5-HAMMOZ: 2. Impact of heterogeneous chemistry on the global
Screening of 18 Mediterranean plant species for volatile organic compound aerosol distributions. 113 (7), D07309.
emission. Atmospheric Environment 31, 101–117. Pruppacher, H.R., Klett, J.D., 1997. Microphysics of Clouds and Precipitation, Second
Owen, S.M., MacKenzie, A.R., Stewart, H., Donovan, R., Hewitt, C.N., 2003. Biogenic Revised and Enlarged Edition. Kluwer Academic Publisher, 953 pp.
volatile organic compound (VOC) emission estimates from an urban tree Pryor, S.C., Soerensen, L.L., 2000. Nitric acid–sea salt reactions: implication for
canopy. Ecological Applications 13, 927–938. nitrogen deposition to wate surfaces. Journal of Applied Meteorology 39,
Palmer, P.I., Jacob, D.J., Fiore, A.M., Martin, R.V., Chance, K., Kurosu, T.P., 2003. Mapping 725–731.
isoprene emissions over North America using formaldehyde column observations Pryor, S.C., Binkowski, F.S., 2004. An analysis of the time scales assocaited with
from space. Journal of Geophysical Research. doi:10.1029/2002JD002153. aerosol processes during dry deposition. Aerosol Science Technology 38,
Palmer, P.I., Abbot, D.S., Fu, T.-M., Jacob, D.J., Chance, K., Kurosu, T., Guenther, A., 1091–1098.
Wiedinmyer, C., Stanton, J., Pilling, M., Pressley, S., Lamb, B., Sumner, A.L., 2006. Pryor, S.C., Klemm, O., 2004. Experimentally derived estimates of nitric acid dry
Quantifying the seasonal and interannual variability of North American deposition velocity and viscous sub-layer resistance at a conifer forest. Atmo-
isoprene emissions using satellite observations of formaldehyde column. Jour- spheric Environment 38, 2769–2777.
nal of Geophysical Research. doi:10.1029/2005JD006689. Pryor, S.C., 2006. Size resolved particle deposition velocities of sub-100 nm diam-
Palmer, P.I., Barkley, M.P., Kurosu, T.P., Lewis, A.C., Saxton, J.E., Chance, K., Gatti, L.V., eter particles over a forest. Atmospheric Environment 40, 6192–6200.
2007. Interpreting satellite column observations of formaldehyde over tropical Pryor, S.C., Larsen, S.E., Sorensen, L.L., Barthelmie, R.J., Gronholm, T., Kulmala, M.,
South America. Philosophical Transactions of the Royal Society. A. 365, 1741–1751. Lauiainen, S., Rannik, Ü, Vesala, T., 2007. Particle fluxes over forests: analyses of
Palmer, P.I., 2008. Quantifying sources and sinks of trace gases using space-borne flux methods and functional dependencies. Journal of Geophysical Research 112
measurements: current and future science. Philosophical Transaction of the (D07205). doi:10.1029/2006JD008066.
Royal Society A – 2008 Triennial Issues, 366, 4509–4528. Pryor, S.C., Barthelmie, R.J., Larsen, S.E., Sorensen, L.L., Sempreviva, A.M.,
Parton, W.J., Morgan, J.A., Wang, G.M., Del Grosso, S., 2007. Projected ecosystem Groenholm, T., Kulmala, M., Rannik, Ü, Vesala, T., 2008a. Upward particle
impact of the Prairie Heating and CO2 enrichment experiment. New Phytologist number fluxes over forest: where, when, why? Tellus B 60, 372–380.
174, 823–834. Pryor, S.C., Gallagher, M.W., Sievering, H., Larsen, S.E., Barthelmie, R.J., Birsan, F.,
Pesava, P., Aksu, R., Toprak, S., Horvath, H., Seidl, S., 1999. Dry deposition of particles Nemitz, E., Rinne, J., Kulmala, M., Groenholm, T., Taipale, R., Vesala, T., 2008b. A
to building surfaces and soiling. Science of the Total Environment 235, 25–35. review of measurement and modelling results of particle atmosphere–surface
Petroff, A., Mailliat, A., Amielh, M., Anselmet, F., 2007a. Aerosol dry deposition exchange. Tellus B 60, 42–75.
on vegetative canopies. Part I: review of present knowledge. Atmospheric Pumpanen, J., Kolari, P., Ilvesniemi, H., Minkkinen, K., Vesala, T., Niinistö, S.,
Environment 42, 3625–3653. Lohila, A., Larmola, T., Morero, M., Pihlatie, M., Janssens, I., Curiel Yuste, J.,
Petroff, A., Mailliat, A., Amielh, M., Anselmet, F., 2007b. Aerosol dry deposition on Grünzweig, J.M., Reth, S., Subke, J.-A., Savage, K., Kutsch, W., Østreng, G.,
vegetative canopies. Part II: a new modelling approach and applications. Ziegler, W., Anthoni, P., Lindroth, A., Hari, P., 2004. Comparison of different
Atmospheric Environment 42, 3654–3683. chamber techniques for measuring soil CO2 efflux. Agricultural and Forest
Pérez, T., Trumbore, S.E., Tyler, S.C., Matson, P.A., Ortiz-Monasterio, I., Rahn, T., Meteorology 123, 159–176.
Griffith, D.W.T., 2001. Identifying the agricultural imprint on the global Raivonen, M., Bonn, M., Sanz, M.J., Vesala, T., Kulmala, M., Hari, P., 2006. UV-induced
N2O budget using stable isotopes. Journal of Geophysical Research 106, NOy emissions from Scots pine: could they originate from photolysis of
9869–9878. deposited HNO3? Atmospheric Environment 40, 6201–6213.
Personne, E., Loubet, B., Herrmann, B., Mattsson, M., Schjoerring, J.K., Nemitz, E., Rannik, U., Petaja, T., Buzorius, G., Aalto, P., Vesala, T., Kulmala, M., 2000. Deposition
Sutton, M.A., Cellier, P., 2009. SURFATM-NH3: a model combining the surface velocities of nucleation mode particles into a Scots pine forest. Environmental
energy balance and bi-directional exchanges of ammonia applied at the field and Chemical Physics 22, 97–102.
scale. Biogeosciences Discussions 6, 71–114. Rattray, G., Sievering, H., 2001. Dry deposition of ammonia, nitric acid, ammonium,
Philippot, L., Hallin, S., Börjesson, G., Baggs, E.M., 2009 Biochemical cycling in the and nitrate to alpine tundra at Niwot Ridge, Colorado. Atmospheric Environ-
rhizosphere having an impact on global change. Plant and Soil Special Issue ment 35, 1105–1109.
S28-RHIZO-BOOK, 321, 1–2. Regener, V.H., 1957. The vertical flux of atmospheric ozone. Journal of Geophysical
Phillips, G.J., Thomas, R., Famulari, D., Williams, P.I., Crosier, J., Allan, J., Coe, H., Research 62, 221–228.
Gallagher, M.W., Flynn, M., Nemitz, E., 2007. Fluxes of submicron organic Reid, J.S., Jonsson, H.H., Smith, M.H., Smirnov, A., 2001. Evolution of the vertical
particles above three UK cities. In: Measurement of Speciated Aerosol Fluxes. profile and flux of large sea-salt particles in the coastal zone. Journal of
Ph.D. thesis. Thomas, R. (Ed.). University of Manchester, Manchester. Geophysical Research 106, 12039–12053.
Phillips, G., Thomas, R., Famulari, D., Williams, P.I., Bower, K.N., Allan, J.D., Gallagher, Ren, L.-Q., Wang, S.-J., Tian, X.-M., Han, Z.-W., Yan, L.-N., Qui, Z.-M., 2007. Non-
M.W., Flynn, M., Nemitz, E. Fluxes of submicron aerosol above three UK cities. smooth morphologies of typical plant leaf surfaces and their anti-adhesion
Atmospheric Chemistry and Physics, in preparation. effects. Journal of Bionic Engineering 4, 33–40.
Pihlatie, M., Ambus, P., Rinne, J., Pilegaard, K., Vesala, T., 2005. Plant-mediated Rich, S., Waggoner, P.E., Tomlinson, H., 1970. Ozone uptake by bean leaves. Science
nitrous oxide emissions from beech (Fagus sylvatica) leaves. New Phytologist 169, 79–80.
168, 93–98. Riedo, M., Milford, C., Schmid, M., Sutton, M.A., 2002. Coupling soil–plant–atmo-
Pilegaard, K., Jensen, N.O., Hummelshøj, P., 1995. Seasonal and diurnal variation in sphere exchange of ammonia with ecosystem functioning in grasslands.
the deposition velocity of ozone over a spruce forest in Denmark. Water, Air and Ecological Modelling 158, 83–110.
Soil Pollution 85, 2223–2228. Rinne, J., Hakola, H., Laurila, T., 1999. Vertical fluxes of monoterpenes above a Scots
Pilegaard, K., 2001. Air–Soil exchange of NO, NO2 and O3 in forests. Water, Air and pine stand in the boreal vegetation zone. Physics and Chemistry of Earth 24,
Soil Pollution: Focus 1 (5–6), 79–88 (10). 711–715.
Pilegaard, K., Skiba, U., Ambus, P., Beier, C., Brüggemann, N., Butterbach-Bahl, K., Rinne, J., Hakola, H., Laurila, T., Rannik, U., 2000. Canopy scale monoterpene emissions
Dick, J., Dorsey, J., Duyzer, J., Gallagher, M., Gasche, R., Horvath, L., Kitzler, B., of Pinus sylvestris dominated forests. Atmospheric Environment 34, 1099–1107.
Leip, A., Pihlatie, M.K., Rosenkranz, P., Seufert, G., Vesala, T., Westrate, H., Zech- Rinne, H.J.I., Guenther, A.B., Warneke, C., de Gouw, J.A., Luxembourg, S.L., 2001.
meister-Boltenstern, S., 2006. Factors controlling regional differences in forest Disjunct eddy covariance technique for trace gas flux measurements.
soil emission of nitrogen oxides (NO and N2O). Biogeosciences 3, 651–661. Geophysical Research Letters 28, 3139–3142.
Pio, C.A., Feliciano, M.S., Vermeulen, A.T., 2000. Seasonal variability of ozone Rinne, H.J.I., Guenther, A.B., Greenberg, J.P., Harley, P.C., 2002. Isoprene and
dry deposition under southern European climate conditions. Atmospheric monoterpene fluxes measured above Amazonian rainforest and their depen-
Environment 34 (2), 195. dence on light and temperature. Atmospheric Environment 36, 2421–2426.
Pio, C.A., Silva, P.A., Cerqueira, M.A., Nuñes, T.V., 2005. Diurnal and seasonal emis- Rinne, J., Ruuskanen, T.M., Reissell, A., Taipale, R., Hakola, H., Kulmala, M., 2005. On-
sions of volatile organic compounds from cork oak (Quercus suber) trees. line PTR-MS measurements of atmospheric concentrations of volatile organic
Atmospheric Environment 39, 1817–1827. compounds in a European boreal forest ecosystem. Boreal Environmental
Pleijel, H., Pihl, G., Karlsson, G., Danielsson, H., Selldén, G., 1995. Surface wetness Research 10, 425–436.
enhances ozone deposition to a pasture canopy. Atmospheric Environment 29 Rinne, J., Taipale, R., Markkanen, T., Ruuskanen, T.M., Hellén, H., Kajos, M.K., Vesala, T.,
(22), 3391–3393. Kulmala, M., 2007. Hydrocarbon fluxes above a Scots pine forest canopy:
Pleijel, H., Danielsson, H., Ojanperä, K., De Temmerman, L., Högy, P., Badiani, M., measurements and modelling. Atmospheric Chemistry and Physics 7, 3361–3372.
Karlsson, P.E., 2004. Relationship between ozone exposure and yield loss in Rinne, J., Douffet, T., Prigent, Y., Durand, P., 2008. Field comparison of disjunct and
European wheat and potato – a comparison of concentration- and flux-based conventional eddy covariance techniques for trace gas flux measurements.
exposure indices. Atmospheric Environment 38, 2259–2269. Environmental Pollution 152, 630–635.
Pleijel, H., Danielsson, H., Emberson, L., Ashmore, M.R., Mills, G., 2007. Ozone risk Roberts, J.M., 1990. The atmospheric chemistry of organic nitrates. Atmospheric
assessment for agricultural crops in Europe: further development of stomatal Environment 24A (2), 243–287.
5264 D. Fowler et al. / Atmospheric Environment 43 (2009) 5193–5267

Robinson, A.L., Donahue, N.M., Shrivastava, M.K., Weitkamp, E.A., Sage, A.M., Simon, V., Dumergues, L., Bouchou, P., Torres, L., Lopez, A., 2005b. Isoprene emission
Grieshop, A.P., Lane, T.E., Pierce, J.R., Pandis, S.N., 2007. Rethinking organic rates and fluxes measured above a Mediterranean oak (Quercus pubescens)
aerosols: semivolatile emissions and photochemical aging. Science 2, 1259–1262. forest. Atmospheric Research 74, 49–63.
Romakkaniemi, S., Kokkola, H., Laaksonen, A., 2005. Soluble trace gas effect on Simpson, D., Winiwarter, W., Börjesson, G., Cinderby, S., Ferreiro, A., Guenther, A.,
cloud condensation nuclei activation: influence of initial equilibration on Hewitt, C.N., Janson, R., Khalil, M.A.K., Owen, S., Pierce, T.E., Puxbaum, H.,
cloud model results. Journal of Geophysical Research D 110, 15. doi:10.1029/ Shearer, M., Skiba, U., Steinbrecher, R., Tarrason, L., Öquist, M.G., 1999. Inventorying
2004JD005364. emissions from nature in Europe. Journal of Geophysical Research 104, 8113–8152.
Rosenstiel, T.N., Potosnak, M.J., Griffin, K.L., Fall, R., Monson, R.K., 2003. Increased Simpson, D., Fagerli, H., Jonson, J.E., Tsyro, S., Wind, P., Tuovinen J.-, P., 2003.
CO2 uncouples growth from isoprene emission in an agriforest ecosystem. Transboundary Acidification, Eutrophication and Ground Level Ozone in
Nature 421, 256–259. Europe. Part I: Unified EMEP Model Description. EMEP Status Report 2003. Det
Ruijgrok, W., Tieben, H., Eisinga, P., 1997. The dry deposition of particles to a forest Meteorologisk Institutt, Oslo.
canopy: a comparison of model and experimental results. Atmospheric Envi- Simpson, D., Fagerli, H., Jonson, J.E., Tsyro, S., Wind, P., Tuovinen, J.-P., 2006.
ronment 31, 399–415. Transboundary Acidification, Eutrophication and Ground Level Ozone in
Rummel, U., Ammann, C., Kirkman, G.A., 2007. Seasonal variation of ozone depo- Europe, Part I, Unified EMEP Model Description. Norwegian Meteorological
sition to a tropical rain forest in southwest Amazonia. Atmospheric Chemistry Office, Blindern. 1, 1–104. 1-8-2003.
and Physics 7, 7399–7450. Simpson, D., Ashmore, M., Emberson, L., Tuovinen, J.-P., 2007. A comparison of two
Russow, R., Sich, I., Neue, H.-U., 2000. The formation of the trace gases NO and N2O in different approaches for mapping potential ozone damage to vegetation. A
soils by the coupled processes of nitrification and denitrification: results of kinetic model study. Environmental Pollution 146, 715–725.
15
N tracer investigations. Chemosphere – Global Change Science 2, 359–366. Simpson, I.J., Meinardi, S., Blake, D.R., Blake, N.J., Rowland, F.S., Atlas, E., Flocke, F.,
Sakamoto, K., Takada, H., Sekiguchi, K., 2004. Influence of ozone, relative humidity 2002. A biomass burning source of C-1-C-4 alkyl nitrates. Geophysical Research
and flow rate on the deposition and oxidation of sulfur dioxide on yellow sand. Letters 29, 2168.
Atmospheric Environment 38, 6961–6967. Singh, H.B., Tabazadeh, A., Evans, M.J., Field, B.D., Jacob, D.J., Sachse, G., Crawford, J.H.,
Sander, R., Kerkweg, A., Jöckel, P., Lelieveld, J., 2005. Technical note: the new Shetter, R., Brune, W.H., 2003. Oxygenated volatile organic chemicals in the oceans:
comprehensive atmospheric chemistry module MECCA. Atmospheric Chem- inferences and implications based on atmospheric observations and air–sea
istry and Physics 5, 445–450. http://www.atmos-chem-phys.net/5/445/2005/. exchange models. Geophysical Research Letters 30 (16), 1862. doi:10.1029/
Sanderson, M.G., Collins, W.J., Hemming, D.L., Betts, R.A., 2007. Stomatal conduc- 2003GL017933.
tance changes due to increasing carbon dioxide levels: projeced impact on Sinha, V., Williams, J., Meyerhofer, M., Riebesell, U., Paulino, A.I., Larsen, A., 2007.
surface ozone levels. B. Tellus 59, 404–411. Air–sea fluxes of methanol, acetone, acetaldehyde, isoprene and DMS from
Scarratt, M.G., Moore, R.M., 1999. Production of chlorinated hydrocarbons and a Norwegian fjord following a phytoplankton bloom in a mesocosm experi-
methyl iodide by the red microalga Porphyridium purpureum. Limnology and ment. Atmospheric Chemistry and Physics 7, 739–755.
Oceanography 44, 703–707. Sitch, S., Cox, P.M., Collins, W.J., Huntingford, C., 2007. Indirect radiative forcing of climate
Schade, G.W., Goldstein, A.H., Gray, D.W., Lerdau, M.T., 2000. Canopy and leaf level change through ozone effects on the land-carbon sink. Nature 488, 791–794.
2-methyl-3-buten-2-ol fluxes from a ponderosa pine plantation. Atmospheric Skiba, U., Fowler, D., Smith, K.A., 1997. Nitric oxide emissions from agricultural soils
Environment 34, 3535–3544. in temperate and tropical climates: sources, controls and mitigation options.
Schade, G.W., Goldstein, A.H., 2001. Fluxes of oxygenated volatile organic Nutrient Cycling in Agroecosystems 48, 139–153.
compounds from a ponderosa pine plantation. Journal of Geophysical Research Skiba, U., 1998. Soil nitrous oxide and nitric oxide emissions as indicators of
106, 3111–3123. elevated atmospheric N deposition rates in seminatural ecosystems. Environ-
Schindlbacher, A., Zechmeister-Boltenstern, S., Butterbach-Bahl, K., 2004. Effects of mental Pollution 102, 457–461.
soil moisture and temperature on NO, NO2, and N2O emissions from European Skiba, U., Di Marco, C., Hargreaves, K., Sneath, R., McCartney, L., 2006. Nitrous oxide
forest soils. Journal of Geophysical Research 109, D17302. emissions from a dung heap measured by chambers and plume methods.
Schjoerring, J.K., Husted, S., Mattsson, M., 1998. Physiological parameters control- Agricultural Ecosystem & Environment 112, 135–139.
ling plant–atmosophere ammonia exchange. Atmospheric Environment Slemr, F., Seiler, W., 1984. Field measurements of NO AND NO2 emissions from
(Ammonia Special Issue) 32 (3), 491–498. 1998. fertilized and unfertilized soil. Journal of Atmospheric Chemistry 2, 1–24.
Schmidt, A., Klemm, O., 2008. Direct determination of highly size-resolved turbu- Slinn, W.G.N., 1982. Predictions for particle deposition to vegetative canopies.
lent particle fluxes with the disjunct eddy covariance method and a 12-stage Atmospheric Environment 16, 1785–1794.
electrical low pressure impactor. Atmospheric Chemistry and Physics Discus- Smart, D.R., Bloom, A.J., 2001. Wheat leaves emit nitrous oxide during nitrate
sions 8, 8997–9034. assimilation. Proceedings of the National Academy of Sciences of the United
Schnitzler, J.P., Graus, M., Kruezwieser, J., Heizmann, U., Renneberg, H., Wisthaler, A., States of America 98, 7875–7878.
Hansel, A., 2004. Contribution of different carbon sources to isoprene biosyn- Smith, R.I., Fowler, D., Sutton, M.A., Flechard, C., Coyle, M., 2000. Regional estima-
thesis in poplar leaves. Plant Physiology 135, 152–160. tion of pollutant gas deposition in the UK: model description, sensitivity
Scholefield, D., Hawkins, J.M.B., Jackson, S.M., 1997. Development of a helium analyses and outputs. Atmospheric Environment 34, 3757–3777.
atmosphere soil incubation technique for direct measurement of nitrous oxide Soerensen, L.L., Pryor, S.C., De Leeuw, G., Schulz, M., 2005. Flux divergence of nitric
and dinitrogen fluxes during denitrification. Soil Biology and Biochemistry 29, acid in the marine atmospheric surface layer. Journal of Geophysical Research –
1345–1352. Atmospheres 110. doi:10.1029/2004JD005403.
Schween, J.H., Dlugi, R., Hewitt, C.N., Foster, P., 1997. Determination and accuracy of Solberg, S., Hov, Ø., Søvde, A., Isaksen, I.S.A., Coddeville, P., De Backer, H., Forster, C.,
VOC-fluxes above the pine/oak forest at Castelporziano. Atmospheric Envi- Orsolini, Y., Uhse, K., 2008. European surface ozone in the extreme summer 2003.
ronment 31, 199–215. Journal of Geophysical Research 113, D07307. doi:10.1029/2007/JD009098.
Seco, R., Peñuelas, J., Filella, I., 2007. Short-chain oxygenated VOCs: emission and Sorimachi, A., Sakamoto, K., Ishihara, H., Fukuyama, T., Utiyama, M., Liu, H.,
uptake by plants and atmospheric sources, sinks, and concentrations. Atmo- Wang, W., Tang, D., Dong, X., Quan, H., 2003. Measurements of sulfur dioxide
spheric Environment 41, 2477–2499. and ozone dry deposition over short vegetation in northern China – a prelimi-
Sempreviva, A.M., Gryning, S.-E., 2000. Mixing height over water and its role on the nary study. Atmospheric Environment 37, 3157–3166.
correlation between temperature and humidity fluctuations in the unstable Sorimachi, A., Sakamoto, K., Sakai, M., Ishihara, H., Fukuyama, T., Utiyama, M.,
surface layer. Boundary-Layer Meteorology 97, 273–291. Liu, H., Wang, W., Tang, D., Dong, X., Quan, H., 2004. Laboratory and field
Seufert, G., Bartzis, J., Bomboi, T., Ciccioli, P., Cieslik, S., Dlugi, R., Foster, P., measurements of dry deposition of sulfur dioxide onto Chinese loess surfaces.
Hewitt, C.N., Kesselmeier, J., Kotzias, D., Lenz, R., Manes, F., Perez Pastor, R., Environmental Science & Technology 38, 3396–3404.
Steinbrecher, R., Torres, L., Valentini, R., Versino, B., 1997. An overview of the Sorimachi, A., Sakamoto, K., 2007. Laboratory measurement of the dry deposition of
Castelporziano experiments. Atmospheric Environment 31, 5–17. sulfur dioxide onto northern Chinese soil samples. Atmospheric Environment
Sharkey, T.D., Singsaas, E.L., Lerdau, M.T., Geron, C.D., 1999. Weather effects on 41, 2862–2869.
isoprene emission capacity and applications in emissions algorithms. Ecological Sorjamaa, R., Svenningsson, B., Raatikainen, T., Henning, S., Bilde, M., Laaksonen, A.,
Applications 9, 1132–1137. 2004. The role of surfactants in Köhler theory reconsidered. Atmospheric
Sharkey, T.D., Yeh, S.S., 2001. Isoprene emission from plants. Annual Review Plant Chemistry and Physics 4 (8), 2107–2117.
Physiology and Molecular Biology 52, 407–436. Sparks, J.P., Roberts, J.M., Monson, R.K., 2003. The uptake of gaseous organic
Shaw, W.J., Spicer, C.W., Kenny, D.V., 1998. Eddy correlation fluxes of trace gases nitrogen by leaves: a significant global nitrogen transfer process. Geophysical
using a tandem mass spectrometer. Atmospheric Environment 32, 2887–2898. Research Letters 30, 2189. doi: 2110.1029/2003GL018578.
Shaw, S.L., Chisholm, S.W., Prinn, R.G., 2003. Isoprene production by Pro- Spindler, G., Teichmann, U., Sutton, M.A., 2001. Ammonia dry deposition over
chlorococcus, a marine cyanobacterium, and other phytoplankton. Marine grassland: micrometeorological flux gradient measurements and bi-directional
Chemistry 80, 227–245. flux calculations using an inferential model. Quarterly Journal of the Royal
Sievering, H., 1983. Eddy flux and profile measurements of small particle dry- Meteorological Society 127, 795–814.
deposition velocity at the Boulder Atmospheric Observatory (BAO). In: Spirig, C., Guenther, A., Greenberg, J.P., Calanca, P., Tarvainen, V., 2004. Tethered
Pruppacher, H.R., Semonin, R.G., Slinn, W.G.N. (Eds.), Precipitation Scavenging, balloon measurements of biogenic volatile organic compounds at a Boreal
Dry Deposition and Resuspension. Elsevier, Amsterdam, NL, pp. 963–977. forest site. Atmospheric Chemistry and Physics 4, 215–229.
Simon, E., Meixner, F.X., Ganzeveld, L., Kesselmeir, J., 2005a. Coupled carbon–water Spirig, C., Neftel, A., Ammann, C., Dommen, J., Grabmer, W., Thielmann, A.,
exchange of the Amazon rain forest, II. Comparison of predicted and observed Schaub, A., Beauchamp, J., Wisthaler, A., Hansel, A., 2005. Eddy covariance flux
seasonal exchange of energy, CO2, isoprene and ozone at a remote site in measurements of biogenic VOCs during ECHO 2003 using proton transfer
Rondônia. Biogeosciences 2, 255–275. reaction mass spectrometry. Atmospheric Chemistry and Physics 5, 465–481.
D. Fowler et al. / Atmospheric Environment 43 (2009) 5193–5267 5265

Spracklen, D.V., Arnold, S.R., Carslaw, K.S., Sciare, J., Pio, C., 2008. Globally significant Sutton, M.A., Nemitz, E., Erisman, J.W., Beier, C., Butterbach Bahl, K., Cellier, C., de
oceanic source of organic carbon aerosol. Geophysical Research Letters 35, Vries, W., Cotrufo, F., Skiba, U., Di Marco, C., Jones, S., Laville, P., Soussana, J.F.,
L12811. doi:10.1029/2008GL033359. Loubet, B., Twigg, M., Famulari, D., Whitehead, J., Gallagher, M.W., Neftel, A.,
Stark, J.D., Smart, D.R., Hart, S.C., Haubensak, K.A., 2002. Regulation of nitric oxide Flechard, C., Herrmann, B., Calanca, P.L., Schjoerring, J.K., Daemmgen, U., Horvath, L.,
emissions from forest and rangeland soils of western North America. Ecology Tang, Y.S., Emmett, B.A., Tietema, A., Peñuelas, J., Kesik, M., Brueggemann, N.,
83, 2278–2292. Pilegaard, K., Vesala, T., Campbell, C.L., Olesen, J.E., Dragosits, U., Theobald, M.R.,
Staudt, M., Bertin, N., Frenzel, B., Seufert, G., 2000. Seasonal variation in amount Levy, P., Mobbs, D.C., Milne, R., Viovy, N., Vuichard, N., Smith, J.U., Smith, P.E.,
and composition of monoterpenes emitted by young Pinus pinea trees – Bergamaschi, P., Fowler, D., Reis, S., 2007. Challenges in quantifying biosphere–
implications for emission modeling. Journal of Atmospheric Chemistry 35, atmosphere exchange of nitrogen species. Environment Pollution 150, 125–139.
77–99. Sutton, M.A., Nemitz, E., Theobald, M.R., Milford, C., Dorsey, J.R., Gallagher, M.W.,
Staudt, M., Mir, C., Joffre, R., Rambal, S., Bonin, A., Landais, D., Lumaret, R., Hensen, A., Jongejan, P.A.C., Erisman, J.W., Mattsson, M.E., Schjoerring, J.K.,
2004. Isoprenoid emissions of Quercus spp. (Q. suber and Q. ilex) in mixed Cellier, P., Loubet, B., Roche, R., Neftel, A., Herrmann, B., Jones, S.K., Lehman, B.E.,
stands contrasting in interspecific genetic introgression. New Phytologist Horvath, L., Weidinger, T., Rajkai, K., Burkhardt, J., Löpmeier, F.J., Dämmgen, U.,
163, 573–584. 2008a. Dynamics of ammonia exchange with cut grassland: strategy and
Stehfest, E., Bouwman, L., 2006. N2O and NO emission from agricultural fields and implementation of the GRAMINAE integrated experiment. Biogeosciences
soils under natural vegetation: summarizing available measurement data and Discussions 5, 3347–3407.
modeling of global annual emissions. Nutrient Cycling in Agroecosystems 74, Sutton, M.A., Erisman, J.W., Dentener, F., Moeller, D., 2008b. Ammonia in the
207–228. environment: from ancient times to the present. Environmental Pollution 156
Steinbrecher, R., Klauer, M., Hauff, K., Stockwell, W.R., Jaeschke, W., Dietrich, T., (3), 583–604.
Herbert, F., 2000. Biogenic and anthropogenic fluxes of non-methane hydro- Sutton, M.A., Reis, S., Baker, S.M.H. (Eds.), 2008c. Atmospheric Ammonia: Detecting
carbons over an urban-impacted forest, Frankfurter Stadtwald, Germany. Emission Changes and Environmental Impacts. Springer.
Atmospheric Environment 34, 3779–3788. Sutton, M.A., Reis, S., Baker, S.M.H. (Eds.), 2009a. Atmospheric Ammonia: Detecting
Stemmler, K., Ammann, M., Donders, C., Kleffmann, J., George, C., 2006. Photo- Emission Changes and Environmental Impacts. Springer, 464 pp.
sensitized reduction of nitrogen dioxide on humic acid as a source of nitrous Sutton, M.A., Nemitz, E., Milford, C., Campbell, C., Erisman, J.W., Hensen, A.,
acid. Nature 440, 195–198. Cellier, P., David, M., Loubet, B., Personne, E., Schjoerring, J.K., Mattsson, M.E.,
Stevenson, D.S., Dentener, F.J., Schulz, M.G., Ellingsen, K., van Noije, T.P.C., Wild, O., Dorsey, J.R., Gallagher, M.W., Horvath, L., Weidinger, T., Meszaros, R.,
Zeng, G., Amann, M., Atherton, C.S., Bell, N., Bergmann, D.J., Bey, I., Butler, T., Dämmgen, U., Neftel, A., Herrmann, B., Lehman, B., Flechard, C., Burkhardt, J.,
Cofala, J., Collins, W.J., Derwent, R.G., Doherty, R.M., Drevet, J., Eskes, H.J., 2009b. Dynamics of ammonia exchange with cut grassland: synthesis of results
Fiore, A.M., Gauss, M., Hauglustaine, D.A., Horowicz, L.W., Isaksen, I.S.A., and conclusions of the GRAMINAE integrated experiment. Biogeoscience
Krol, M.C., Lamarque, J.-F., Lawrence, M.G., Montanaro, V., Müller, J.-F., Pitari, G., Discussions 6, 1121–1184.
Prather, M.J., Pyle, J.A., Rast, S., Rodriguez, J.M., Sanderson, M.G., Savage, N.H., Ta, W., Wei, C., Chen, F., 2005. Long-term measurements of SO2 dry deposition over
Shindell, D.T., Strahan, S.E., Sudo, K., Szopa, S., 2006. Multimodel ensemble Gansu Province, China. Atmospheric Environment 39, 7095–7105.
simulations of present-day and near-future tropospheric ozone. Journal of Takahashi, A., Sato, K., Wakamatsu, T., Fujita, S., 2001. Atmospheric deposition of
Geophysical Research 111, D08301. doi:10.1029/2005JD006338. acidifying components to a Japanese cedar forest. Water, Air and Soil Pollution
Sun, J., Ariya, P.A., 2006. Atmospheric organic and bio-aerosols as cloud conden- 130, 559–564.
sation nuclei (CCN): a review. Atmospheric Environment 40, 795–820. Takahashi, A., Sato, K., Wakamatsu, T., Fujita, S., Yoshikawa, K., 2002. Estimation of
Sutka, R.L., Ostrom, N.E., Ostrom, P.H., Gandhi, H., Breznak, J.A., 2003. Nitrogen dry deposition of sulfur to a forest using an inferential method. Influence of
isotopomer site preference of N2O produced by Nitrosomonas europaea and canopy wetness on SO2 dry deposition. Journal of Japan Society for Atmospheric
Methylococcus capsulatus Bath. Rapid Communications in Mass Spectrometry Environment 37 (3), 192–205.
17, 738–745. Tang, Y.S., Simmons, I., van Dijk, N., Di Marco, C., Nemitz, E., Dämmgen, U., Gilke, K.,
Sutton, M.A., Fowler, D., Moncrieff, J.B., 1993a. The exchange of atmospheric Djuricic, V., Vidic, S., Gliha, Z., Borovecki, D., Mitosinkova, M., Hanssen, J.E.,
ammonia with vegetated surfaces. I: unfertilized vegetation. Quarterly Journal Uggerud, T.H., Sanz, M.J., Sanz, P., Chorda, J.V., Flechard, C.R., Fauvel, Y., Ferm, M.,
of the Royal Meteorological Society 119, 1023–1045. Perrino, C., Sutton, M.A., 2009. European scale application of atmospheric
Sutton, M.A., Pitcairn, C.E.R., Fowler, D., 1993b. The exchange of ammonia between the reactive nitrogen measurements in a low-cost approach to infer dry deposition
atmosphere and plant communities. Advanced Ecology Research 24, 301–393. fluxes. Agriculture. Ecosystems & Environment 133, 183–195.
Sutton, M.A., Fowler, D., 1993. A model for inferring bi-directional fluxes of Tarrason, L., Nyiri, A., 2008. Transboundary Acidification, Eutrophication and Ground
ammonia over plant canopies. In: Proceedings of the WMO Conference on the Level Ozone in Europe in 2006. Norwegian Meterological Institute. 2008.
Measurement and Modelling of Atmospheric Composition Changes Including Theobald, M.R., Dragosits, U., Place, C.J., Smith, J.U., Sozanska, M., Brown, L.,
Pollutant Transport. World Meteorological Organization, Geneva, pp. 179–182. Scholefield, D., Del Prado, A., Webb, J., Whitehead, P.G., Angus, A., Hodge, I.D.,
Sutton, M.A., Asman, W.A.H., Schjørring, J.K., 1994. Dry deposition of reduced Fowler, D., Sutton, M.A., 2004. Modelling nitrogen fluxes at the landscape scale.
nitrogen. Tellus 46B, 255–273. WASP Focus 4 (6), 135–142.
Sutton, M.A., Schjørring, J.K., Wyers, G.P., 1995. Plant atmosphere exchange of ammonia. Thomas, R., 2007. Measurement of speciated aerosol fluxes. PhD thesis. In Physics
Philosophical Transactions of the Royal Society, London A351 (1656), 261–276. Department, University of Manchester, Manchester, 298 pp.
Sutton, M.A., Perthue, E., Fowler, D., Storeton-West, R.L., Cape, J.N., Arends, B.G., Thomas, R., Trebs, I., Otjes, R., Jongejan, P.A.C., ten Brink, H., Phillips, G., Korner, M.,
Möls, J.J., 1997. Vertical distribution and fluxes of ammonia at Great Dun Fell. In: Meixner, F.X., Nemitz, E., 2009. An automated analyser to measure surface–
Fuzzi, S. (Ed.), Special Issue on the Great Dun Fell Cloud Experiment 1993. atmosphere exchange fluxes of water soluble inorganic aerosol compounds and
Atmospheric Environment, vol. 31, pp. 2615–2625. reactive trace gases. Environmental Science & Technology. doi:10.1021/
Sutton, M.A., Burkhardt, J.K., Guerin, D., Nemitz, E., Fowler, D., 1998. Development of es8019403.
resistance models to describe measurement of bi-directional ammonia surface– Tingey, D.T., Turner, D.P., Weber, J.A., 1991. Factors controlling the emissions of
atmosphere exchange. Atmospheric Environment 32, 473–480. monoterpenes and other volatile organic compounds. In: Sharkey, T.D.,
Sutton, M.A., Nemitz, E., Fowler, D., Wyers, G.P., Otjes, R.P., Schjoerring, J.K., Holland, E.A., Mooney, H.A. (Eds.), Trace Gas Emissions by Plants.
Husted, S., Nielsen, K., San José, R., Moreno, J., Gallagher, M.W., Gut, A., 2000a. Academic Press, Inc., San Diego/New York/Boston/London/Sydney/Tokyo/
Fluxes of ammonia over oilseed rape: overview of the EXAMINE experiment. Toronto, pp. 93–119.
Agricultural and Forest Meteorology (Ammonia Special Issue) 105 (4), Topping, D.O., McFiggans, G.B., Kiss, G., Varga, Z., Facchini, M.C., Decesari, S.,
327–349. Mircea, M., 2007. Surface tensions of multi-component mixed inorganic/organic
Sutton, M.A., Nemitz, E., Milford, C., Fowler, D., Moreno, J., San Jose, R., Wyers, G.P., aqueous systems of atmospheric significance: measurements, model predic-
Otjes, R., Harrison, R., Husted, S., Schjoerring, J.K., 2000b. Micrometeorological tions and importance for cloud activation predictions. Atmospheric Chemistry
measurements of net ammonia fluxes over oilseed rape during two vegetation and Physics 7 (9), 2371–2398.
periods. Agricultural and Forest Meteorology (Ammonia Exchange Special Tost, H., Jöckel, P., Kerkweg, A., Pozzer, A., Sander, R., Lelieveld, J., 2007. Global cloud
Issue) 105 (4), 351–369. and precipitation chemistry and wet deposition: tropospheric model simula-
Sutton, M.A., Milford, C., Nemitz, E., Theobald, M.R., Hill, P.W., Fowler, D., tions with ECHAM5/MESSy1. Atmospheric Chemistry and Physics 7, 2733–2757.
Schjoerring, J.K., Mattsson, M.E., Nielsen, K.H., Husted, S., Erisman, J.W., Otjes, R., http://www.atmos-chem-phys.net/7/2733/2007/.
Hensen, A., Mosquera, J., Cellier, P., Loubet, B., David, M., Genermont, S., Neftel, A., Trebs, I., Lara, L.L., Zeri, L.M.M., Gatti, L.V., Artaxo, P., Dlugi, R., Slanina, J.,
Blatter, A., Herrmann, B., Jones, S.K., Horvath, L., Führer, E., Mantzanas, K., Andreae, M.O., Meixner, F.X., 2006. Dry and wet deposition of inorganic
Koukoura, Z., Gallagher, M., Williams, P., Flynn, M., Riedo, M., 2001. Biosphere– nitrogen compounds to a tropical pasture site (Rondonia, Brazil). Atmospheric
atmosphere interactions of ammonia with grasslands: experimental strategy and Chemistry and Physics 6, 447–469.
results from a new European initiative. Plant and Soil 228, 131–145. Tuovinen, J.P., Aurela, M., Laurila, T., 1998. Resistance to ozone deposition to a flark
Sutton, M.A., Milford, C., Nemitz, E., Theobald, M.R., Riedo, M., Hargreaves, K.J., Hill, P.W., fen in the northern aapa mire zone. Journal of Geophysical Research-
Dragosits, U., Fowler, D., Schjoerring, J.K., Mattsson, M.E., Husted, S., Erisman, J.W., Atmospheres 203, 16953–16966.
Hensen, A., Mosquera, J., Otjes, R., Cellier, P., Loubet, B., David, M., Neftel, A., Tuovinen, J.-P., 2000. Assessing vegetation exposure to ozone: properties of the
Hermann, B., Jones, S.K., Blatter, A., Horvath, L., Weidinger, T., Meszaros, R., Raso, J., AOT40 index and modifications by deposition modelling. Environmental
Mantzanas, K., Koukoura, Z., Gallagher, M., Dorsey, J., Flynn, M., Lehman, B., Pollution 109, 361–372.
Burkhardt, J., Daemmgen, U., 2002. Ammonia exchange with vegetation: measure- Tuovinen, J.-P., Simpson, D., Mikkelsen, T.N., Emberson, L.D., Ashmore, M.R.,
ment, modelling & application. In: Midgley, P.M., Reuther, M. (Eds.), Transport and Aurela, M., Cambridge, H.M., Hovmand, M.F., Jensen, N.O., Laurila, T.,
Chemical Transformation in the Troposphere. Margraf Verlag, pp. 49–60. Pilegaard, K., Ro-Poulsen, H., 2001. Comparisons of measured and modelled
5266 D. Fowler et al. / Atmospheric Environment 43 (2009) 5193–5267

ozone deposition to forests in Northern Europe. Water, Air and Soil Pollution: Werner, C., Butterbach-Bahl, K., Haas, E., Hickler, T., Kiese, R., 2007. A global inventory
Focus 1, 263–274. of N2O emissions from tropical rainforest soils using a detailed biogeochemical
Tuovinen, J.-P., Ashmore, M., Emberson, L., Simpson, D., 2004. Testing and model. Global Biogeochemical Cycles 21, GB3010. doi:10.1029/2006GB002909.
improving the EMEP ozone deposition module. Atmospheric Environment 38, Wesely, M.L., 1983. Turbulent transport of ozone to surfaces common in the eastern
2373–2385. half of the United States. Advances in Environmental Science and Technology
Tuovinen, J.-P., Simpson, D., Altimir, N., 2008. Modelling the Stomatal and Non- 12, 345–370.
stomatal Ozone Fluxes to Coniferous Forests in Northern Europe. EGU General Wesely, M.L., Cook, D.R., Hart, R.L., Speer, R.E., 1985. Measurements and parame-
Assembly, 14–18 April 2008, Vienna, Austria, Geophysical Research Abstracts, terization of particulate sulfur dry deposition over grass. Journal of Geophysical
vol. 10. EGU2008-A-08272. Research – Atmospheres 90, 2131–2143.
Tuovinen, J.-P., Simpson, D., 2008b. An aerodynamic correction for the European Wesely, M.L., 1989. Parameterization of surface resistance to gaseous dry deposition
ozone risk assessment methodology. Atmospheric Environment 42, 8371–8381. in regional scale, numerical models. Atmospheric Environment 23, 1293–1304.
Turner, N.C., Waggoner, P.E., Rich, S., 1974. Removal of ozone from the atmosphere Wesely, M.L., Hicks, B.B., 2000. A review of the current status of knowledge on dry
by soil and vegetation. Nature 250, 486–489. deposition. Atmospheric Environment 34, 2261–2282.
Turnipseed, A.A., Huey, L.G., Nemitz, E., Stickel, R., Higgs, J., Tanner, D.J., Slusher, D.L., Whitehead, J.D., Twigg, M., Famulari, D., Nemitz, E., Sutton, M.A., Gallagher, M.W.,
Sparks, J.P., Flocke, F., Guenther, A., 2006. Eddy covariance fluxes of peroxyacetyl Fowler, D., 2008. Evaluation of laser absorption spectroscopic techniques for
nitrates (PANs) and NOy to a coniferous forest. Journal of Geophysical Research eddy covariance flux measurements of ammonia. Environmental Science &
– Atmospheres 111, D09304. doi: 09310.01029/02005JD006631. Technology. doi:10.1021/es071596u.
Twigg, M., Famulari, D., Fowler, D., Gallagher, M., Nemitz, E., Sutton, M.A., Whitehead, J.D., McFiggans, G.B., Gallagher, M.W., Flynn, M.J., 2009. Direct linkage
Whitehead, J., Rippey, B., 2005. Principles and development of tunable laser between tidally driven coastal ozone deposition fluxes, particle emission fluxes
diode absorption spectroscopy for measuring fluxes of ammonia and nitric acid. and subsequent CCN formation. Geophysical Research Letters 36, L04806.
First ACCENT Symposium Proceedings. Urbino, 12–16. September 2005. Wichink Kruit, R.J., van Pul, W.A.J., Otjes, R.P., Hofschreuder, P., Jacobs, A.F.G.,
Ulbrich, I.M., Canagaratna, M.R., Zhang, Q., Worsnop, D.R., Jimenez, J.L. 2009. Holtslag, A.A.M., 2007. Ammonia fluxes and derived canopy compensation
Interpretation of organic components from positive matrix Factorization of points over non-fertilized agricultural grassland in The Netherlands using the
aerosol mass spectrometric data. Atmospheric Chemistry and Physics Discus- new gradient ammoniadhigh accuracydmonitor (GRAHAM). Atmospheric
sions 9, 2891–2918. Environment 41, 1271–1287.
UNECE, 2004. Mapping critical levels for vegetation. In: Manual on Methodologies Wiedinmyer, C., Greenberg, J., Guenther, A., Hopkins, B., Baker, K., Geron, C.,
and Criteria for Modelling and Mapping Critical Loads & Levels and Air Pollution Palmer, P.I., Long, B.P., Turner, J.R., Petron, G., Pierce, T.E., Lamb, B., Westberg, H.,
Effects, Risks and Trends. UNECE Convention on Long-range Transboundary Air Baugh, W., Harley, P., Koerber, M., Jannsen, M., 2005. The Ozarks Isoprene
Pollution, Geneva Available at: www.icpmapping.org (Chapter 3). Experiment (OZIE): measurements and model interpretations of the ‘isoprene
Utiyama, M., Fukuyama, T., Sakamoto, K., Ishihara, H., Sorimachi, A., Tanonaka, T., volcano’. Journal of Geophysical Research. doi:10.1029/2005JD005800.
Dong, X., Quan, H., Wang, W., Tang, D., 2005. Sulfur dioxide dry deposition on Williams, J., Pöschl, U., Crutzen, P.J., Hansel, A., Holzinger, R., Warneke, C.,
the loess surface–surface reaction concept for measuring dry deposition flux. Lindinger, W., Lelieveld, J., 2001. An atmospheric chemistry interpretation of
Atmospheric Environment 39, 329–335. mass Scans obtained from a proton transfer mass spectrometer flown over the
Vaattovaara, P., Huttunen, P.E., Yoon, Y.J., Joutsensaari, J., Lehtinen, K.E.J., O’Dowd, C.D., tropical rainforest of Surinam. Journal Atmospheric Chemistry 38, 133–166.
Laaksonen, A., 2006. The composition of nucleation and Aitken modes particles Williams, J., 2004. Organic trace gases: an overview. Environmental Chemistry 1,
during coastal nucleation events: evidence for marine secondary organic contri- 125–136.
bution. Atmospheric Chemistry and Physics 6, 4601–4616. Williams, J., Holzinger, R., Gros, V., Xu, X., Atlas, E., Wallace, D.W.R., 2004.
Van Hove, L.W.A., Adema, E.H., Vredenberg, W.J., Pieters, G.A., 1989. A study of the Measurements of organic species in air and seawater from the tropical Atlantic.
adsorption of NH3 and SO2 on leaf surfaces. Atmospheric Environment 23, Geophysical Research Letters 31 (23), L23S06.
1479–1486. Wolfe, G.M., Yatavelli, R.L.N., Thornton, J.A., McKay, M., Goldstein, A.H., LaFranchi, B.,
Van Oijen, M., Rougier, J., Smith, R., 2005. Bayesian calibration of process-based Min, K.-E., Cohen, R.C., 2008. Eddy covariance fluxes of acyl preoxy nitrates
forest models: bridging the gap between models and data. Tree Physiology 25, (PAN, PPN, and MPAN) above a Ponderosa pine forest. Atmospheric Chemistry
915–927. and Physics Discussions 8, 17495–17548.
Van Oss, R., Duyzer, J., Wyers, P., 1998. The influence of gas-to-particle conversion Wolke, R., Sehili, A.M., Simmel, M., Knoth, O., Tilgner, A., Herrmann, H., 2005.
on measurements of ammonia exchange over forest. Atmospheric Environment SPACCIM: a parcel model with detailed microphysics and complex multiphase
32, 465–471. chemistry. Atmospheric Environment 39, 4375–4388.
Vautard, R., Honoré, C., Beekmann, M., Rouil, L., 2005. Simularion of ozone during Wrage, N., van Groenigen, J.W., Oenema, O., Baggs, E.M., 2005. A novel dual-isotope
the August 2003 heat wave and emission control scenarios. Atmospheric labelling method for distinguishing between soil sources of N2O. Rapid
Environment 39, 2957–2967. Communications in Mass Spectrometry 19, 3298–3306.
Velasco, E., Lamb, B., Pressley, S., Allwine, E., Westberg, H., Jobson, B.T., Wuebbles, D.J., Hayhoe, K., 2002. Atmospheric methane and global change. Earth-
Alexander, M., Prazeller, P., Molina, L., Molina, M., 2005. Flux measurements of Science Reviews 57, 177–210.
volatile organic compounds from an urban landscape. Geophysical Research Wyers, G.P., Otjes, R.P., Slanina, J., 1993. A continuous-flow denuder for the
Letters 32 (20). measurement of ambient concentrations and surface-exchange fluxes of
Venterea, R.T., Rolston, D.E., 2000. Mechanisms and kinetics of nitric and nitrous ammonia. Atmospheric Environment 27, 2085–2090.
oxide production during nitrification in agricultural soil. Global Change Biology Wyers, G.P., Duyzer, J.H., 1997. Micrometeorological measurement of the dry
6, 303–316. deposition flux of sulphate and nitrate aerosols to coniferous forest. Atmo-
Venterea, R.T., Rolston, D.E., Cardon, Z.G., 2005. Effects of soil moisture, physical, spheric Environment 31, 333–343.
and chemical characteristics on abiotic nitric oxide production. Nutrient Cycling Wyers, G.P., Erisman, J.W., 1998. Ammonia exchange over coniferous forest.
in Agroecosystems 72, 27–40. Atmospheric Environment 32 (3), 441–451.
Vesala, T., Jarvi, L., Launiainen, S., Sogachev, A., Rannik, Ü., Mammarella, I., Siivola, E., Xu, Y., Carmichael, G.R., 1998. Modeling the dry deposition velocity of sulfur dioxide
Keronen, P., Rinne, J., Riikonen, A., Nikinmaa, E., 2007. Surface–atmosphere and sulfate in Asia. Journal of Applied Meteorology 37, 1084–1099.
interactions over complex urban terrain in Helsinki, Finland. Tellus B 60, Xu, C.K., Hu, Z.Y., Cai, Z.C., Wang, T.J., He, Y.Q., Cao, Z.H., 2004. Atmospheric sulfur
188–199. deposition for a red soil broadleaf forest in southern China. Pedosphere 14 (3),
Vigano, I., van Weelden, H., Holzinger, R., Keppler, F., McLeod, A., Rockmann, T., 323–330.
2008. Effect of UV radiation and temperature on the emission of methane from Yan, X., Shi, S., Du, L., Xing, G., 2000. Pathways of N2O emission from rice paddy soil.
plant biomass and structural components. Biogeosciences 5, 937–947. Soil Biology and Biochemistry 32, 437–440.
Vitel, H., Kromer, B., Moessner, M., Platt, U., 2002. New technique for measurements Yassaa, N., Peeken, I., Zöllner, E., Bluhm, K., Arnold, S., Spracklen, D., Wernli, H.,
of atmospheric vertical trace gas profiles using DOAS. Environmental Science Williams, J., 2008. Evidence for marineproduction of monoterpenes, Environ-
and Pollution 9, 17–26. mental Chemistry 6, 391–401.
Voisin, D., Legrand, M., Chaumerliac, N., 2000. Scavenging of acidic gases (HCOOH, Yienger, J., Levy, H., 1995. Empirical model of global soil–biogenic NOx emissions.
CH3COOH, HNO3, HCl, and SO2) and ammonia in mixed liquid-solid water Journal of Geophysical Research 100, 11447–11464.
clouds at the Puy de Dôme mountain (France). Journal of Geophysical Research Yin, Y., Carslaw, K.S., Feingold, G., 2005. Vertical transport of aerosols in a mixed-
105 (D5), 6817–6836. phase convective cloud and the feedback on cloud development. Quarterly
Vong, R.J., Vickers, D., Covert, D.S., 2004. Eddy correlation measurements of aerosol Journal of the Royal Meteorological Society 131, 221–245.
deposition to grass. Tellus Series B – Chemical and Physical Meteorology 56B, Yoon, Y.J., Ceburnis, D., Cavalli, F., Jourdan, O., Putaud, J.P., Facchini, M.C., Decesari, S.,
105–117. Fuzzi, S., Jennings, S.G., O’Dowd, C.D., 2007. Seasonal characteristics of the
Walker, J.T., Robarge, W.P., Wuc, Y., Meyers, T.P., 2006. Measurement of bi-direc- physico-chemical properties of North Atlantic marine atmospheric aerosols.
tional ammonia fluxes over soybean using the modified Bowen-ratio technique. Journal of Geophysical Research 112, D04206. doi: 0.1029/2005JD007044.
Agricultural and Forest Meteorology 138, 54–68. Yun, H.J., Yi, S.M., Kim, Y.P., 2002. Dry deposition fluxes of ambient particulate heavy
Wang, T.J., Zhang, Y., Zhang, M., Hu, Z.Y., Xu, C.K., Zhao, Y.W., 2003. Atmospheric metals in a small city, Korea. Atmospheric Environment 36, 5449–5458.
sulfur deposition and the sulfur nutrition of crops at an agricultural site in Zhang, L.F., Boeckx, P., Chen, G.X., Van Cleemput, O., 2000. Nitrous oxide emission
Jiangxi province of China. Tellus 55B, 893–900. from herbicide-treated soybean. Biology and Fertility of Soils 32, 173–176.
Webb, E.K., Pearman, G.I., Leuning, R., 1980. Correction of flux measurements for Zhang, L., Gong, S., Padro, J., Barrie, L., 2001. A size-segregated particle dry depo-
density effects due to heat and water vapour transfer. Quarterly Journal of the sition scheme for an atmospheric aerosol module. Atmospheric Environment
Royal Meteorological Society 106, 85–100. 35, 549–560.
D. Fowler et al. / Atmospheric Environment 43 (2009) 5193–5267 5267

Zhang, L., Brook, J.R., Vet, R., 2002. On ozone dry deposition – with emphasis on non- Zimmermann, F., Plessow, K., Queck, R., Bernhofer, C., Matschullat, J., 2006.
stomatal uptake and wet canopies. Atmospheric Environment 36, 4787–4799. Atmospheric N- and S-fluxes to a spruce forestdcomparison of inferential
Zhang, L., Brook, J.R., Vet, R., 2003a. Evaluation of a non-stomatal resistance param- modelling and the throughfall method. Atmospheric Environment 40,
eterization for SO2 dry deposition. Atmospheric Environment 37, 2941–2947. 4782–4796.
Zhang, L., Brook, J.R., Vet, R., 2003b. A revised parametrization for gaseous dry depo- Zorn, S.R., Drewnick, F., Schott, M., Hoffmann, T., Borrmann, S., 2008. Charac-
sition in air-quality models. Atmospheric Chemistry and Physics 3, 2067–2082. terization of the South Atlantic marine boundary layer aerosol using an
Zhang, L., Vet, R., 2006. A review of current knowledge concerning size-dependent aerodyne aerosol mass spectrometer. Atmospheric Chemistry and Physics 8,
aerosol removal. China Particuology 4 (6), 272–282. 4711–4728.
Zimmerman, P.R., Greenberg, J.P., Westberg, C.A., 1988. Measurements of atmo- Zou, J.W., Huang, Y., Sun, W.J., Zheng, X.H., Wang, Y.S., 2005. Contribution of plants
spheric hydrocarbons and biogenic emission fluxes in the Amazon boundary to N2O emissions in soil-winter wheat ecosystem: pot and field experiments.
layer. Journal of Geophysical Research 93, 1407–1416. Plant and Soil 269, 205–211.

You might also like