Professional Documents
Culture Documents
Urban Sound Planning
Urban Sound Planning
ISBN 9789163918599
90000 >
URBAN SOUND PLANNING
9 789163 918599 - the SONORUS project
URBAN SOUND PLANNING
- the SONORUS project
1
Foreword
SONORUS is a European Training Net- sound environment with all its related
work inside the Marie Skłodowska- aspects such as city and traffic planning,
Curie Actions funded by the European architectural aspects, noise control and
Commission. The goal of those training soundscaping, as well as political pro-
networks is to bring together universities, cesses.
research centres and companies from Second, a key issue for such holistic
different countries to train a new genera- urban planning is the successful commu-
tion of researchers with a profile beyond nication between all people involved in
existing professions. the process together with the inclusion
SONORUS’ overall objective was of tools and methods for communica-
and is to offer young researchers in the tion. To achieve this, the tools and meth-
early-stage of their career the opportun- ods that are needed go beyond today’s
ity to develop their knowledge and skills state of the art.
in the area of urban sound planning. The Third, one rarely finds acoustic ex-
idea of SONORUS stems from several perts mastering the whole spectrum of
observations. knowledge. The demanded tools inside
First, the complexity to handle the the field of acoustics needed for such a
sound environment in our cities de- task include applying state-of-art sound
mands a trans-disciplinary approach to prediction tools for shielded and quiet
master the planning process of an urban areas, being able to work with the con-
3
Members of the SONORUS consortium: Members of the Educational Program:
Senior Researchers: Francesco Aletta
Wolfgang Kropp, coordinator - Chalmers University of Sonia Alves
Technology, Sweden Gemma Echevarría Sánchez
Jens Forssén , coordinator - Chalmers University of Laura Estévez Mauriz
Technology, Sweden Karlo Filipan
Beate Altreuther - Müller BBM, Germany Ignacio García Merino
Dick Botteldooren - Gent University, Belgium Fotis Georgiou
Salvatore Curcuruto - ISPRA, Italy Like Jiang
Kurt Heutschi - EMPA, Switzerland Mercury Kounturas
Maarten Hornikx - Eindhoven University of Technology, Efstathios Margaritis
the Netherlands Raúl Pagán Muñoz
Nicole van Hout - Eindhoven University of Technology, Virginia Puyana Romero
the Netherlands Frederik Rietdijk
Jian Kang - Sheffield University, UK Georgios Zachos
Krister Larsson - SP, Sweden
Luigi Maffei - Second University of Naples, Italy City representatives:
Timothy van Renterghem - Gent University , Belgium Rebecca Beeckman (Antwerp)
Joachim Scheuren - Müller BBM, Germany Matthew Easteal & Simon Banister (Brighton & Hove)
Rosalba Silvaggio - ISPRA Martin Knape, Henrik Nystedt & Maria Holmes
(Gothenburg)
Guido Carati (Rome)
Introduction 7
veloped in SONORUS. In Section 2 the macroscale, related to urban planning;
control of the sound environment is re- on the mesoscale, related to urban
lated to e.g. decisions on traffic planning design; and on the microscale.
and urban form. Methods for predicting Although the process of urban
and auralising the sound environment, sound planning has been initiated and
as essential tools for communicating first attempts have been made toward
the acoustic consequences of different using such a process, it is essential to
planning scenarios, are presented in recognize that the work documented
Section 3. Section 4 focuses on the use in this booklet is just the beginning of a
of the soundscape approach as a tool to development.
design the perceived acoustic environ- The on-going urbanization de-
ment (i.e. the sonic environment) from an mands a different way of planning,
end-user perspective. Although tools are where a holistic view is essential to
important, the most important advance- create attractive cities. A positive sound
ment is the implementation of urban environment is an essential part of the
sound planning in real life cases. Section perception of cities. Only an attractive
5 presents four test sites where the idea urban environment will be successful on
of urban sound planning is converted to an economic level, by being able to de-
realistic scenarios for the cities of velop, attract and to keep competence,
Antwerp, Brighton, Rome and Gothen- enterprises and financial resources, as
burg. The discussions in the following well as creating means and capabilities
text is adapted to the general view on for further development of future social,
planning by discussing urban sound cultural, environmental and economic
planning on three different scales: on the sustainability.
Introduction
The process of Urban Sound Plan- use of greenery in the context of noise
ning requires the possibility to control engineering.
However, when it comes to Urban
control the sound environment. Sound Planning, the needs are beyond
Traditionally we talk about noise just controlling the noise. Tools are called
control in this context, and during for that allow for shaping and exchang-
the past a multitude of noise mit- ing dominating sounds in an area. In ad-
dition, approaches are demanded which
igation measures have been de- can be applied on all three scales:
veloped and practical applications On the macroscale, where an acous-
explored. tic master plan may define the needs and
also the ambition of a city with respect
This includes tools such as noise barriers, to the sound environment, tools and
low-noise road surfaces or the use of approaches are required for controlling
green roofs and facades. Here, the out- the overall sound environment. There is
comes of the EC project HOSANNA may also still a need to develop a theoretical
be a useful source of inspiration for the framework supporting the planning on
9
this scale. noise control engineering work. How-
On the mesoscale, where urban ever, in an urban sound planning pro-
areas are considered, transport manage- cess, measures are motivated by thor-
ment and traffic design are important ough pre-considerations, e.g. of the
tools for controlling the sound environ- architectural design.
ment. In the following, approaches de-
On the microscale, local modifica- veloped within SONORUS are presented
tions are made to adjust the sound en- which have the goal to control the sound
vironment to the needs in a very limited environment on all these three scales.
area. This is most similar to traditional
Controlling the
sound environment
at microscale level
CONTROLLING THE SOUND unaware of the acoustical consequences
ENVIRONMENT BY LOCAL that the presence of the different urban
elements can have on the exposure of
ARCHITECTURAL ELEMENTS
these people as well as on those living in
In the urban context, motorized traffic
the flanking dwellings. The facade shape,
and pedestrians or cyclists are often
the width of the street, and urban fur-
found in the same street canyon. Urban
niture are natural elements in the urban
designers and architects are frequently
environment that can transform the 20 m wide canyon and 8-floor buildings
propagation of sound. Add-on small (See Case F1.1 in Figure 2a), organized in
barriers, absorptive layers, and shaped street cases (S) and facade cases (F). The
balconies can alleviate exposure. (See source is two-lane urban traffic at speed
example in Figure 1). 50 km/h. The results from this study,
To quantify the potential benefit of given below sorted by efficiency, should
each of these effects, a range of over 60 be taken as an instructive guideline to
cases were calculated using a numer- understand the effect of different urban
ical method that can account for small geometries on noise. The optimal design
geometrical details and all propagation to achieve the greatest noise reduction
effects they cause: A 2-D Finite Difference for any other urban situation would need
Time Domain method. All cases analysed new calculations. The conclusions from
are variations of a typical geometry of a these calculations are summarized below.
Facade cases:
(a)
(c)
(d)
(e)
Figure 2 - Studied cases: General Building Shapes (a),
Setback in lower storeys (b), Balcony geometry (c),
Triangular prominences on facade (d), Shielded inclined
windows (e), Low barrier shape (f), Absorption on a
vertical low barrier (g), Absorption on an inclined low
F5.1 F5.2 barrier (h), Depressed road (i), Two level street ( j).
(f)
(g)
(h)
(i)
(j)
have one of its facades protected from noise mapping prediction software. The
high noise levels. The recommendation is main focus is on the diffraction over the
to keep that level below 50 dB and buildings combined with multiple reflec-
preferably below 45 dB as a day- tions in both the street canyon and the
evening-night noise level. inner-yard.
What makes these areas so im- Results between the implementation
portant is their tremendous capacity to model and the noise mapping software
influence the urban decision-making show differences around 10 dB for low
process, having an impact at all urban frequencies for hard facades, increasing
scales. These conditions are making the with frequency (Figure 4). This is in the
quiet side an incredibly powerful tool in case we have 1 reflection order, which
the development of new urban areas, is usually how the noise maps from our
especially when densification of consoli- cities are calculated. Results get closer
dated cities is pursued. (See illustration in to each other when the noise mapping
Figure 3). software calculations include a higher
Normally, noise mapping has been order of reflections (20 in this case). The
the main tool to obtain noise levels at calculations are made for hard ground
the quiet areas. However, they are regu- (dense asphalt) and for both soft facade
larly underestimating the noise levels in (SF, 20% absorption) and hard facade
such type of areas. This is mainly due to (HF, 3% absorption).
the fact that the software used are de- The Qside implementation and
veloped to study the most exposed fa- noise mapping software calculations are
cade. In SONORUS the Qside engin- compared with noise measurements at
eering model that accounts for quiet an inner-yard in the city of Gothenburg
areas is expanded, in order to implement (see Figure 5). Similar spectra are found
it in real case studies and compare it with for the Qside implementation and the
Controlling the sound environment 17
Figure 4 - Contribution to the inner-yard noise level from a line source (road).
Differences between noise mapping software with different reflection order and
Qside implementation with soft and hard facade materials.
TRANSPORT MANAGEMENT
AND NOISE EMISSIONS:
(b)
INTERSECTIONS
Among the microscale traffic study in
cities, one of the most important designs
is of the intersection. A common practice
has been to replace crossings by round-
abouts as a safer alternative. To compare
these two intersection types, the influ-
ence of vehicle kinematics is studied in
SONORUS through a microscopic traffic
assessment. The two intersection types Figure 6 - Two intersection types. Crossing and
are based on the future urban develop- roundabout. (a) Amount of traffic, (b) microscopic
ment of Frihamnen in Gothenburg, with traffic simulation.
a different number of vehicles coming
from the different streets approaching In the scenarios, the same amount of
the intersection. The intention is to iso- traffic was handled in both intersection
late key features that could help to un- types, adjusting the road layout. In Figure
derstand their behaviour and the sound 6a, the amount of traffic is indicated, with
environment impact. For this, we study a smaller total flow in the E-W direction
several indicators based on time patterns compared with the N-S direction. Figure
related to human annoyance for three 6b represents the microscopic traffic sim-
scenarios of each intersection type. ulations for both scenarios.
Case % Vehicles
Light (LV) Medium-heavy (MHV) Heavy vehicles (HV)
1) LV-MHV-HV 92 4 4
2) LV-MHV 96 4 -
3) LV 100 - -
Since vehicle types also have a strong pends on how traffic is handled:
influence on people’s perception of the • Queues at certain lanes make it diffi-
sound environment, we study alternat- cult to enter the roundabout. In this
ives of including heavy-vehicles (>12 tons case, if points are located close to
as large buses and heavy duty vehicles) the intersection, crossing has higher
and medium-heavy vehicles (3.5-12 tons) noise levels (1-4 dB);
in comparison with having only light • For sidewalks in the E-W direction,
vehicles (<3.5 tons) for the peak hour as the roundabout tends to have high-
the worst-case scenario (Table 1). er noise levels (probably due to low
To study the differences between traffic flow and the resulting higher
these two intersection types, 12 study driving speeds);
points are included. The results, dis- • For location points at 100 m of the
played in Figure 7, show that not all intersection, the behaviour is similar
study points are less noisy for a certain for both.
intersection type, since it strongly de-
(c)
The equivalent sound pressure level
difference maps present a useful tool to
observe the different behaviours of the
two intersections (Figure 8):
• With different vehicle types, the sig-
nalized crossing intersection is having
a higher sound pressure level almost
in the entire area (blue colour), be-
ing more equal when removing the
heavier vehicles;
• The effect of vehicle kinematics is
present and it is observable as a stop-
and-go behaviour represented by
blue dots that indicate higher noise
Figure 8 - Equivalent noise level difference maps levels due to interrupted traffic flow;
for the crossing and the roundabout for the three • With these types of maps is clearly
vehicle types: light, medium-heavy and heavy observable that noise mapping soft-
vehicles (a), light and medium-heavy vehicles (b), ware, which gives as output a con-
light vehicles (c)
stant traffic flow, is not enough in case
we want to study how traffic behaves
in urban areas and the impact on
people’s health and behaviour.
(c)
At high traffic variations, as the ones
present in dense urban environments,
time patterns, as shown in Figure 9,
become relevant since noise annoyance
is partly determined by the noise events
resulting from traffic flow. Moreover, if we
are interested in the types of activities
and uses that citizens can do of such
urban spaces, the sound environment
becomes even more important. For
example, if one wants to experience the
city-life, have a coffee with friends with-
out being disturbed by road traffic noise,
Figure 9 - Time patterns at (a) study point etc.
4 (at the intersection), (b) study point 8 (at In the study of number of events
100 m of intersection) and (c) study point above 60 dBA for these two intersections
11 (at sidewalk) for the signalized crossing
(see Figure 10), one can conclude that:
and the roundabout
• There is a strong influence of heavy
vehicles, resulting in a larger number
of events, specially in the roundabout
scenarios;
• As soon as the heavy vehicles are re-
moved, the differences start to smear
out.
Controlling the sound environment 23
(a)
• The behaviour within this type of
analysis is rather different than in the
study of sound pressure level. The
implicit rule to yield to vehicles in
the roundabout results in a higher
congestion of certain parts of the
network, as these vehicles need a
larger gap to enter the roundabout,
turning it into a complex situation in
the case of high traffic flow.
• The signalised crossing maintains a
more constant behaviour through-
out the inclusion or exclusion of
(b) different vehicles types. In this sense,
research has appointed that the
presence of heavy vehicles led to
higher unpleasantness scores in the
roundabout cases.
(b)
Figure 11 - Equivalent
sound pressure level (a)
and number of events
above 60 dB (b) for all
study points and plaus-
ible scenarios
Figure 12 - (a) Case with dispersed green space pattern (left) (ANN=0.79),
(b) case with clustered green space pattern (right) (ANN=0.84).
THE EFFECT OF GREEN SPACE green space coverage itself. Small areas
PATTERN ON TRAFFIC NOISE with similar green space coverage and
different green space pattern exhibit dif-
DISTRIBUTION
ferent noise levels (Figure 12).
Green space becomes important in the
The same conclusions were found in
analysis of the sound environment. The
a city-scale approach. Eight average-
analysis of green space coverage itself
sized UK cities were investigated using
is sometimes not enough to be used as
a sample area of 30 km2. A prelimi-
a predictor for the average noise levels.
nary study is presented in Figure 13a,
Another issue is that frequently, cities
where the population density ranges
express an unequal spatial distribution of
between 4,300 and 7,200 residents/km2
their green space pattern depending on
with Brighton being the most densely
the morphology of the place and the so-
populated city. Regarding the car avail-
cioeconomic status of the area. Overall,
ability (Figure 13b), there are relatively
the green space pattern as a parameter
small differences among the cities and a
is sometimes more important than the
Controlling the sound environment 29
(a) (b)
Figure 17 – Description
of the relationship of
radial and linear forms
with: (a) Green space
coverage ratio,
(b) Noise levels
evidence for the cost of possible noise THE EFFECT OF SPECIFIC GREEN
interventions or the policy that should SPACE VARIABLES ON TRAFFIC
be followed. Characteristic examples are
NOISE
the ones referring to the Outline Zoning
To relate traffic noise prediction and the
Plans and the compatibility of different
effect of green space, two typical aver-
land uses.
age-sized UK cities were chosen as
Sheffield and Brighton. Sheffield, apart
from being a radial city, is also con-
Controlling the sound environment 33
Figure 18 - Com-
parison of the road
coverage in the two
cities in primary,
minor and local
roads
sidered one of the greenest cities in ated backyards or front yards, could be
England. On the other hand, Brighton a good predictor for traffic noise, since it
is a mid-size linear city with less natural contributed up to 38% in the Lden ex-
green spaces within the selected study planation for the current case studies
area. Both cities present similar charac- (see Figure 19). The combination of de-
teristics in terms of the road coverage in tached or semi-detached houses and
the different categories as presented in their gardens (Figure 20) reflects the
Figure 18. They also have similar values noise resistant areas, favoured also by
in the car availability per resident, which the small local roads around them and
allows for further comparison between their distance from the city centre. Such
them. results can be further used in the plan-
The analysis on the green spaces ning field, when designing residential
in these two typical UK cities has shown areas in combination with the traffic
that green areas, such as parks, urban management system. Another parameter
forests and gardens can affect traffic that was also proved significant is the
noise distribution. This was especially total number of cars, which in this partic-
noticeable in purely residential areas with ular case was negatively correlated with
high green space coverage. For example, the amount of green areas within the
the “Gardens ratio” referring to veget- same study sites (Figure 21).
Figure 20 - Example of
houses with vegetated
backyards or front yards in
(a) Sheffield, (b) Brighton
(a) (b)
Introduction
A key role in the acoustic planning Nowadays, prediction methods are
and design process of urban areas widely used in Europe for noise mapping
purposes in order to fulfil the EC noise
is played by the numerical predic- mapping requirements and, additionally,
tion of sound propagation. Being they are of large importance for eval-
able to simulate the sound field in uating the impact of noise control meas-
urban environments can be used to ures. Here, we make a distinction
between the prediction methods typic-
facilitate the decision-making and ally used for noise mapping, referred to
closing the communication gap as engineering methods, and the com-
among the diverse groups particip- putational urban acoustics methods,
ating in the planning process. which predict the urban sound field with
high accuracy by numerically solving the
governing physical equations after some
simplifications.
37
The predictions also form the basis for model, whereby the same propagation
auralisation purposes, i.e. making the modelling can be applied to different
urban environment audible in a virtual sound sources. The usually unavoid-
reality sense. It can be seen as the pro- able propagation effects of importance
cess of simulating a listening experience. to model are the distance decay, the
While developed originally for room ground reflection, and the air absorp-
acoustical purposes during the recent tion, as well as the Doppler effect in case
decades, auralisation of outdoor environ- of a moving source (or receiver). Other
ments has gained an increased interest propagation effects may involve reduc-
during the last decade. Being able to tion due to screening objects, reflection/
listen to a planned environment before it scattering in facades and other objects,
has been built is not only informative for focusing due to wind and temperature
decision makers and users at all levels, profiles, and wave distortion due to tur-
including the citizens, but can also be bulence in the air.
used as input for further computer aided Since the calculation of the propaga-
analysis and tools. tion effects may be numerically expens-
A sound environment can consist ive, it is of interest to try and simplify
of audible contributions from a large the physical modelling. For instance,
number of sources. The sources may be for auralisation puposes, the modelling
stationary in space, like a splashing water of a distant road (macro or mesoscale
fountain in a park or a humming bus background sounds) may be simplified if
on idle, or moving through space, like there is a more prominent sound event,
a flying seagull making its alarm call or e.g. a car passing by on a nearby local
the siren on a passing ambulance. The road (microscale foreground sound).
source signal of the auralisation can be a Simplifications leading to a reduction of
recorded or a synthesised sound. Due to the numerical cost may thus allow for
the large possible variation in output of a higher level of detail in the modelling
environmental sound sources, it is at- of the more prominent sound events,
tractive to use models for the synthesis enabling real-time auralisation of more
of the source signals rather than relying complex situations.
on recordings. Conventionally urban environmental
The sound propagation effects, noise is assessed with noise measure-
during the travel from the source to the ments and noise mapping software. Also,
listener, are separated from the source noise control measures are usually ex-
of PSTD models, and the parabolic equa- pass-by. Other projects within SONORUS
tion and the equivalent source method have used numerical simulations in their
have been used to characterize the ex- investigations. For example, at microscale
cess attenuation of intermediate canyons level the influence of the urban canyon
to obtain a correction factor for engin- shape has been investigated using the
eering methods. FDTD method and at macroscale level
SONORUS has contributed to fur- several SONORUS projects have worked
ther developing some of these numer- on combining noise control and urban
ical methodologies. For instance, a planning by using engineering meth-
novel hybrid method combining PSTD ods for noise and exposure assessment.
and DG has been implemented to allow More details about these projects can
the computation of arbitrary boundary be found in the “Controlling the sound
conditions and complex geometries environment” Chapter.
as shown in the schematic example of As emphasized above, wave based
Figure 2 and in an application of the methods (FDTD, PSTD, etc.) are more ac-
methodology to a 2D irregular shape, as curate than geometrical acoustics meth-
shown in Figure 3. Additionally, source ods, while being computationally heav-
directivity has been incorporated in PSTD ier. Furthermore, geometrical acoustics
by using spherical harmonics technique methods may become more appropriate
and currently these developments are at high frequencies where complex wave
used for the auralisation of inner city car based effects may be neglected.
b)
Auralisation
For the sound synthesis of the source, a certain gear and engine speed (rpm)
several methods exist. Synthesis can be of a car and can be reconstructed to a
achieved by determining the physical seemingly continuous sound stream. This
properties of a source, with possible way, a car with a variable speed across its
sub-sources, and determining their route can be given a sound with smooth
sound spectrum. Then noise according and controlled transitions. Aside from the
to these spectra can be generated and spectrum of the source, its temporal
updated. Spectral modelling for example, behaviour is generally also needed.
will construct a sound spectrum by Examples are amplitude modulations
adding frequency components. Sub- in the emission of wind turbines and jet
tractive synthesis on the other hand will engines as well as impulse-like sounds.
remove unwanted components from Directivity of a source refers to the
a noise spectrum. Another method is angular distribution of the sound field
synchronous granular synthesis, an idea generated by the source. The major
first conceived for musical purposes. sources of noise in urban environments,
Granular synthesis will acquire small road, rail and air traffic, have a directional
grains of sound, usually from recordings, character. A realistic auralisation of these
saved as a library for a source and picked noise sources in urban environments re-
up according to the preferred source quires taking into account these aspects
property. This has been proven useful in the prediction method.
for sources that contain cycles of opera- Another factor that affects the
tion, just like a car engine. For example, perceived sound field is the effect of the
a group of grains has been recorded for head, outer ears and torso of the listener.
This is referred to as the head related a PSTD simulated source are shown for
transfer function (HRTF) or head related three different octave bands.
directivity. An HRTF is the response that A valid auralisation tool will not
characterises how the human ear re- necessarily simulate accurately all of the
ceives the sound from a point in space. physical properties and processes
With the use of HRTFs a 3D sound ex- of an environment and its sources the
perience can be achieved with the use way they are taking place in reality, but
of only two audio channels (headphone through translating these will give useful
playback). By incorporating directivity results for the situation. Source models
and HRTFs in auralisation the quality of can be constructed by using psycho-
auralisation is significantly improved. acoustic properties of the human
Source directivity and HRTFs have hearing, for example taking into consid-
been incorporated in various geometrical eration temporal and frequency masking.
acoustics methods and computational In the SONORUS project, a method for
methods. Within the SONORUS project auralisation of background road traffic
a methodology to include source and has been developed, with the aim to
head related directivity in the pseudo- concentrate computational power to
spectral time-domain method (PSTD) has foreground events, e.g. a car passing
been developed. In Figure 5, directivity by on a local road where the listener is
patterns of an analytically derived and located. The approach uses modulation
transfer functions, i.e. rippled noise spec- account typical effects like spherical
tra that shift with time, which appears spreading and air absorption. Especially
to be a compact and promising way to important with aircraft auralisation
model a time varying noise event. Initial are the often strong Doppler shift and
listening tests have been carried out for ground effect. Also, while listening to
this approach and further work is ongo- aircraft noise, one can typically hear
ing. fluctuations that are relatively slow and
Aircraft are a major contributor random. These fluctuations are often due
to noise in urban environments. The to atmospheric turbulence. An important
amount of people exposed to aircraft contribution of the project was a model
noise increases every year due to urban to simulate the effects of atmospheric
densification and increasing flight move- turbulence on sound propagation result-
ments. Within the SONORUS project ing in more realistic sounding auralisa-
tools have been developed to simulate tions.
how it sounds when an aircraft flies over Figure 6 shows a spectrogram of an
an urban environment. The developed auralisation of an Airbus A320 taking off
tools can be used to study the impact of from Zurich Airport. Strong tonal com-
aircraft noise on humans. ponents can be observed during the first
The propagation model takes into seconds as the aircraft approaches the
More multichannel setups exist, where area, and the listeners can navigate
the most popular ones are higher order themselves within this acoustic field. Both
ambisonics (HOA) and wave field syn- HOA and WFS can be configured to cre-
thesis (WFS) rendering techniques. The ate a correct acoustic field in either a 2D
main difference between these two is plane or a 3D volume. It should be noted
that HOA has a sweet spot, although it that the headphone setup coupled with
can be expanded and controlled, where- HRTFs, HOA or WFS techniques, can be
as WFS, with its large number of loud- enhanced by using sensors tracking the
speakers, avoids the sweet-spot limita- position and rotation of a listener, and
tion. For the latter, virtual sound sources adjusting the output.
are spatially located within the listening Designing a subjective listening
What is (urban)
soundscape?
The approach to the evaluation The concept of ‘soundscape’ was origi-
and design of the urban sonic nally rooted in the music and acoustic
ecology research areas. It quickly ex-
environment can be tackled from panded to other disciplines, such as
different perspectives regarding the environmental noise, architecture, envir-
scale, the potential users of the onmental health, psychology, sociology
areas and the complementarity and urban studies, claiming for a holistic
approach to the way we conceive the
with other kinds of interventions. sound around us. Soundscape research
In this Chapter, an overview of the investigates how humans experience the
whole design process has been sonic environments and tries to establish
undertaken, from the establish- the relationships between the physical
world and the human response to it.
ment of the general objectives to It involves the cooperation of human,
the definition of the soundscape social and engineering sciences. Sound-
proposal. Examples of the applica- scape planning and design represents a
tion of the soundscape design can paradigm shift as it explores the diversity
of soundscapes across countries and
be found in the “Applied Urban cultures, and considers environmen-
Sound Planning” Chapter in the tal sounds as a ‘resource’ rather than a
sections referring to the cities of ‘waste’. Soundscape is a key approach in
Brighton, Rome and Antwerp. the urban sound planning process advo-
53
Example of a positive
soundscape transformation
Urban soundscape 55
setting the objectives. The urban sound- of the private home have some specific
scape design can be part of a wider demands. Here, the feeling of being in
intervention comprising different fields control of the sound environment be-
such as urbanism, community affairs or comes particularly important.
environmental protection, so in this case Different methodologies could be
the objectives must be compatible with envisaged to derive soundscape object-
a set of global objectives for the area. ives. The process could be part of design
For soundscape design to succeed, the competitions for (re)development of an
drawing up of these objectives should area. The question “How will it sound
involve the active engagement of the during different parts of the day or in
stakeholders from the outset of the de- different seasons?” can be answered to-
cision making process. gether with more usual questions such as
Urban soundscape methodolo- “What will it look like?” or “How will it be
gies have been mostly applied for open used?” The technology is there to
outdoor public spaces with a typical use. illustrate this vision to a broader audi-
For these, and in particular for parks and ence (see Chapter on “Prediction and
squares, literature on sound preference auralisation of urban sound environ-
has established different kinds of wanted ments”). Co-creation of the public space,
sound sources, such as natural (sound including its soundscape, by several
of rustling leaves, wind, water sounds, stakeholders together is certainly also an
birds…) and cultural sounds (bells, tra- option.
ditional festival sounds, sound marks…), Some examples of acoustic object-
contrasting the unwanted sound (me- ives include: “I want the sound of water
chanical and technological sounds). to be noticeable”, “I want the sound of
However, soundscapes along routes the traditional music of the festival to
connecting different functions of the city dominate completely” or “I do not want
should not be forgotten. Indeed, trans- to hear the noise of the trains passing
ition gives stronger impressions than a by”. The achievements of these goals at
sustained sonic environment as the the soundscape scale can lead to meet
human ear quickly adapts and filters. the global goals that influence people’s
Finally, objectives for the environment mood, behaviour and quality of life.
Urban soundscape 57
unexpected or incongruent sounds could could both be used. The main acoustic
also be more salient because they dis- strategies applied in the urban sound-
turb the overall expectations based on scape management and planning are
an earlier experience of the environment. listed below.
Therefore, it needs to be clear which vis- Noise control technologies (see
itors will visit the newly designed urban Chapter on “Controlling the sound en-
soundscape and what are the possible vironment”) are necessary tools for the
influences of their personal beliefs to noise abatement in soundscape inter-
perception. ventions. They could be applied at the
source, on the transmission path
between the source and receiver, or at
the receiver, and tend to mitigate the
negative effects of noise in the environ-
ment.
Masking techniques are often used
when it is not possible to adequately re-
duce the environmental noise with noise
control techniques. Energetic masking
may be applied in the cases in which
we want to avoid certain sources to be
heard. This energetic masking is intro-
duced via sounds that are created natur-
ally or artificially (i.e. by means of elec-
troacoustic devices). Attentional masking
DEFINING THE ACOUSTIC consist of the introduction of positive
STRATEGY sounds that attract the attention of vis-
Once we have established the objectives itors and decrease the perception of un-
of our intervention, we have to define wanted sounds. However, limitations
an acoustic strategy on how to achieve include environments with high back-
them. To define this strategy, evaluation ground noise levels, above 65 dB accord-
by listening to the existing or new scen- ing to some literature, where the noise
arios and prediction of the improvement levels should firstly be reduced through
(see Chapter on “Prediction and aural- noise control measures.
isation of urban sound environments”) A different approach can be undertaken
Urban soundscape 59
Analysis of the Example of a soundwalk
around a city
implemented urban
soundscape
Once the soundscape objective is trans- associated with a different perception
formed into a specific implementation, of soundscape. It is important to decide
several steps have to be observed to test how we are going to gather information
the characteristics of the changed envir- on urban soundscape, according to the
onment. different types of experience of the urban
environments. Previous soundscape
COLLECTING PERCEPTUAL DATA studies have used various methods and
There are many ways to experience the today the most common extend to:
acoustic environment. This can happen • Soundwalk
when we are actually on site, in a labor- • Interview
atory, or we can just recall it from • Behavioural observation
memory. Every such experience will be • Laboratory experiment
Every method will likely be related to part of the acoustic environment, so that
different ‘tools’ necessary to achieve the information later can be further pro-
the desired goals. In particular, all of the cessed with respect to its ‘perceptual’
mentioned methods except laboratory counterpart. Several techniques for
experiment have the advantage of being recording exist and the most commonly
administered in situ and to the people used today are: Ambisonics recording,
that have recently experienced the envir- binaural recording and recording with
onment in question. On the other hand, portable devices.
laboratory experiment brings the possib- Ambisonics recordings are used for
ility of having the controlled environ- capturing 3D sound environments. The
ment when investigating the particular equipment is placed usually in a station-
characteristics of the soundscape. To ary measurement position. Main advan-
summarize the main approaches used by tage of Ambisonics is the possibility of
researchers and practitioners, methods the listening to a real sound environment
and associated tools are presented in in laboratory conditions. However, this
Figure 2. implies that a listening room with specifi-
cally positioned loudspeakers has to
MEASURING PHYSICAL be created. Although used widely in
ENVIRONMENT research on spatial audio, the custom
Every time we collect soundscape data, it created solution is always necessary. An-
is crucial to measure also the ‘physical’ other advantage is the possibility to
convert the Ambisonics recordings to a
Urban soundscape 61
Recording of an
urban soundscape
using an artificial
head
range of available formats (stereo, Dolby tionary recording while the latter one is
formats, binaural). This gives the prac- used in mobile systems. Binaural format
titioners the possibility of having their provides an easy transition from meas-
recordings stored in a universal format urements to a listening experience since
with an easy conversion to any reproduc- the necessary equipment in the whole
tion format required. chain consists of microphone, recorder,
A widely used type of capturing playback device with the equalizer and
physical environment is recording the headphones. Finally, the commercial
binaural signals. They are based on the solutions both for recording and repro-
characteristics of human hearing where duction are easily accessible.
only two signals coming to each ear are Latest trends in soundscape research
enough to capture the spatiality of the focus on the use of portable measure-
sound. These recordings can be made ment devices. Their applications extend
using an artificial head or by placing the mostly to the acquisition of a sound in
microphones in or on the listener’s ears. dynamic and large urban environments.
The former type is usually used as sta- The measurement device is often set
Urban soundscape 63
is still in the early stages. However, the evaluation of the input sound. The model
possibilities that such models provide for can be used when (single-channel)
automatic analysis and long term mon- recordings from real or auralised sonic
itoring in the future should encourage environments are available. To the plan-
their use within the practitioners’ com- ners, its output would present a clear
munity. indication when the visitors of a recorded
place would likely pay attention to the
MODELS DEVELOPED DURING sound. As displayed in Figure 3, when
SONORUS PROJECT given a recording, the model marks
A new model for extracting auditory auditory saliency with a signal that peaks
saliency from environmental sound has where noticeable events are detected. As
been developed. The features of the a result, analysis of recordings of sound-
model are based on spectro-temporal scape could be greatly simplified by ex-
modulations found to induce direct tracting only the parts that are deemed
response in the human brain. Imple- salient and discarding the other sublim-
mented in the model are the procedures inal (i.e. non-noticeable) signal portions.
that mimic processing in human auditory This automatic procedure would find its
pathways, as well as the simulation of application in large sensor networks that
activation and inhibition of neurons in gather ‘big data’ and help the planners
the human brain. to focus on what is most important in
The output of the model represents the soundscape.
a single-number time-changing saliency
(b)
Figure 3 - Saliency
output from a traffic
noise with honk
sounds recorded
in Ghent (BE): (a)
original recording; (b)
recording with added
speech
Urban soundscape 65
Presenting information
about urban soundscape
Gathering data from several sources can LA10-LA90. In addition, psychoacoustic
imply that the structure and internal co- parameters related to the sound percep-
herence useful for the study needs to be tion can also be evaluated with examples
created. According to the goals, we have of: Loudness, Sharpness, Roughness and
to select and organize information into Fluctuation Strength.
more meaningful structures for repres- Spatial metrics are algorithms that
entation. define features of the landscape structure
on the basis of land use. In the analysis
“WHAT ARE WE GOING TO of urban soundscape, these metrics
REPRESENT?” are used to investigate the influence of
Most of the authors agree that the urban landscape on the perceived sonic
perception of the sonic environment is environment. Some other spatial and
influenced by the context in which the visual parameters can be used, such as
sounds are perceived and by the indi- the height of the buildings, distances
vidual differences. Regarding this gen- from the sound sources, and percentage
eral approach, we can represent acoustic, portion of elements perceived in an area
contextual (related with visual or spatial and finally visibility features.
information, odours, temperature, illu- Different measures for evaluation
minance…) and individual data (demo- of the sonic environment according to
graphic data, noise sensitivity, mood perceptual responses have been con-
scales…). Other classifications of the sidered. These include indicators related
represented data may be done regarding to soundscape quality, acoustic com-
the way of gathering the data – physical, fort and noise annoyance. To obtain a
perceptual or from computational detailed description of the sonic atmo-
models. sphere, semantic scales may be used to
In soundscape research the most evaluate people’s appraisals on sound-
frequently used acoustic variables related scape or to study how people experience
to the physical environment are Leq, an environment.
LAeq, LA10, LA50, LA90, LCeq-LAeq and
Urban soundscape 67
calculated with specific engineering ation can be cross validated with other
prediction methods (see the Chapter on predictive statistics.
“Controlling the sound environment”). These maps can also be intended
Representation of the sound propagation to show different perceptual informa-
of other artificial (engines, commercial tion than the acoustic one, for example
and leisure activities…) or natural sounds related with appraisals on the security of
(birds, water…) may also be used as com- a place, the beauty of landscape or the
plementary information in the sound- potential benefits to offer an alternative
scape evaluation. In contrast, maps with information to soundscape data, which
alternative information to that offered by can affect the overall environment per-
the traditional noise maps may be de- ception.
veloped, showing for example the classi- Finally, as people experience sound-
fications of areas regarding typologies of scape by visiting the space and listen-
soundscape or the distinctive spatial and ing (attending) to the sound, it is very
temporal variability of sound (Figure 4a). important for the planners and stake-
The perception of a single sound- holders to do the same. Although other
scape attribute from semantic differential methods we presented provide good
scales may also be represented using representations and general overview, to
maps (Figure 4b). Geo-statistical meth- experience soundscape by listening will
ods of interpolation, based on the con- provide a significant advantage to the
tinuous spatial variation of the same pat- planner. Therefore, in order to imagine
tern may be used for the representation a specific soundscape that matches an
of these variables. The main assumption environment – a goal that every urban
of interpolation is that the collected sound planner should aim to achieve –
data samples are correlated in space. listening in every possible occasion will
Such procedure can be understood as a provide not only the ideas but also the
prediction of the spatial behaviour of the feedback necessary to successfully ac-
variables, and the results of the interpol- complish this rewarding task.
Introduction
Worldwide, cities are expanding at well-planned infrastructure develop-
an unprecedented rate. Population ments are therefore needed to face this
challenge.
growth in city centres places un- The requirement to improve peo-
precedented demands on existing ple’s mobility within cities will increase,
city infrastructure systems that which inevitably affects the urban in-
are bringing many EU cities to a frastructure, resulting in the expansion
of the transportation networks. Conse-
breaking point. quently, negative impacts such as noise
and air pollution are expected to rise.
Traffic noise, people noise, pollution, In order to meet these challenges, the
poorly planned and managed places integration of urban and transport plan-
within our cities threaten the health and ning will be fundamental to the future
wellbeing for all of us. Large scale and of successful cities. The delivery of col-
69
laborative transport, public realm and of SONORUS to urban sound planning
regeneration projects can achieve this. relies on preventing the occurrence of
It has therefore never before been more noise, by not limiting the interventions to
important for technical, social and eco- the obvious noise engineering solutions
logical systems to work together. but to include a combined approach that
In such a dynamic environment, the coordinates actions of different acoustic
newly established discipline of Urban fields to provide an integrated solution.
Sound Planning combines synergies In this Chapter, we are summarizing
with the other planning processes in a examples where urban sound planning
unique way. The Urban Sound Planner is applied within the SONORUS project.
brings a different perspective and con- The intention of including practical cases
tributions to the process of delivering in the project is to develop exemplary
well-designed cities that work for people application to real case scenarios within
rather than harm them. This is achieved the urban sound planning approach. This
by improving the quality of sonic urban work aims to contribute to improving
environments, not simply trying to make the current situation and reversing the
things quieter, but proactively designing growth of poor urban sound envir-
to avoid noise generation and defining onments. Four different scenarios are
policies and strategies to value, introduce included, varying in scales, methods and
and preserve the characteristics of a outcomes: Rivierenhof Park in Antwerp
good sonic environment. (Belgium), Frihamnen area in Gothen-
Presently, the main objective of an burg (Sweden), Valley Gardens in
acoustic intervention is defined in the Brighton & Hove (UK) and the Colos-
regulations as a noise ceiling and is seum, Palatine and Roman Forum area in
usually considered after the urban plan Rome (Italy).
project is already decided, limiting the
opportunities of approaches to other DIFFERENT SCALES, DIFFERENT
than traditional engineering noise con- NEEDS, DIFFERENT TOOLS
trol. This approach is usually restricted The holistic approach has the ability to
in space (to the most exposed receivers) assess the sonic environment at different
and in time (short-term perspective), urban scales, from growing mega-urban
missing the opportunity to contribute developments to the small urban park.
to improved environments, sometimes Mega (more than 10 million inhabitants)
missing to take advantage of high- or Meta cities (more than 20 million
quality potentials. The holistic approach
2
4
Highway running
on the south side
of the park
sound sources. Therefore, such people ing that an increase in the amount of
hear these antagonizing sounds more trees within the parks can provide fur-
than the sounds that they actually want ther noise level attenuation (Figure 4). A
and expect to hear in a tranquil environ- second outcome is that the noise levels
ment. within the “tree coverage” are lower than
Rivierenhof Park is also the objec- the noise levels in the “grass coverage”.
tive of study for the interaction of green To continue with the application of
space parameters and sound levels. In urban sound planning, the study conti-
this case there is a negative correlation nued with the development of a model
between the tree coverage and the for human perception of environmental
average levels of L10 and L90, suggest- sounds and its translation to an artificial
Applied urban sound planning 77
sound perception model (see Figure 5). attentively to either of the sounds, as a
The current status of technology allows human park visitor would do.
a monitoring of cities with a high spa- To complement this study applying
tial resolution. However, the challenge an urban sound planning approach, the
starts within a soundscape approach that study goes beyond the analysis of the
analyses the person-environment inter- distribution of different activities within
action, i.e. the perception. The inclusion the park, with the aim to make it more
of such translation will strongly help in attractive to the users. The proximity of
the understanding and assessment of different paths, the accessibility of the
urban sound environments. Although park and different activities have been
the models will not be described in detail studied, as well as the quality of inte-
(for more information, check the “Urban gration in the city. It was concluded that
soundscape” Chapter), it is worth to the southern part of the park has more
mention that they describe and imple- chances to be visited than the northern
ment two basic listening styles: the first part due to the presence of water (lakes)
characterizes the holistic background lis- and the activities around it. Improving
tening experience and the second imple- accessibility, creating new paths, and en-
ments the analytic listening considering hancing the lake as the final destination
the person’s attention and the noticed could result in more visitors for these
sounds areas.
The model is expected to give statis-
tical information about individual per- Controlling the urban sound en-
ception. For example, mechanical sounds vironment: improvement of road
were almost constantly noticed next to layout
the busy roads. Human sounds appear Rivierenhof Park is divided by a 1 km
mostly in the centre area of the park. road with two traffic lanes, each with an
Natural sounds were often noticed in adjacent bicycle and pedestrian path.
the north area of the park. However, the The current urban layout decreases
model also includes an important por- pedestrian safety and interrupts the co-
tion of human sounds activation in areas hesion of the park as only four crossing
without many people during the mea- points are available. At the same time,
surement campaign. As a consequence the linear geometry of the road en-
of the implementation of attention hances vehicle acceleration. From the
processes, the model allows not listening aesthetic and visual point of view, it gives
the misleading impression to the visitor • Separating the two traffic lanes;
that this is the end of the park. Addition- • Reducing the number of lanes: redis-
ally, this road is a significant source of tribution of the traffic flow to other
road traffic noise, with LEq values around possible routes;
70 dBA. • Reducing traffic speed with calming
In this regard, the working group measures, such as the inclusion of
proposed the following solutions to chicanes along the road, known to
mitigate the problems generated by road reduce traffic speed. This measure will
traffic, reducing road traffic noise emis- also avoid the linear perception of
sion and increasing pedestrian safety and the road, giving visual continuity to
urban green space quality: the park;
(b)
• Adding a porous road surface mater- Most of the proposed solutions were
ial, reducing noise emission; acoustically calculated using a wave
• Locating vegetated low barriers next model (a Finite-Difference Time-Domain
to the source (see “Controlling the method, FDTD) (to know more, see the
urban sound environment” Chapter); Chapter on “Prediction and auralisation
• Including absorbent vegetated areas of urban sound environments”). The ad-
between source and receiver, reduc- dition of different shapes of low barriers
ing noise levels at the pedestrian was also assessed (see the “Controlling
paths while giving visual continuity to the urban sound environment” Chapter).
the park; The different noise abatement meas-
• Locating pedestrian paths and bike ures are included in two new scenarios:
lanes at a further distance from the scenario 1 (Sc1) keeping the two-lane
road. road, and scenario 2 (Sc2) modelling a
one-lane road (see Figure 6). Both scen-
arios are compared to the current situa-
sensorial (especially, the visual sense is This demonstrates that noise control
related to the auditory sense) and such in the context of soundscape design
interactions can have an important effect should not only consider reducing levels
on people’s noise perception. As an of unwanted sounds, but also improve
example, the effect of the noise barriers the audio-visual perception of the urban
is not only a noise reduction at the ear environment. Consequently, architects
of the pedestrian. It also partially hides and urbanists play an important role in
the sound source from sight, and the the perception of the urban environment
visual design of the barrier may help to and the participation of an urban sound
improve the overall perception of the planner is recommended as an integral
environment. part of future urban planning.
already shown in scenario 9 in Figure 12, is analysed. (See Figure 13, top, for equ-
(acceleration noise omitted) in compari- ivalent as well as peak levels.) To study
son with base scenario (1), with diffe- data through maps, roads are grouped
rences around 1-3 dBA for the selected into segments (here ca 150 segments in
receivers. Also, scenario 8, without heavy total).
vehicles, is giving differences in terms of In general, when assessing LAeq,1h
equivalent sound pressure level (1-2.5 the most favourable scenarios are 3, 5, 8
dBA). and 9. Speed reduction in the highway
With this type of tool is also possible (scenario 5) might be a good solution to
to see the contribution to the equivalent reduce noise levels in the majority of the
sound pressure level, for e.g. 15 minutes study points. Keeping only light vehicles
(LAeq,900s), from each road segment to (scenario 8) reduces levels up to 2.5 dBA.
a certain study point (e.g. receiver), cre- A study on time patterns should be
ating a kind of contribution noise map. made to give further information about
The same way, the largest LAeq,1s value the test site opportunities (to see more
during the period, here denoted Lpeak, about time patterns and number of
Applied urban sound planning 89
Figure 12 – Dynamic noise maps for the second simulated quarter hour
reflect the equivalent sound pressure level for the second quarter hour
(LAeq,900s)
Map of Brighton
(Valley Gardens 7
marked in blue)
6
A narrow strip
of the park
surrounded by
traffic
strategies has been proposed in order the majority of the study points are the
to assess potential benefits of an overall banning of heavy vehicles and the intro-
solution aimed at improving the Valley duction of a continuous absorbing noise
Gardens sound environment. For this, 15 barrier around the park. However, the
study points were selected (see Figure 16 noise barrier solution will need a fur-
and Table 1). ther study in order to be adapted to the
Different noise mitigation actions several particularities of the urban layout,
were proposed and discussed with the such as crossings, bike lanes, etc.
city planners. The most effective ones for
02
00
Cases
-02
1
-04 2
3
-06
4
-08 5
6
-10
-12
-14
Case Description
1 No reflections from buildings
4 No heavy vehicles
5 Buses on the West bound and remaining traffic on
the East bound
6 All traffic to the East bound
SPL
(dBA)
Time (s)
Provide positive Perceived sound envir- • Introduce ‘natural’ sounds through elements such
soundscapes onment as bad and not as wind in vegetative foliage or flowing water
appropriate to the site • Allow for ‘artificial’ sounds via infrastructure such as
new lampposts
• Encourage bird song by adequate planting
• Encourage positive activity such as children’s play
Accessibility and sound • The passage through the area should be attractive
attractiveness: access to and accessible. For this, the city could use sound
both the city centre and to connect the space as a kind of heritage, reflect-
the north area is one of ing the possibilities to keep and recall its past as a
the key sets in this project former industrial area and harbour.
• Build a landmark/soundmark throughout the
piers. Taking advantage of the positive sounds
that water features may bring to Frihamnen, in-
corporating the waterfront history (e.g. sounds of
waves on resting boat hulls), floating bridges and
shipyard sculptures as a variation of the sound en-
vironment as well as different pavements capable
to reduce attention to road traffic noise.
Park area and piers: to • The park activities could be oriented according to
cope with the uses and the noise exposure of the area. Possible functions
functions of this area as might be a recreational park with a large number
a park, special acoustic of activities e.g. concerts, playground, sports, etc.
qualities are needed. The • Topography as an alley through the construction
area will be submitted to of a railroad-oriented slope, incorporating the
high noise levels during attractive idea to Gothenburg citizens of a running
the construction period, track in the park, which could block the noise from
which will be around 20 the northern infrastructures
years.
Economic aspects The impact of the low • Careful study on the impact of future measures
attractiveness of the area intended to act only at a single city system. For
due to its high levels of example, the traffic can be planned as a deterrent,
noise providing a design difficult to drive through.
However, this approach might drive the area and
its surroundings to a higher noise exposure as
a consequence of the spread of traffic and the
increase in the distance travelled.
Figure 22 – Work-
ing process and
outcomes. Focus
on soundscape
design.
Test site S W O T
City
Several professional Indirect contact with Interventions made Sensitive areas related
Antwerp backgrounds in the the city planning previous to urban to noise annoyance
working group office decisions, observation with possible public
of consequences due opinion and news me-
to the short-term dia repercussions
accomplishment
Indirect participation Acoustic interven- Authorities are aware The group is not an
Rome within the first phase tions are limited due of the noise problem active part in the de-
of the decision to the protective and are open to intro- cision making process
making process legislation, despite duce noise mitigation
that high acoustic actions to improve the
standards are de- sonic quality of the
fined for the site area
117
118