whole

River-aquifer interactions in a semi-arid environment
stressed by groundwater abstraction
Author:
McCallum, Andrew Murray
Publication Date:
2012
DOI:
https://doi.org/10.26190/unsworks/16111
License:
https://creativecommons.org/licenses/by-nc-nd/3.0/au/
Link to license to see what you are allowed to do with this resource.
Downloaded from http://hdl.handle.net/1959.4/52606 in https://

unsworks.unsw.edu.au on 2024-07-13
River-aquifer interactions in a semi-arid
environment stressed by groundwater abstraction
Andrew Murray McCallum

(2250887)
A Thesis submitted for the Degree of

Doctor of Philosophy
School of Civil and Environmental Engineering

Water Research Laboratory
Supervisors:
Martin S. Andersen
R. Ian Acworth
3 September 2012 · Sydney · Australia

PLEASE TYPE
THE UNIVERSITY OF NEW SOUTH WALES
Thesis/Dissertation Sheet
Surname or Family name: McCallum
First name: Andrew Other name/s: Murray
Abbreviation for degree as given in the University calendar: PhD
School: Civil and Environmental Engineering Faculty: Engineering
Title: River-aquifer interactions in a semi-arid environment

stressed by groundwater abstraction
Abstract 350 words maximum: (PLEASE TYPE)
The connec ons between rivers and aquifers need to be understood and quan fied for successful water management. However, the interac ons are complex
and remain poorly understood in many catchments worldwide. To advance our understanding of river-aquifer interac ons in semi-arid environments stressed
by groundwater abstrac on, the Maules Creek Catchment, Murray Darling Basin (Australia) is inves gated.
First, an analysis of exis ng records (precipita on, evapotranspira on, river flow, groundwater levels, and groundwater abstrac on) shows that the
groundwater levels have predominantly declined due to groundwater abstrac on rather than changes in the climate. Furthermore, the river has altered from
being gaining to losing during low-flow periods, with a significant delay in this transi on since the onset of abstrac on.
Next, two novel methods are presented: (1) using pairs of temperature me series at the river-aquifer interface to quan fy the interac ons and thermal
diffusivity; (2) using hydrographs recorded at the upstream and downstream ends of a river reach to es mate the total recharge and rate of recharge. Both
methods offer several advantages over previously developed approaches.
Finally, the spa al and temporal variability in the interac ons are determined. It is found that high and low river flows correspond to high and low riverbed
interac ons, which were strongly losing during high-flows, and slightly losing, neutral, or gaining under low-flows. It is demonstrated that episodic high-flow
events account for a majority of the total groundwater recharge. It is suggested that differences in permeability between river bed/banks and the floodplain
control the efficiency of recharge from the river. The results highlight that up-scaling of the interac ons, especially during high-flows, is problema c.
Implica ons for the modelling and management of water resources are discussed throughout.
The process understanding gained, and methods developed, can be applied to other semi-arid environment stressed by groundwater abstrac on worldwide.
Declaration relating to disposition of project thesis/dissertation
I hereby grant to the University of New South Wales or its agents the right to archive and to make available my thesis or dissertation in whole or in
part in the University libraries in all forms of media, now or here after known, subject to the provisions of the Copyright Act 1968. I retain all
property rights, such as patent rights. I also retain the right to use in future works (such as articles or books) all or part of this thesis or dissertation.
I also authorise University Microfilms to use the 350 word abstract of my thesis in Dissertation Abstracts International (this is applicable to doctoral
theses only).
3 September 2012
…………………………………………………………… ……………………………………..……………… ……….……………………...…….…

Signature Witness Date
The University recognises that there may be exceptional circumstances requiring restrictions on copying or conditions on use. Requests for
restriction for a period of up to 2 years must be made in writing. Requests for a longer period of restriction may be considered in exceptional
circumstances and require the approval of the Dean of Graduate Research.
FOR OFFICE USE ONLY Date of completion of requirements for Award:
THIS SHEET IS TO BE GLUED TO THE INSIDE FRONT COVER OF THE THESIS

Abstract
The connections between rivers and aquifers need to be understood and quantified
for successful water management. However, the interactions are complex and remain
poorly understood in many catchments worldwide. To advance our understanding of
river-aquifer interactions in semi-arid environments stressed by groundwater abstrac-
tion, the Maules Creek Catchment, Murray Darling Basin (Australia) is investigated.
First, an analysis of existing records (precipitation, evapotranspiration, river
flow, groundwater levels, and groundwater abstraction) show that the groundwater
levels have predominantly declined due to groundwater abstraction rather than
changes in the climate. Furthermore, the river has altered from being gaining to
losing during low-flow periods, with a significant delay in this transition since the
onset of abstraction.
Next, two novel methods are presented: (1) using pairs of temperature time series
at the river-aquifer interface to quantify the interactions and thermal diffusivity; (2)
using hydrographs recorded at the upstream and downstream ends of a river reach
to estimate the total recharge and rate of recharge. Both methods offer several
advantages over previously developed approaches.
Finally, the spatial and temporal variability in the interactions are determined.
It is found that high and low river flows correspond to high and low riverbed inter-
actions, which were strongly losing during high-flows, and slightly losing, neutral, or
gaining under low-flows. It is demonstrated that episodic high-flow events account
for a majority of the total groundwater recharge. It is suggested that differences in
permeability between river bed/banks and the floodplain control the efficiency of
recharge from the river. The results highlight that up-scaling of the interactions,
especially during high-flows, is problematic. Implications for the modelling and
management of water resources are discussed throughout.
The process understanding gained, and methods developed, can be applied to
other semi-arid environment stressed by groundwater abstraction worldwide.
i
Contents
Abstract i
List of Figures v
List of Tables vi
Research Outcomes ix
1 Introduction 1
1.1 Outlook . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.3 Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
2 Groundwater abstraction and interactions 7

2.1 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
2.2 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
2.3 Study area . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
2.4 Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
2.4.1 Hydrological stresses . . . . . . . . . . . . . . . . . . . . . . . 14
2.4.2 Hydrological impacts . . . . . . . . . . . . . . . . . . . . . . . 15
2.4.3 Linking groundwater gradients and river low-flows . . . . . . . 17
2.5 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
2.5.1 Hydrological stresses . . . . . . . . . . . . . . . . . . . . . . . 18
2.5.2 Hydrological impacts . . . . . . . . . . . . . . . . . . . . . . . 21
2.5.3 Linking groundwater gradients and river low-flows . . . . . . 28
2.6 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
2.6.1 Groundwater abstraction and changes in groundwater levels . 30
2.6.2 Changes in river flows . . . . . . . . . . . . . . . . . . . . . . 32
2.6.3 Relationship of groundwater declines to river flow losses . . . 35
2.6.4 Summary of processes . . . . . . . . . . . . . . . . . . . . . . 36
2.6.5 Significance for resource management . . . . . . . . . . . . . 37
2.7 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
3 Estimating interactions using riverbed temperature time series 41

3.1 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
3.2 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
3.3 Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
3.3.1 Theoretical background . . . . . . . . . . . . . . . . . . . . . . 43
iii
Contents
3.3.2 Proposed method . . . . . . . . . . . . . . . . . . . . . . . . . 46

3.3.3 Field example . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
3.4 Results and Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . 48
3.4.1 Theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
3.4.2 Interactions in Maules Creek Catchment . . . . . . . . . . . . 50
3.4.3 Method comparison . . . . . . . . . . . . . . . . . . . . . . . . 53
3.5 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
4 Estimating interactions using differential river gauging 61

4.1 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
4.2 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
4.3 Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
4.3.1 Theoretical background . . . . . . . . . . . . . . . . . . . . . . 63
4.3.2 Proposed method . . . . . . . . . . . . . . . . . . . . . . . . . 64
4.3.3 Field example . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
4.4 Results and Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . 68
4.4.1 Steps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
4.4.2 Recharge in Maules Creek Catchment . . . . . . . . . . . . . 71
4.4.3 Method limitations . . . . . . . . . . . . . . . . . . . . . . . . 74
4.5 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
5 Spatial-temporal variation in interactions 77

5.1 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
5.2 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
5.3 Study area . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
5.4 Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
5.4.1 Point measurements from temperature time series . . . . . . . 81
5.4.2 Reach measurements from differential gauging . . . . . . . . . 85
5.5 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
5.5.1 Point measurements from temperature time series . . . . . . . 87
5.5.2 Reach measurements from differential gauging . . . . . . . . . 91
5.6 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
5.6.1 Influence of river flow events on groundwater recharge . . . . . 98
5.6.2 Influence of channel geomorphology on groundwater recharge . 99
5.6.3 Role of riverbed sediment heterogeneity and hyporheic flow . . 102
5.6.4 Implications for numerical modelling and resource management104
5.7 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
6 Conclusion 107
6.1 Modelling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107
6.2 Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
Bibliography 111
iv
List of Figures
2.1 Maules Creek study area . . . . . . . . . . . . . . . . . . . . . . . . . 11

2.2 Historical groundwater flow direction . . . . . . . . . . . . . . . . . . 13
2.3 Monthly precipitation and evapotranspiration time series . . . . . . . 19
2.4 Distribution of groundwater abstraction and head differences . . . . . 21
2.5 Representative observation bore hydrographs . . . . . . . . . . . . . . 23
2.6 Abstraction and groundwater head relationships . . . . . . . . . . . . 24
2.7 River flows I . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
2.8 River flows II . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
2.9 River-aquifer gradients . . . . . . . . . . . . . . . . . . . . . . . . . . 29
2.10 River-aquifer gradient and fluxes . . . . . . . . . . . . . . . . . . . . 30
2.11 Conceptualized river-aquifer interactions . . . . . . . . . . . . . . . . 37
3.1 Schematic of a temperature array installed in a riverbed . . . . . . . 49

3.2 Theory of using temperature data . . . . . . . . . . . . . . . . . . . . 50
3.3 Field data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
3.4 Field data interpretation . . . . . . . . . . . . . . . . . . . . . . . . . 52
3.5 Comparison of methods I . . . . . . . . . . . . . . . . . . . . . . . . . 54
3.6 Comparison of methods II . . . . . . . . . . . . . . . . . . . . . . . . 56
4.1 Namoi Catchment study area . . . . . . . . . . . . . . . . . . . . . . 66

4.2 Schematic geological cross section . . . . . . . . . . . . . . . . . . . . 66
4.3 Photographs of study area . . . . . . . . . . . . . . . . . . . . . . . . 67
4.4 Method steps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
4.5 Recharge results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
5.1 Study area . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80

5.2 Temperature array deployments . . . . . . . . . . . . . . . . . . . . . 84
5.3 Example of temperature data . . . . . . . . . . . . . . . . . . . . . . 88
5.4 Results for pool scale . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
5.5 Results for reach scale . . . . . . . . . . . . . . . . . . . . . . . . . . 92
5.6 Flow gauge data I . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
5.7 Flow gauge data II . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
5.8 Flow gauge data III . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
v
List of Tables
2.1 Precipitation statistics . . . . . . . . . . . . . . . . . . . . . . . . . . 19

2.2 Evapotranspiration statistics . . . . . . . . . . . . . . . . . . . . . . . 20
2.3 River flow statistics . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
3.1 Monte Carlo analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
5.1 River-aquifer statistics . . . . . . . . . . . . . . . . . . . . . . . . . . 90

5.2 Analysis of high-flow events . . . . . . . . . . . . . . . . . . . . . . . 95
vii
Research Outcomes
Conference Presentations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
McCallum, A. M., Andersen, M. S., Kelly B. F. J., Giambastiani, B. M. S., and Ac-
worth, R. I. (2009), Hydrological investigations of surface water–groundwater
interactions in a sub-catchment in the Namoi Valley, NSW, Australia, Oral
presentation at Joint IAH/IAHS International Convention, Hyderabad, India,
7-12 Sept 2009.
McCallum, A. M., Andersen, M. S., Rau, G. C., and Acworth, R. I. (2010), In-
vestigation of surface water groundwater interactions and temporal variability
of streambed hydraulic conductivity using streambed temperature data, Oral
presentation at XXXVIII - IAH Congress, Krakow, Poland, 12-17 Sept 2010.
McCallum, A. M., Andersen, M. S., Rau, G. C., and Acworth, R. I. (2011), Using
Combined Temperature, Flow and Level Data to Investigate River-Aquifer
Interaction Scaling Issue, Oral presentation at AGU Fall Meeting 2011, San
Francisco, USA, 5-9 Dec 2011.
Journal Papers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
McCallum, A. M., Andersen, M. S., Rau, G. C., and Acworth, R. I. (2012), A 1D
analytical method for estimating surface water groundwater interactions and
effective thermal diffusivity using temperature time series, Water Resources
Research.
McCallum, A. M., Andersen, M. S., Giambastiani, B. M. S., Kelly B. F. J., and
Acworth, R. I. (2013a), River-aquifer interactions in a semi-arid environment
stressed by groundwater abstraction, Hydrological Processes.
McCallum, A. M., Andersen, M. S., and Acworth, R. I. (2013b), Estimating recharge
to unconfined aquifers using differential river gauging, Groundwater.
McCallum, A. M., Andersen, M. S., Rau, G. C., Larsen J. R., and Acworth, R. I. (in
prep.), River-aquifer interactions in a semi-arid environment established using
point and reach measurements.
ix
Chapter 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Introduction
1.1 Outlook
The hydrological cycle has been studied for millennia. In the first century BC Marcus
Pollio, for example, wrote of such concepts as evaporation, condensation, infiltration,
and springs [Acworth, 2009]. The interactions between rivers and aquifers are an
essential part of the hydrological cycle generally [Winter et al., 1998; Jones and
Mulholland, 2000] and within semi-arid environments in particular [Nemeth and
Solo-Gabriele, 2003]. However, these interactions have generally been overlooked,
and rivers and aquifers have often been considered as separate entities [Acworth,
2009].
Consequently, rivers and aquifers have traditionally been managed as separate
resources [Braaten and Gates, 2003]. Over the last two decades, however, a paradigm
shift has been taking place which views rivers and aquifers as parts of a continuum
[Bencala, 1993]. This has been especially due to the seminal works of Winter et al.
[1998], Woessner [2000] and Sophocleous [2002]. There is thus a growing awareness
that rivers and aquifers should be jointly managed for water quantity [e.g. Winter
et al., 1998], water quality [e.g. Postma et al., 2007], and ecological reasons [e.g.
Brunke and Gonser, 1997]. Added to this are the facts that irrigated areas have
increased exponentially globally in the last century [Biemans et al., 2011] and that
water used for irrigation now accounts for about 70% of the global water withdrawal
1
Chapter 1. Introduction
[Wada et al., 2012], thus there is a now real impetus for beneficial change in the
outlook of water management. Accordingly, there is a global move towards a legal
framework that considers rivers and aquifers as two interconnected parts of the one
resource [e.g. WFD, 2008].
A good understanding of river-aquifer interactions is therefore clearly essential
for good water resource management. However, despite this knowledge being of
fundamental importance, the interactions between rivers and aquifers remain poorly
understood in many catchments worldwide [Ivkovic et al., 2009]. Consequently, the
study of river-aquifer interactions is a rapidly growing area of research [Fleckenstein
et al., 2010].
1.2 Overview
The aim of this section is to give an overview of the research to date on how surface
water (SW) and groundwater (GW) are conceptually understood to interact. Whilst
the basic concepts are first covered, the focus in this section is on the complexity
involved and the scientific questions which remain unanswered. Further literature
is reviewed in the following chapters.
Movement of SW and groundwater GW is largely controlled by physiography
(i.e., geomorphology and geology) and climate (i.e., precipitation and evapotran-
spiration) [Winter, 1999]. As a result of these factors, local, intermediate, and
regional flow systems are formed in the groundwater [Toth, 1962, 1963]. Where
groundwater intersects a surface water body (e.g., river, lake, wetland, and so on)
interactions take place.
Fundamentally, there are only two basic ways in which rivers and aquifers inter-
act: rivers can gain water by inflow through the riverbed/bank from the aquifer or
lose water by outflow though the riverbed/bank to the aquifer. These are termed
“gaining systems” and “losing systems”, respectively. For a river to be gaining, the
elevation of the aquifer water table near the river must be higher than the elevation
2
of the river stage. For a river to be losing, the opposite must be true. A river may
also be gaining on one side and losing on the other, termed a flow-through system
[Woessner, 2000].
Losing systems can be either connected or disconnected from the aquifer. The
disconnected system (where there is an unsaturated zone beneath the SW body)
may be completely disconnected or transitional [Banks et al., 2011]. In the first
instance, changes in the water table do not affect the infiltration rate, while in
the transitional system the infiltration rate is affected by the depth of the water
table. Thus a transition region describes a transition from a system where the flux
is directly proportional to the hydraulic gradient to a system where the flux is only
dependent on the river stage [Brunner et al., 2009a,b]. The term “disconnected”
can be misunderstood to imply that no water exchange is occurring; within the
Australian context there is a move to replace it with “non-contiguous connected”
which seeks to imply that there is still exchange.
The interaction of rivers and aquifers can also vary spatially and temporally. A
river may entirely and persistently be gaining or losing, but it can also have some
reaches that are gaining and other reaches that are losing. In addition one reach can
be gaining at one point in time and losing at another. These temporal changes in
flow direction can be caused by hydrological stresses such as flooding events and/or
seasonal groundwater abstraction [Winter et al., 1998; Theis, 1941].
In the case of flooding, or large flow events, caused by precipitation or release
of water from upriver reservoirs, two other processes occur. The rapid rise in the
river stage causes water to flow into the riverbanks, leading to a temporary storage
of water in the adjacent unsaturated zone. This loss of water from the river causes
a reduction in the flood peak. Depending on the relative river aquifer levels a
proportion of the water stored in the bank will be slowly released back into the
river system once the flood peak has passed and the river stage recedes (i.e., “bank
storage”). In the case that the event causes water to overtop the river banks, river
3
water will recharge groundwater throughout the flooded area. There have not been
many studies evaluating the significance of intermittent river flows on river-aquifer
interactions, such as in semi-arid environments, where rainfall events, recharge, and
flow events are erratic [Xi et al., 2010].
The other important source of system perturbation is groundwater abstraction.
Extracting water from aquifers causes a decline in groundwater levels surrounding
the abstraction well. The abstraction results in the capture of some of the ground-
water which would potentially have discharged into the river. Direct flow from the
river to the aquifer is induced if the abstraction is large enough. The sum of these
two effects is termed “streamflow depletion” [Sophocleous, 2002].
Whilst it can be of importance to understand the connectivity at a particular
place of potential impact, it is also important to know how these are linked in
contiguous river reaches from the catchment headwaters to the discharge point at
the sea [Banks et al., 2011]. Water management policies sometimes suggest that
groundwater development will not impact the river flows if a river and aquifer are
disconnected. However, while that may be true at one location, the groundwater
development could potentially expand the length of river that is disconnected, thus
increasing the overall loss rate [Brunner et al., 2011]. How the connection state
varies along a river reach in response to groundwater abstraction is therefore an
important research topic.
Furthermore, the mechanisms of interaction are not always straightforward.
Rivers can gain water in different ways: directly through the riverbed, as superficial
flow from the flood plain, or as a combination of these. The second is likely to occur
where a wide floodplain receives diffuse recharge, which then becomes superficial
flow and enters the river via the banks [Langhoff et al., 2006]. There are various
mechanisms of river-aquifer interaction; in particular, seepage from the aquifer can
occur through the riverbed or riverbanks. Shanafield et al. [2010] showed that flux
from the banks can account for up to 50% of the total flux.
4
Not only is there complexity at the large scale (e.g., reach or catchment scale),
but it exists also on the smaller scale (i.e., “Hyporheic exchange”). Interactions
can occur at a smaller scale due to bed forms and river curvature [Cardenas et al.,
2004]. Currents interact with bed topography or channel sinuosity which leads to
head gradients (e.g., across point bars) which cause SW-GW exchange to occur
[Cardenas, 2008, 2009]. Between the large (river reach) and small (riverbed) scales,
a river reach may gain water overall, and lose water coincidentally at the smaller
spatial scale [McCallum et al., 2012b]. Consequently, de Vries and Simmers (2002)
state that how point, reach and catchment scales relate remains an open question
which requires further investigation.
Finally, another level of complexity arises because of the heterogeneous nature
of fluvial deposits [Woessner, 2000]. Hydrogeology generally has been forced to deal
with heterogeneous geologic formations [e.g. Fleckenstein and Fogg, 2008]. This is
obviously important for river-aquifer interactions [Cardenas et al., 2004], however,
how best to deal with this question is not yet clear.
To summarise, river-aquifer interactions occur at different spatial and temporal
scales, which superimpose on one another, causing dynamic and complex patterns
of connectivity [Angermann et al., in press]. Consequently, as stated by Sophocleous
[2002], the spatial-temporal dynamics of river-aquifer interactions in different geo-
morphological settings needs further elucidation. Fundamental questions of pro-
cesses understanding yet remain.
1.3 Objectives
In light of the existing body of research, the objective of this thesis is to advance
the understanding of river-aquifer interactions in semi-arid environments stressed
by groundwater abstraction. An “observational science” approach is used in a
specific catchment, from which conclusions can be generalised and applied to other
catchments worldwide.
5
The main objective is addressed in four technical chapters, each of which has a
separate sub-objective:
– Chapter 2: Understand the links between groundwater abstraction and river-

aquifer interactions along a reach of the Namoi River in semi-arid eastern
Australia.
– Chapter 3: Develop a new methodology for using heat as a tracer to investigate

river-aquifer interactions.
– Chapter 4: Develop a new methodology for using river hydrographs to invest-

igate river-aquifer interactions.
– Chapter 5: Understand the spatial-temporal variation in river-aquifer interac-

tions along the studied reach of the Namoi River.
The thesis thus consists of an interplay of process understanding and method de-
velopment. Chapters 2 and 5 focus on process understanding, while chapters 3 and
4 focus on method development. Each of the technical chapters has been written
as a stand-alone journal paper. The first three have been published in Hydrological
Processes, Water Resources Research, and Groundwater, respectively, and the fourth
is presently a draft manuscript which will be published in due course. I am the
first author of each. The thesis is thus essentially a compilation of four journal
papers which together tell the story of “River-aquifer interactions in a semi-arid
environment stressed by groundwater abstraction”.
6
Chapter 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Groundwater abstraction and in-

teractions
2.1 Summary
Rivers and aquifers are, in many cases, a connected resource and as such the
interactions between them need to be understood and quantified for the resource to
be managed appropriately. The objective of this chapter is to advance the under-
standing of river-aquifer interaction processes in semi-arid environments stressed by
groundwater abstraction. This is performed using data from a specific catchment
where records of precipitation, evapotranspiration, river flow, groundwater levels,
and groundwater abstraction are analysed using basic statistics, hydrograph analysis
and a simple mathematical model to determine the processes causing the spatial and
temporal changes in river-aquifer interactions. This combined approach provides a
novel but simple methodology to analyse river-aquifer interactions, which can be
applied to catchments worldwide.
The analysis revealed that the groundwater levels have declined (~ 3 m) since the
onset of groundwater abstraction. The decline is predominantly due to the abstrac-
tion rather than climatic changes (r = 0.84 for relationship between groundwater
abstraction and groundwater levels; r = 0.92 for relationship between decline in
groundwater levels and magnitude of seasonal drawdown). It is then demonstrated
7
Chapter 2. Groundwater abstraction and interactions
that, since the onset of abstraction, the river has changed from being gaining to
losing during low-flow periods, defined as periods with flow less than 0.5, 1.0 or 1.5
GL/day (1 GL/day = 1 x 106 m3 /day). If defined as < 1.0 GL/day, low-flow periods
constitute approximately 65% of the river flows; the periods where the river is losing
at low-flow conditions are thus significant. Importantly, there was a significant delay
(> 10 years) between the onset of groundwater abstraction and the changeover from
gaining to losing conditions. Finally, a relationship between the gradient towards
the river and the river flow at low-flow is demonstrated. The results have important
implications for water management as well as water ecology and quality.
2.2 Introduction
Rivers and aquifers are, in many cases, a connected resource and as such the
interactions between them need to be understood and quantified for the resource to
be managed appropriately. There are two basic ways in which rivers and aquifers
interact: rivers gain water by inflow through the riverbed/bank from the aquifer or
lose water through the riverbed/bank to the aquifer, depending on the respective
river stage and aquifer head. A river may entirely and persistently be gaining
or losing but may also have reaches that are gaining and other reaches that are
losing. In addition, one reach can be gaining at one point in time and losing at
another [Winter et al., 1998; Woessner, 2000]. With increasing demands worldwide
on water resources, it is becoming apparent that we need to consider a conjunctive
management approach based on a quantitative process understanding of these river-
aquifer connections [Fullagar et al., 2006; Ivkovic et al., 2009].
The understanding of how groundwater abstraction interacts with surface water
systems has been developed over many decades. Theis [1941] and Glover and Balmer
[1954] were the first to consider the effects of a well on a nearby river. Later, this was
further refined by Hantush [1965] who included a semi-permeable barrier between the
river and aquifer representing the riverbed, and Hunt [1999] who produced a solution
8
to the case where the river is only partially penetrating the aquifer. However,
although these analytical solutions provide insights into river-aquifer interactions,
they are limited in their applicability by their conceptual model of river, aquifer and
abstraction well. Because of this simplicity, they cannot include the influences of
varying recharge, variable transmissivity, other outlets and inlets for the aquifer, and
so on. [Rushton, 2002]. This fact likely lies behind the ambivalence in the literature
as to the usefulness of such models when applied to specific catchments [Kollet
and Zlotnik, 2003; Fox, 2004] and the rapid development of increasingly complex
analytical models [e.g., Chen, 2003; Fox and Durnford, 2003; Zlotnik, 2004] as well
as the alternative use of numerical models for site to catchment scale studies [e.g.,
Sophocleous et al., 1995; Zume and Tarhule, 2008; Sanz et al., 2011]. The fact that
there can be substantial difference between field data and analytical or numerical
models [e.g., Sophocleous et al., 1995; Kollet and Zlotnik, 2003] highlights the need
for a field-driven approach [e.g., Mair and Fares, 2010; Tao et al., 2011] where the
hydrology of catchments is to be understood and the impacts of management policy
predicted.
These limitations are particularly pertinent when considering semi-arid catch-
ments. Critically, the models often treat the river as perennial which is clearly not
a valid assumption for such environments. Furthermore, there have not been many
studies evaluating the significance of intermittent river flows on river-aquifer interac-
tions for semi-arid regions [Xi et al., 2010]. Additionally, although it is acknowledged
that groundwater abstraction has led to a reduction of baseflow in different semi-arid
catchments worldwide [Winter et al., 1998], a broader understanding of river-aquifer
interactions seems to be lacking. Fundamental resource management questions such
as: ’What have been the impacts of historical groundwater abstraction on river
flows? Can the impacts be quantified to consider water scarcity and ecosystem
health? How do the impacts vary with different levels of groundwater abstraction?
What role does climate play in influencing the impacts? Are current surface water
9
and groundwater allocations sustainable?’, largely remain unanswered [Ivkovic et al.,

2009].
The objective of this study is to advance the understanding of river-aquifer
interaction processes in semi-arid environments stressed by groundwater abstraction.
A field-driven approach is taken using data from a specific catchment where records
of precipitation, evapotranspiration, river flow, groundwater levels, and groundwater
abstraction are analysed using statistics, hydrograph analysis and a simple math-
ematical model to determine the processes causing the spatial and temporal changes
in river-aquifer interactions. Although these methods individually are not new
their combined application provides a novel methodology to analyse river-aquifer
interactions, which can be applied to catchments worldwide.
2.3 Study area
Maules Creek Catchment (surface area of approximately 1500 km2 ) is located in the
semi-arid region of north western New South Wales, Australia (Figure 2.1), and is
a sub-catchment of the much larger Namoi Catchment. The plains of the Maules
Creek Catchment are 200 to 250 m above sea level, and the mountain range has
peaks rising to 1500 m at Mount Kaputar. The Namoi River runs from south to
north through the catchment. Maules Creek is the main tributary and is generally
ephemeral, although it does have perennial flow in its mid-reach [Rau et al., 2010].
Because of its intermittent nature, Maules Creek only contributes flow to the Namoi
River during infrequent floods [Sinclair et al., 2005; Andersen and Acworth, 2009].
10
Figure 2.1: Maules Creek study area. (a) and (b) Catchment located in New South
Wales, Australia. (c) Surface geology of the catchment (modified from surface
geology map of Australia 1:1000000 scale). Locations of the stations for collection
of precipitation (black crosses for Mt Lindsay (Mt L) and Turrawan (T)) and river
flow gauging (red triangles for Boggabri (B) and Turrawan (T)). Also shown are the
locations of the groundwater observation bores (black dots) and bore number for
bores close (< 350 m) to the Namoi River.
The geology of the catchment is complex. The eastern edge of the catchment
is bounded by the Hunter-Mooki Thrust, which has controlled the development
of the catchment. On the eastern side of the thrust is the Nandewar Range.
11
Rocks on this side are part of the New England Origin and are Carboniferous in
age, predominantly consisting of consolidated conglomerates and sandstones, with
multiple beds of rhyolitic to andesitic tuff. On the western side of the thrust are
the westerly dipping Triassic, Jurassic and Permian formations, which are part of
the Gunnedah Basin, a foreland basin. The rocks are predominantly consolidated
claystones, sandstones and conglomerates; however, Permian rhyolitic to dacitic lava
and ash flow tuffs form the southern boundary, and there are multiple coal beds in
the Permian formations. The Permian units now form the majority of the basement
for the unconsolidated aquifer. Immediately to the north is a Tertiary volcanic
centre, consisting of trachytes sills, dykes and plugs, and volcanic flows. These
volcanic rocks cap the ridges that form the north and north-east boundaries of the
catchment [Abbott, 1969; Martin, 1994; DMR, 1998; Young et al., 2002; Roberts,
2006].
In the late Cretaceous, streams incised a valley through the Gunnedah Basin
formations. Since the mid-Miocene, this palaeovalley has been slowly filled with
sediments. It is from these unconsolidated sediments that groundwater is abstracted.
The Cenozoic sediments that fill the palaeovalley on the eastern margin are colluvial
sediments from the weathered and reworked formations that form the Nandewar
Range. The western side of the catchment contains predominantly alluvial sediments
transported by the Namoi River. The upper 6 metres in the flood plains are
dominated by Holocene clay and silt rich vertosols. Beneath and to the east of
the present day Namoi River, relatively coarse grained sediments are deposited in
a palaeochannel, which is up to 120 m deep [Gates, 1980; Gates and Ross, 1980;
Williams, 1997].
Groundwater historically flowed from the Mountain Front in the east to the
Namoi River in the west (Figure 2.2). Since the 1980s, groundwater has been
abstracted for flood irrigation farming of cotton, sorghum and wheat on the alluvial
flood plains, particularly from depth within the palaeochannel. Minor abstractions
12
for stock and domestic use also occur. Some surface water from the regulated Namoi
River is also diverted for irrigation purposes [Andersen and Acworth, 2009].
Figure 2.2: Historical groundwater flow direction. Grey lines show piezometric
head distribution within the shallow aquifer (< 30 m) for 1978 (a year without a
major flood event and which is illustrative of flow directions prior to groundwater
development). Heads are expressed in meters above the Australian Height Datum
(m AHD). Also shown are the locations of the groundwater observation bores (black
dots) and appoximate location of the palaeochannel (grey shaded area).
2.4 Methodology
Throughout the catchment there is a network of government monitoring infra-

structure from which data has been collected. Four types of time series data are
available for the catchment: precipitation and evapotranspiration (daily), river flow
(daily), groundwater levels (monthly to quarterly) and groundwater abstraction
13
(yearly; the water year runs from October to September). The data are publicly
available and were sourced from the Bureau of Meteorology [BOM , 2009] and
Department of Natural Resources [DNR, 2009]. All data were divided into two
periods, before and after the onset of major groundwater abstraction for irrigation.
These periods have been termed “pre-development” and “post-development”. The
changeover point between the periods (October 1984) was identified visually by
observing the first seasonal drawdown because of abstraction in the catchment
groundwater hydrographs.
The methodology consists of two stages: first, an analysis of the hydrological
stresses (i.e., precipitation, evapotranspiration and groundwater abstraction) and
hydrological impacts (i.e., changes in groundwater levels and river flows) by means
of basic statistics and hydrograph analysis, and second, linking the observed changes
in the groundwater levels and river flows via a simple mathematical model.
2.4.1 Hydrological stresses
Precipitation and evaporation data for the mountain range are represented by the Mt
Lindsay weather station and for the plains by the Turrawan weather station (i.e., Mt
L and T in Figure 2.1). The data were analysed for the 20 year period preceding and
following the onset of groundwater abstraction (i.e., October 1964 to September 1984
and October 1984 to September 2004). Daily data were summed for each month
to create monthly data. Initial estimates of any trends were determined using a
low pass filter [5-year moving average; Hoppe and Kiely, 1999]. Standard statistics
(minimum, maximum, average, and standard deviation) were computed for each
period for each station. The significance of differences between the pre-development
and post-development data sets was assessed by calculating the t-statistic value (t).
Monthly groundwater abstraction is reported for each licensed well; however,
only yearly aggregates were available for this study. The totals at each abstraction
well were plotted spatially to show major zones of abstraction.
14
2.4.2 Hydrological impacts
Analysis of groundwater levels
The hydrographs for each observation bore location were plotted on a map of the
catchment so that spatial variations in hydrograph shape could be analysed visually.
Where there was more than one piezometer screened at different depths at a given
observation bore location, the hydrographs for the upper (< 30 m) and lower (> 30
m) portions of the aquifer have been plotted. The observation bores are separate
from the abstraction wells and were installed for the purpose of monitoring.
In contrast to the climate data, groundwater hydrograph data do not exist from
October 1964. The groundwater levels were analysed for the 10 year period preceding
and 20 year period following the onset of groundwater abstraction (i.e., October 1974
to September 1984 and October 1984 to September 2004). Comparable data sets
were created from the raw data by fitting a spline to the raw data and re-sampling
all bores on a monthly interval.
The effect of the abstraction on groundwater levels was investigated in three
stages. (i) Computing the difference in average recovered heads (i.e., the head value
in October prior to the start of the irrigation season) for the pre-development and
post-development periods at each observation bore (∆h). (ii) Calculating for each
observation bore the difference between the average pre-development head and the
recovered head in the post-development period for each year and then averaging
all observation bores to get a catchment average. Then the correlation coefficient
between these data and the total catchment groundwater abstraction for each year
was calculated. (iii) Calculating for each observation bore the average annual
drawdown (i.e., the difference between the recovered head and the minimum head
in each year; used as a proxy for the magnitude of groundwater abstraction) in the
post-development period, and computing the correlation coefficient between these
data and ∆h.
15
Analysis of river flows
As for the climate data, the river flow data at the Boggabri and Turrawan gauging
stations (i.e., B and T in Figure 2.1) were analysed for the 20 year period preceding
and following the onset of groundwater abstraction (i.e., October 1964 to September
1984 and October 1984 to September 2004). The Turrawan gauging stations was not
in operation between 1987 and 1996; the corresponding data at the Boggabri gauging
station was not used in the analysis. Standard statistics (minimum, maximum,
average, and standard deviation) were computed for each period for each gauge.
The significance of differences between the pre-development and post-development
data sets was assessed by calculating the t-statistic value (t).
The river flow data were used to create “Flow Duration Curves” [Vogel and
Fennessey, 1994]. The data also were used to make a comparison between the water
entering the catchment at Boggabri and that leaving the catchment at Turrawan.
This was done by computing the cumulative flow for the period October 1964 to
September 2004 at each gauge. The difference between these cumulative curves was
then calculated, showing periods when more water entered than left the catchment
via the river (positive slope), and vice versa, which was taken to represent the net
river-aquifer flux [Opsahl et al., 2007].
To analyse the low-flow component of the flow regime, three band-pass filters
were applied to the river flow data. Flows less than 0.5, 1.0 and 1.5 GL/day
were considered (n.b., GL is Gigalitres; e.g., 1 GL/day = 1 x 106 m3 /day). The
cumulative river flow difference between the two gauging stations (i.e., T minus B)
was calculated and plotted for these filtered data.
The effect of abstraction on river flow was investigated by dividing the data
into the pre-development and post-development periods. Flow duration curves
were computed using all data and for those months when the river flow difference
between gauging stations was positive and negative for each period. The within-year
distributions of river flows and river flow differences were averaged for each month
16
for each period.
2.4.3 Linking groundwater gradients and river low-flows
The flux across the interface between a hydraulically connected river and aquifer can
be considered to be proportional to the near-river hydraulic gradient in the aquifer
[Anderson and Woessner, 1992], which mathematically is expressed as:
Q=c·i·l (2.1)
where Q is river-aquifer flux (L3 /T), c is a constant (L2 /T), i is near-river hydraulic
gradient in the aquifer (-) (positive indicating gaining river, and negative indicating
losing river), and l is river reach length (L).
Between the Boggabri and Turrawan gauging stations, there are six observation
bores adjacent to the Namoi River on both western and eastern sides (i.e., 36008,
36057, 36016, 36096, 36004 and 30231 in Figure 2.1). Their proximity to the river
(< 350 m) and shallow depths (i.e., 17.7, 39.0, 22.3, 20.7, 10.4 and 13.1 meters below
ground level for 36008, 36057, 36016, 36096, 36004 and 30231, respectively) makes
it possible to compare the river stage to the groundwater level at these locations.
A digital GPS survey of the gauging stations was conducted on 24 May 2009. The
surveyed elevations were interpolated using a one-dimensional linear interpolation
along a polyline representing the river, thus establishing the elevation of the “0 m”
water reference level in the river adjacent to each observation bore. For each month,
the average river stage at Turrawan (T) was calculated. This time series was then
added to the reference level adjacent to the observation bores to obtain time series
of river heads at each site.
The river-aquifer gradient was calculated monthly at each of the six locations
by subtracting the aquifer head from the computed river head and dividing by the
shortest distance from the observation bore to the river. A weighted average river-
aquifer gradient for the catchment was constructed by weighting each of the six
17
calculated gradients by the length of river reach which they represent. This length
was calculated as the distance to the mid-points between the bore of interest and the
nearest upriver and downriver bores. The constant c, which represents the transfer
term for the river and aquifer, was then adjusted until the cumulative river-aquifer
flux computed using Equation 2.1 matched the river-aquifer flux computed from the
river flow data. For the purpose of illustrating the link between the groundwater
gradients and river flows, the 1.0 GL/day cut-off was adopted. The correlation
coefficient between the two flux time series was calculated.
2.5 Results
2.5.1 Hydrological stresses
The precipitation and evapotranspiration time series, raw and filtered, for the moun-
tain range (Mt Lindsay) and plain area (Turrawan) are shown in Figure 2.3. Visually
there are no overall trends in the series. The statistics for the pre-development and
post-development periods are shown Tables 2.1 and 2.2. Maximum and average
precipitation is approximately 60% higher on the mountain range than that on
the plain for both periods. Maximum evapotranspiration is approximately 15%
higher on the plain than that on the mountain range for both periods, whereas
the average is approximately 25% higher on the plain. There are no statistically
significant differences between the pre-development and post-development periods
in the precipitation (t = 0.85 and 0.57 for Mt L and T) or evapotranspiration (t =
0.79 and 0.86 for Mt L and T) data.
18
Figure 2.3: Monthly precipitation and evapotranspiration time series (grey) with
5-year moving average (black) for Mt Lindsay (Mt L in (a) and (c)) and Turrawan
(T in (b) and (d)).
Pre-development Post-development
Mt L T Mt L T
Minimum 2 0 0 0
Maximum 432 304 354 218
Average 81 52 83 50
Standard Deviation 73 52 66 45
Table 2.1: Precipitation statistics for pre-development (before October 1984) and
post-development (from October 1984) periods for Mt Lindsay (Mt L) and Turrawan
(T) weather stations (mm/month).
19
Mt L T Mt L T
Minimum 30 41 32 45
Maximum 200 235 203 232
Average 100 125 101 126
Standard Deviation 45 51 46 51
Table 2.2: Evapotranspiration statistics for pre-development (before October 1984)

and post-development (from October 1984) periods for Mt Lindsay (Mt L) and
Turrawan (T) weather stations (mm/month).
The spatial distribution of the total abstraction since the onset of development
is plotted in Figure 2.4. The majority of the abstraction (at some locations up to 15
GL) is confined to a narrow north-south running zone, which roughly corresponds
to the higher transmissivity palaeochannel.
20
Figure 2.4: Spatial distribution of cumulative groundwater abstraction since the

onset of development across the catchment (grey circles). The black dots show
the location and name of the government observation bores. Grey lines show the
average head differences between the pre-development and post-development periods
for upper portion of aquifer (> 30 m). Also shown is the approximate location of
the palaeochannel (grey shaded area).
2.5.2 Hydrological impacts
Declines in groundwater levels
The hydrographs in the catchment (Figure 2.5) show that the heads are higher in
the east than near the river, implying a general westward groundwater flow towards
the river. There is a general decline in head with time, especially since 2000. The
declines have been greater immediately to the east of the river compared to the area
further east, generally correlated to the areas with greatest abstraction (see also
21
contours in Figure 2.4).

Along an east-west transect in the northern part of the catchment (i.e., bores
30231, 30232, 30233, 30234, 30235, 30236, 30237, 30132, 30133 and 30134; refer
to Figure 2.5), episodic (intra-annual) but rapid head increases (up to 3.5 m) are
evident in all observation bores, particularly in the east (e.g., bore 30237). In
contrast, there are pronounced temporary yearly drawdowns (up to 7 m) further
west (e.g., bore 30233). The heads in the upper portion of the aquifer (blue lines)
are higher than those in the lower portion (green lines) (e.g., bore 30235).
In the central zone (i.e., bores 36004, 30447, 36094 and 30446; refer to Figure
2.5), the episodic and rapid head increases evident in the northern zone are apparent
in all observation bores to some degree. Pronounced yearly drawdowns (up to 12
m) are observable in all bores, except bore 36004 and the upper piezometer at bore
30446. The heads in the upper portion of the aquifer are higher than those in
the lower portion, except at bores near the Namoi River (e.g., bore 36004) where
from 1970 to 2000 there was an upward gradient, and following 2000 there was a
downwards gradient.
Along a transect parallel to Maules Creek (i.e., bores 36005, 36096, 36093, 36003,
36164, 36186, 36187, 30129, 30130 and 30131; refer to Figure 2.5), there are pro-
nounced yearly drawdowns (up to 11 m) for bores located within the palaeochannel.
The heads in the upper portion of the aquifer are higher than those in the lower
portion (particularly at bore 30130), except at bore 30132 and at 36005. The
patterns of head changes over time are similar between bore 36186 and 36187 as
well as bores 30129 and 30130.
22
Figure 2.5: Location of selected representative observation bore hydrographs within

the study area (21 piezometers at depths less than 30 m; 20 piezometers at depths
greater than 30 m). All hydrograph inserts have a vertical scale of 20 m (elevation
varies depending on location) and a horizontal scale of 40 years (1970 to 2010) with
tick marks every 5 years. Heads are expressed in meters above the Australian Height
Datum (m AHD). Hydrographs for the upper portion of the aquifer (< 30 m) are
shown in blue and the lower portion (> 30 m) in green. The depth of screen are in
meters below ground level (m bgl) and shown below each hydrograph. Also shown
is the approximate location of the palaeochannel (grey shaded area).
The correlations between groundwater heads and abstraction are shown in Figure
2.6. The abstraction for the catchment varies between 5 and 18 GL/year over the
period 1985 to 2004. Generally the abstraction has been increasing from approx-
imately 5 GL/year in the mid-1980s to 16 GL/year in 2004, with a maximum of
just less than 18 GL/year in 2002. The post-development head, on a catchment-
average basis, varies from 0.15 m above (in 1990 and 1998) to 3.6 m below (in 1994)
23
the pre-development head value. The correlation coefficient between this and the
groundwater abstraction is high (r = 0.84) (Figure 2.6a). The correlation coefficient
between the difference in average pre-development and post-development heads and
the average yearly drawdown is also high (r = 0.92) (Figure 2.6b).
Figure 2.6: (a) The total abstraction in the catchment for each water year
(i.e., October to September) (grey bars) and average pre-development head minus
recovered post-development head each year, averaged for catchment (black line).
(b) Relationship between average pre-development head minus average post-
development head and the average yearly drawdown at each observation bore. For
(a) and (b), piezometers with depths greater than 30 m only were used.
Losses in river flows
The river flow statistics for Boggabri and Turrawan for the pre-development and
post-development periods are shown in Table 2.3. The average monthly flow is an
order of magnitude less than the maximum monthly flow. There are no statistically
significant differences in the pre-development and post-development averages for
either Boggabri (t = 0.92) or Turrawan (t = 0.92).
B T B T
Minimum 0.3 0.3 0.0 0.1
Maximum 967.5 990.8 1039.3 990.2
Average 64.0 60.2 59.7 57.4
Standard Deviation 140.2 126.1 147.3 146.9
Table 2.3: River flow statistics for pre-development (before October 1984) and
post-development (from October 1984) periods for Boggabri (B) and Turrawan (T)
gauging stations (GL/month).
24
The river flow regime at Boggabri and Turrawan is shown in Figure 2.7. This
reach of the Namoi River is characterised by episodic large flow events (Figure 2.7a).
Infrequently very large (~ 2.5 GL/day less than 10 % of the time) and very small
(~ 0.1 GL /day less than 10 % of the time) flow events occur (Figure 2.7b). The
cumulative plots (Figure 2.7c, primary axis) show overall more flow at Boggabri
than at Turrawan (~ 1000GL for the period October 1974 to September 2004).
The cumulative difference plot (Figure 2.7c, secondary axis) shows that this overall
negative trend is interrupted with shorter periods of positive trend. The cumulative
difference plots for the low-flow periods (Figure 2.7d) reveal a very different picture
with two main periods: initially, an overall positive trend followed by an overall
negative trend. This observation can be seen regardless of which low-flow cut-off is
considered. However, the timing of the changeover point varies depending on which
low-flow cut-off value is used; the timing of the changeover point is later for a smaller
low-flow cut-off value (early 2000s for 0.5 GL/day cut-off compared with 1990s for
1.5 GL/day).
25
Figure 2.7: (a) Daily river flow at Boggabri (B) and Turrawan (T). Note the data
gap between 1987 and 1996 represents a period when the Turrawan gauge was
not in operation. (b) Flow duration curves for Boggabri and Turrawan gauging
stations (computed on a daily basis). (c) Cumulative river flow at Boggabri and
Turrawan gauging stations (black and grey) and the difference in cumulative river
flows (dashed). (d) Cumulative river flow differences with flows larger than 0.5, 1.0
and 1.5 GL/day excluded by filtering.
The river flow regimes for the pre-development and post-development periods
(i.e., before and after October 1984) are shown in Figure 2.8. Pre-development
the flows are generally greater than for the post-development period (Figure 2.8a;
i.e., black curves are above grey curves). Furthermore, during the pre-development
period flows at Turrawan are marginally higher than those at Boggabri except
at high-flow events. In the post-development period the flows at Turrawan are
smaller than those at Boggabri for flows greater than 0.2 GL/day (i.e., 59 % of the
time). Comparing the pre-development and post-development periods, there is also a
change in the magnitude of flow and frequency of months where the flow at Turrawan
is greater than that at Boggabri (Figure 2.8b). The magnitude and frequency of
positive differences are greater for the pre-development period (i.e., black curve
is to the right of grey curve in Figure 2.8b) while the magnitude and frequency
26
of negative differences are greater for the post-development period. Comparing the
pre-development and post-development periods, there is also a pronounced change in
the distribution of river flows throughout the year when analysed on a monthly basis
(Figure 2.8c-d). In the pre-development period there were much larger summer flows
(> 150 compared to ~ 75 GL/month). In the post-development period there are very
low-flows in March-June (< 20 GL/month). Furthermore, in the pre-development
period the river flow difference is positive for all months except possibly December-
January, with maximums in March-May and November (Figure 2.8e). In contrast
to this, for the post-development period the river flow difference is negative with a
minimum in December-January, with the exception of March-July where the river
flow difference is positive (Figure 2.8f).
27
Figure 2.8: (a) River flow duration curves for pre-development and post-development
periods (i.e., October 1964 to September 1984 and October 1984 to September
2004). (b) Flow duration curves for positive (i.e., gaining) and negative (i.e., losing)
river flow differences (i.e., flow at Turrawan (T) minus flow at Boggabri (B)). (c)
Within-year distribution of river flow for pre-development period. (d) Within-year
distribution of river flow for post-development period. (e) Within-year distribution
of river flow difference for pre-development period for river flows smaller than 0.5,
1.0 and 1.5 GL/day. (f) Within-year distribution of river flow difference for post-
development period for river flows smaller than 0.5, 1.0 and 1.5 GL/day.
2.5.3 Linking groundwater gradients and river low-flows
The river-aquifer gradients for six locations (i.e., bores 36008, 36057, 36016, 36096,
36004 and 30231; refer Figure 2.1) are shown in Figure 2.9. At observation bores
36008, 36057 and 36016, located in the upriver portion of the catchment, the
gradient initially varied from positive (i.e., gaining river) to negative (i.e., losing
28
river) values (ranging from 0.010 to -0.005), but from the mid-1990s onwards, the
gradient was generally negative (up to -0.015) (Figure 2.9a-c). At observation
bores 36096 and 36004, located in the centre of the catchment, the gradient was
initially consistently positive (approximately 0.015) but, from the mid-1990s, became
increasingly negative, particularly at observation bore 36005 (where the gradient
became -0.015 by 2007) (Figure 2.9d-e). Finally at observation bore 30231, located
at the lowest end of the catchment, the gradient was consistently positive for most
of the studied period except from the last couple of years where the gradient was
zero or slightly negative (Figure 2.9f).
Figure 2.9: River-aquifer gradient between river and observation bores (a) 36008 (b)
36057 (c) 36016 (d) 36096 (e) 36004 (f) 30231 (for locations see Figure 2.1). The
gradients are presented as computed and smoothened, where smoothened results
were obtained using 12-month moving average filter. A positive value indicates
potentially gaining conditions and a negative value losing conditions.
29
Initially the weighted average river-aquifer gradient was positive (i.e., gaining)
and varied between 0.005 and 0.015 (Figure 2.10a). However since September
1990 there has been an overall negative (i.e., losing) trend in the gradient. The
corresponding cumulative river-aquifer flux shows a positive period follow by a
negative, with the changeover occurring in November 2001. Using a value of c
= 69 m2 /d, the match between the computed river-aquifer flux and the river flow
difference is high (r = 0.99) (Figure 2.10b).
Figure 2.10: (a) Weighted average river-aquifer gradient (based on 6 river segments)
for the near-river (< 350 m from river) zone for the entire reach between Boggabri (B)
and Turrawan (T) gauging stations. A positive value indicates potentially gaining
conditions and a negative value losing conditions. (b) Cumulative river-aquifer fluxes
computed using river flow data (Figure 2.7d, grey line) and groundwater gradient
data (Q = c·i·l wherec is 69 m2 /day, i is near-river hydraulic gradient in the aquifer
and l is river reach length). The gap in the curves is because the cumulative flux
based on the river flow difference could not be calculated due to the lack of data
from the Turrawan gauging station which was not in operation between 1987 and
1996.
2.6 Discussion
2.6.1 Groundwater abstraction and changes in groundwater
levels
The Nandewar Range on the eastern side of the catchment causes a difference in
precipitation and evapotranspiration between the plain and mountain range with
precipitation being consistently lower and evapotranspiration being consistently
30
higher, on the plain compared to the mountain range (Tables 2.1 and 2.2). There are
small trends within the data but the differences between the pre-development and
post-development periods are not significant (Figure 2.3). Recharge is not expected
to vary much between the periods.
In addition to climate, the other major stress on the system is groundwater
abstraction. The majority of abstraction has occurred within the palaeochannel
with minor abstraction along Maules Creek (Figure 2.4, grey circles). Associated
with the groundwater abstraction is an overall decline in hydraulic head (Figure 2.4,
grey lines).
The variability in the groundwater levels is controlled by a number of hydrological
processes (Figure 2.5). The greatest decline in groundwater levels is occurring within
the palaeochannel where most of the groundwater abstraction is occurring. The
groundwater levels in general respond to mountain front/diffuse recharge events (i.e.,
the episodic and rapid head increases) and to the seasonal groundwater abstraction
(i.e., the pronounced yearly drawdowns). At some locations there is little response
to the seasonal abstraction, such as those where the river dampens the effect and in
some upper zones, which appear to have poor connectivity to the lower portion of
the aquifer. In general, there is a downward gradient from upper to lower portions
of the aquifer. Exceptions to this are the discharge zones adjacent to the Namoi
River and in an area where previous research has observed groundwater discharge
[Andersen and Acworth, 2009; Rau et al., 2010]. Although in some locations it
appears that Maules Creek is disconnected from the aquifer, there are additional
head variations because of interactions with the creek in other areas. Observation
bores located to the east of the creek are similar to each other and show no response
to the groundwater abstraction.
While there are complex processes occurring which vary spatially and temporally,
it can be concluded that overall the hydraulic heads show a long term decline.
A number of factors indicate that the groundwater abstraction rather than the
31
impact of climatic variability on recharge is the major cause. First, the abstraction
has gradually increased over time (Figure 2.6a, grey bars) with the variability
approximately reflected by the catchment average groundwater level (Figure 2.6a,
black line). The correlation coefficient between these two variables is high. Second,
there is a close relationship between those areas where there has been a large decline
between the pre-development and post-development periods and where there is a
large seasonal drawdown (used as a proxy for large groundwater abstraction) (Figure
2.6b).
2.6.2 Changes in river flows
The flow regime in the Namoi River in the catchment is due to a combination
of climatic and anthropogenic stresses. The Namoi River is regulated at Keepit
Dam (completed 1961) located 130 km upriver of the catchment (measured along
the river). Dam water is released primarily to satisfy irrigation water demands as
well as releases for environmental requirements. Downriver of the dam there are
three major tributaries to the river (Peel River, Mooki River and Cox’s Creek)
as well as numerous minor tributaries. Consequently, the compounded river flow
entering the Maules Creek sub-catchment at Boggabri is highly variable (Figure
2.7a-b) and consists of regulated and natural flow components. Comparison of
the raw flow data at the gauging stations at Boggabri and Turrawan indicate that
overall there is a net loss of water from the river reach (Figure 2.7c). However
this overall picture is completely dominated by the large flow events. In addition,
large uncertainties related to the high-flows made the comparison noisy. Regardless,
the large flow events tend to induce steep but often short-lived gradients from the
river to the aquifer because of the high river stage. These events naturally cause
losing conditions. From inspecting the total river flows, the apparent conclusion
is that only minor changes have occurred from the pre-developmentto the post-
development period (refer to overall river flow statistics in Table 2.3). However,
32
when the low-flow conditions are considered, by filtering out the high-flows from
the data, a more nuanced picture emerges. Generally, there was a net gain in river
flow until some point in the 1990s (or early 2000s) after which the reach became a
net losing system (Figure 2.7d). This is an important observation since the low-flow
periods (if defined as < 1.0 GL/day) constitute approximately 65% of the river flows
(Figure 2.7b). The timing of this change depends on the definition of “low-flow”,
which is an ongoing issue in the literature [e.g. Smakhtin, 2001]. However, regardless
of which definition is accepted, there was a significant delay (i.e., > 10 years) between
the onset of groundwater abstraction and the change from overall gaining to overall
losing river conditions at low-flows. The analysis above highlights the necessity of
not only analysing the total river flows but the low-flows in particular [see Mair
and Fares, 2010].
The methodology used of computing river-aquifer flux by means of using differ-
ential flow measurements has inherent limitations [Halford and Mayer, 2000; Harte
and Kiah, 2009; Opsahl et al., 2007]. The losing and gaining trends were obtained by
computing the flow differences between two gauging stations. However, in principle
these differences could be caused by other factors than river-aquifer exchanges, such
as surface water evaporation, river diversions for irrigation, and precipitation-runoff
contributions (from tributaries, overland flow and interflow). Evaporation from the
narrow tree-lined river can be assumed to be negligible compared with the over-
all river water balance. Prior to the groundwater development, there was more
reliance on surface water in the catchment. Therefore, if the diversion data (which
were not available) were to be included in the analysis, the computed gains for
the pre-development period would be larger and the change from gaining to losing
would be even more apparent. Some evidence suggests that runoff is an insignificant
contribution to the river flow differences. First, the river is surrounded by a very
flat plain with, based on field observations, little evidence of runoff. Second, the
flows that are generated by large precipitation events are mainly confined to Maules
33
Creek, which rarely flows at the confluence with the Namoi River, and when it does
the contribution is small when the Namoi itself is in flood [Andersen and Acworth,
2009]. In any case, runoff caused by these large storm events correlates to periods
where the river flow is high which are the periods filtered out of the data for the
analysis of river-aquifer interactions. Although there is an inherent uncertainty in
the analysis because of lack of data, the basic conclusion of a gaining period followed
by a losing period remains robust.
Due to a combination of the climatic and anthropogenic stresses as discussed
above, the flow regimes in the pre-development and post-development periods are
different (Figure 2.8a, c-d). While it is not possible to tell what exactly is driving
these changes, it is possible to see the influence of groundwater abstraction in the
river flow data. The flow duration curve has been utilised in a different way to
normal practice by considering the magnitude and frequency of those months where
more water enters the catchment at Boggabri than leaves at Turrawan, and vice
versa. The resultant curves show that the magnitude and frequency of losing months
has increased over time (Figure 2.8b). The use of band-pass filtering on the river
flow differences has also shown the changes occurring in the catchment. For low-flow
conditions, the river was gaining throughout the year for the pre-development period
except possibly December-January (Figure 2.8e). However, any gains were low (i.e.,
river flow differences close to zero) around September and during December-January,
which corresponds to the timing of irrigation and possible surface water diversions.
In the post-development period the river was losing, during spring and summer
(August to February), with only a short period of gaining conditions during winter
from March to July (Figure 2.8f). By considering the data in this pre-development
and post-development period manner and by looking at changes that have occurred
in the catchment (i.e., between the Boggabri and Turrawan gauging stations), it is
clear that the river flow regime is being influenced by the groundwater abstraction
[see Mair and Fares, 2010; Tao et al., 2011].
34
2.6.3 Relationship of groundwater declines to river flow
losses
The trends in the river-aquifer gradients (Figure 2.9 and Figure 2.10a) show that,
over time, there has been a reversal in the river-aquifer gradient and, therefore, a
change from overall potentially gaining to losing conditions. The river-aquifer flux
computed using the river stage and aquifer head data gives a very good match to
that calculated using the river flow data (Figure 2.10b). The excellent match was
in part produced by adjusting c in Equation 2.1. It should be noted though that
while adjusting c changes the magnitude of the computed flux, there are features in
the curves that match well in addition to simply matching maximum cumulative
flux values: the magnitude of cumulative river-aquifer flux at the start of the
data gap (i.e., 1987), the timing of the changeover from overall gaining to losing
(i.e., 2001), and the magnitude of the cumulative river-aquifer flux at the end of
the computation (i.e., 2004). The value of c also has a physical meaning: the
transmissivity of the aquifer (i.e., the product of aquifer hydraulic conductivity and
aquifer thickness). If the hydraulic conductivity varies from 2 to 10 m/d [Domenico
and Schwartz, 1990] and the thickness varies from 10 to 120 m, then the true value of
c is within 20 to 1200 m2 /d. The best fit value of c = 69 m2 /d is clearly within this
range. This simple agreement of river and aquifer data demonstrates the connection
between groundwater declines because of abstraction and river flow depletion at the
catchment scale.
From this analysis, it can be concluded that any further lowering of the ground-
water levels by groundwater abstraction will, by implication of the relationship in
Equation 2.1, lead to an increasing loss from the reach. Ultimately, either the reach
will become disconnected from the groundwater table, in which case the loss rate
will become constant and independent of the head change [Brunner et al., 2009a] or
the reach simply runs dry at low-flow due to the loss. However, it shoudl be noted
that while the relationship established above using Equation 2.1 illustrates the link
35
between groundwater gradients and river flows, it is not a predictive tool. For
example, changes in river management, and therefore in river levels, would require a
new relationship to be established. Furthermore, the relationship is only applicable
when the river and aquifer are hydraulically connected.
Sophocleous [2002] argues that the primary challenge in understanding river-
aquifer interactions to date has been in spatially defining the hydraulic properties
of the riverbed, making quantification of fluxes a major challenge. However, in the
approach used here, the transfer term between the river and aquifer is treated as a
single term and the temporal [Zammouri and Feki, 2005] and spatial [Fleckenstein
et al., 2006; Kalbus et al., 2009] variability in the riverbed conductivity has been
ignored. Furthermore, the approach simplifies the river-aquifer interactions which
have been reported at a small-scale because of factors such as river morphology and
curvature [Woessner, 2000; Cardenas, 2009]. The results indicate that the “lumped
parameter” approach used has value in understanding river-aquifer interactions at
the catchment-scale.
2.6.4 Summary of processes
The processes involved in river-aquifer interactions in a semi-arid environment

stressed with groundwater abstraction are conceptually summarised in Figure 2.11
(numbers in the figure correspond to numbers in parentheses in the following). There
has been an overall decline in groundwater levels between the pre-development (1)
and post-development (2) periods. This decline can be predominantly attributed
to groundwater abstraction (3) rather than climatic changes (4). For both periods,
high-flow conditions (i.e., dam releases and flooding events) (5) caused temporary
losing conditions (6). However, during low-flow conditions (7) there has been a
change in the river-aquifer interaction corresponding to the decline in groundwater
levels. In the pre-development period the river was overall gaining (8), whereas in
the post-development period, the river is now overall losing (9). However, during
36
very low-flow conditions (11), the river is still gaining in the post-development
period. The decline in groundwater levels between the periods may appear modest
(~ 3 m), but is sufficient to lower the near-river groundwater level below the low-flow
river level thereby causing losing conditions. There has been a significant lag (> 10
years) from the onset of groundwater abstraction to the changeover from gaining to
losing conditions for the catchment.
Figure 2.11: Cross-section (not to scale) of river, aquifer and abstraction well
with conceptualized river-aquifer interactions (see explanation in text). Grey lines
represent water levels.
2.6.5 Significance for resource management
Given that groundwater abstraction, where the river and aquifer are hydraulically
connected, will inevitably lead to reduction in river flows through captured discharge
or induced river recharge as illustrated in this analysis, this consideration may be the
limiting factor when considering sustainable water use in any given catchment rather
than the magnitude of recharge from mountain front or diffuse sources [Braaten and
Gates, 2003].
The change from gaining to losing river conditions occurred with only a modest
decline in groundwater levels (~ 3 m) and with a significant delay (> 10 years)
after the onset of groundwater abstraction. For this situation to be reversed,
37
the groundwater levels in the near-river part of the aquifer would have to rise
to elevations at least above the base of the river channel. However, as pointed
out by Ivkovic et al. [2009], it can take decades or longer to recharge aquifers
when groundwater resources have been overexploited. Furthermore, rarely will it be
realistic for groundwater abstraction to cease altogether, although reductions may
be possible [Rushton, 2002]. Therefore, the use of the groundwater water stored
above the base of the river channel may prove to have been a one-time event and
led to a permanent change from gaining to losing conditions.
The reduced baseflow may mean a deteriorated aquatic environment for in-river
fauna and benthic species [Hancock, 2002; Humphreys, 2009]. In addition, the
induced river recharge may have implications for the groundwater quality in the
vicinity of the river as has been document at other sites [Doussan et al., 1997;
Massmann et al., 2004; Postma et al., 2007]. Both of these aspects should be
researched further to inform management policy.
These types of management issues will only become more apparent as the demand
on groundwater resources continues to increase. The study highlights the crucial
importance of constructing and maintaining a strategic monitoring system with
observation bores placed both near rivers and well fields.
2.7 Conclusions
Overall, there has been a decline in groundwater levels between the pre-development
and post-development periods (up to ~ 3 m). This is particularly the case within
the palaeochannel where seasonal drawdowns due to groundwater abstraction are
most evident (up to ~ 12 m). With no visible trends or statistical differences in
precipitation or evapotranspiration between the periods, the decline can be pre-
dominantly attributed to the groundwater abstraction rather than any changes in
climate. This is confirmed by the high correlation between groundwater abstraction
and groundwater levels on a catchment scale (r = 0.84) as well as between the
38
head decline between the periods and the magnitude of seasonal drawdown at the
observation bore scale (r = 0.92).
By computing the difference between the cumulative volumes of water entering
compared to leaving the catchment for low-flows (i.e., flows less than a cut-off value
of 0.5, 1.0 or 1.5 GL/day), it is seen that there is a period of gaining conditions
followed by one of losing conditions. If defined as < 1.0 GL/day low-flow periods
constitute approximately 65% of the river flows. The periods where the river is
losing at low-flow conditions are thus significant. In the pre-development period,
the river was gaining all year (with the possible exception of December-January),
whereas in the post-development period the river is losing year round (especially in
December-January which also corresponds to the groundwater abstraction season)
with the exception of March-June, which corresponds to the period of low-flow in
the river.
Importantly, it has been illustrated quantitatively that the losses in the river
flows are linked to the declines in groundwater levels. The river-aquifer gradients
computed for different reaches show the same trend seen in the river flow data: a
gaining period followed by a losing period. Using a simple mathematical model
(Equation 2.1) the flux computed using the river-aquifer gradients matched well
with the river-aquifer flux computed using the river flow data (r = 0.99).
Two important insights resulted from the analysis. With only modest declines
in groundwater levels (up to ~ 3 m) because of groundwater abstraction a situation
has arisen where the groundwater levels are below the river low-flow level thereby
resulting in losing conditions. Second, there was a significant delay (> 10 years)
between the onset of groundwater abstraction and the changeover from gaining to
losing conditions. These results have implications for management of the ground-
water resource. For example, reductions in groundwater abstraction will now not
lead to increased baseflow until the groundwater levels rise above the river low-flow
level, which may take decades.
39
The analysis approach used, which considered relationships within and between
precipitation, evapotranspiration, river flow, groundwater levels and abstraction
data sets, can be transferred to other catchments worldwide where there has been
a similar level of baseline monitoring.
40
Chapter 3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Estimating interactions using river-

bed temperature time series
3.1 Summary
In order to manage surface water (SW) and groundwater (GW) as a single resource,
it is necessary that the interactions between them are understood and quantified.
Heat, as a natural tracer of water movement, is increasingly being used for this
purpose. However, analytical methods that are commonly used are limited by
uncertainties in the effective thermal diffusivity of the sediments at the SW-GW
interface. We present a novel 1D analytical method. It utilises both the amplitude
ratio and phase shift of pairs of temperature time series at the SW-GW interface
to estimate the Darcy velocity. This eliminates both the need to specify a value
for effective thermal diffusivity and the need for iteration. The method also allows
for an estimation of effective thermal diffusivity, which can indicate periods where
assumptions to the analytical solution are violated. Riverbed temperature data from
the Murray Darling Basin (Australia) are used to illustrate the method.
41
Chapter 3. Estimating interactions using riverbed temperature time series
3.2 Introduction
In water management globally, there is a move towards a conjunctive approach which

recognizes that surface water (SW) and groundwater (GW) are a connected resource
[Winter et al., 1998; Woessner, 2000]. For this to be successful, the interactions
between SW and GW need to be understood and quantified.
Heat as a natural tracer of water movement has increasingly become popular for
the study of SW-GW interactions [Anderson, 2005; Constantz, 2008]. The principle
behind this approach is that daily temperature fluctuations within a SW body
(e.g., river) due to solar radiation lead to a temperature response in the sediments
at the SW-GW interface (e.g., riverbed) due to conduction and convection. In
particular, the sinusoidal components of the diel temperature fluctuations which
propagate through the SW-GW interface have been exploited to quantify the Darcy
velocity [Stallman, 1965; Hatch et al., 2006; Keery et al., 2007]. However, analytical
methods that are commonly used are limited by uncertainties in the effective thermal
diffusivity value [Shanafield et al., 2011].
In this chapter, a novel method is presented which by-passes this issue by
estimating the Darcy velocity from pairs of temperature time series without the need
to specify the effective thermal diffusivity value. Estimations of the effective thermal
diffusivity are also obtained from the time series. The presented method has the
same basis, and inherent limitations, as the original Stallman [1965] solution. The
method is illustrated using data from Maules Creek Catchment, Murray Darling
Basin (Australia). The method would be equally applicable to other SW-GW
settings such as streams, lakes, springs and channels.
42
3.3 Methodology
3.3.1 Theoretical background
The 1D convection-conduction equation is
∂T ∂ 2T ∂T
=D 2 −v (3.1)
∂t ∂z ∂z
where T is temperature, t is time, z is depth, v is thermal front velocity and D is

effective thermal diffusivity [Suzuki, 1960].
In this thesis the convention used is that v is positive for gaining conditions (i.e.,
when the SW body gains water), and negative for losing conditions (i.e., when the
SW body loses water).
For saturated conditions, the thermal front velocity and effective thermal diffus-
ivity are given by
ρ w cw
v= q (3.2)
ρc
and
!2
κ ρ w cw
D= +β q (3.3)
ρc ρc
where q is Darcy velocity, ρw and cw are the density and heat capacity of water, ρc
is the bulk heat capacity, κ is bulk thermal conductivity, and β is thermal dispersivity
(in the direction of fluid flow) [Rau et al., 2012].
The first term in Equation 3.3 is the thermal diffusivity and the second term the
increase in thermal diffusivity due to thermal dispersivity; together they constitute
the effective thermal diffusivity.
The bulk heat capacity is defined as the arithmetic average of the phases
43
ρc = ρw cw + (1 − n) ρs cs (3.4)
where n is porosity, and ρs and cs are the density and heat capacity of solids
[Buntebarth and Schopper, 1998].
The bulk thermal conductivity is defined as the geometric average of the phases
κ = κnw κ1−n
s (3.5)
where κw and κs are the water and solid thermal conductivity [Woodside and
Messmer, 1961].
As the solid phase is constituted of various minerals, κs can be approximated
using
κs = κQ 1−Q
q κo (3.6)
where Q is the quartz content of the total solids, κq is the solid thermal con-
ductivity of quartz, and κo is the solid thermal conductivity of the fraction of other
minerals [Johansen, 1975].
Equations 3.3 through 3.6 show that the effective thermal diffusivity is a function
of a number of different parameters, each of which has a possible range, which leads
to the uncertainty in its value.
For a semi-infinite half space where the temperature of the upper surface is
varying sinusoidally, the temperature at any depth and time is given by
 s   s 
vz z α + v2  2πt z α − v2 
T(z,t) = A · exp  − cos  − (3.7)
2D 2D 2 P 2D 2
s
8πD
2
α= v4 + (3.8)
P
where A is amplitude and P is period [Stallman, 1965; Goto et al., 2005].
44
Equation 3.7 is subject to the assumptions that (1) fluid flow is in the vertical
direction only, (2) fluid flow is steady-state, and (3) fluid and solid properties
are constant in both space and time, and (4) fluid and solid temperatures at any
particular point in space are equal at all times [Stallman, 1965].
Starting from Equation 3.7, previous researchers [e.g. Hatch et al., 2006; Keery
et al., 2007] have developed analytical methods to estimate the thermal front ve-
locity. These utilize amplitude ratio or phase shift data from pairs of temperature
time series recorded at different depths.
The equation for use with amplitude ratio data is
s
2D α + v2
−vAr = ln (Ar ) + (3.9)
∆z 2
where Ar is the amplitude ratio (i.e., Ar = A2 /A1 ; see Figure 3.1).

Similarly, the equation for use with phase shift data is
s
4πD
2
|v∆Φ | = α − 2 ∆Φ (3.10)
P ∆z
where ∆Φ is the phase shift of temperature maxima, minima, or zero crossings

(i.e., ∆Φ = t2 − t1 ; see Figure 3.1).
When using the convention v is positive for gaining and negative for losing
conditions, a negative sign is required on the left hand side of Equation 3.9. The
absolute sign on the left hand side of Equation 3.10 indicates that this equation
provides the magnitude, but not the direction of the thermal front velocity.
In this approach, v is a function of Ar and D (Equation 3.9) or a function of ∆Φ
and D (Equation 3.10). Since D varies, applying Equations 3.9 and 3.10 results in
uncertainty in the estimated v. When β is assumed to be negligible, D varies from
0.02 to 0.13 m2 /d with an average of 0.075 m2 /d [Shanafield et al., 2011].
45
3.3.2 Proposed method
Given the assumptions of Equation 3.7, the vAr and results v∆Φ should be consistent.
Equations 3.9 and 3.10 can therefore be combined and rearranged (done here using
Mathematica, version 7) to give an expression for thermal front velocity as a function
of Ar and ∆Φ:
∆z (P 2 ln2 Ar − 4π 2 ∆Φ2 )
v= √ (3.11)
∆Φ 16π 4 ∆Φ4 + 8P 2 π 2 ∆Φ2 ln2 Ar + P 4 ln4 Ar
Similarly, the effective thermal diffusivity can be expressed as a function of Ar

and ∆Φ:
∆z 2 P 2 lnAr (4π 2 ∆Φ2 − P 2 ln2 Ar )

D= (3.12)
∆Φ (P 2 ln2 Ar + 4π 2 ∆Φ2 ) (P 2 ln2 Ar − 4π 2 ∆Φ2 )
In this approach, v is a function of Ar and ∆Φ but not D; the information about

D is implicit in the Ar and ∆Φ data. Likewise, D is a function of Ar and ∆Φ but
not v. Furthermore, v and D appear on the left hand side of the equations only,
and so no iteration is required.
It is thus possible to quantify the SW-GW interaction (i.e., Darcy velocity, q)
using temperature time series by multiplying the estimated thermal front velocity
(v) with a constant (i.e., Equation 3.2 rearranged, γ):
q = γv (3.13)
nρw cw + (1 − n) ρs cs
γ= (3.14)
ρ w cw
It should be noted that there is some confusion in the literature concerning the
relationship of Darcy velocity to thermal front velocity. This appears to have arisen
via [Goto et al., 2005] and Hatch et al. [2006]. Goto et al. [2005] use the symbol vf for
Darcy velocity in their Equation 5 [S. Goto, personal communication, 2012 ] while
46
Hatch et al. [2006] use vf for Darcy velocity divided by porosity in their Equation
1 (and possibly for Darcy velocity in the remainder of the paper). Gordon et al.
[2012] correctly and clearly set out the terms.
The value of γ is also uncertain as it depends on physical and thermal parameters
that vary from site to site. Consequently, there is a range of possible Darcy velocities
for a given value of thermal front velocity. To assess the uncertainty, γ10 , γ50 and
γ90 , the 10th, 50th and 90th percentiles of γ, were computed using a Monte Carlo
analysis of literature values (see Table 3.1). It was assumed that the porosity and
solid heat capacity follow a normal distribution, with standard deviations equal to
one fourth of the maximum range in values.
Parameter Min. Max. Mean Std Dev. Units

n 0.25 0.50 0.375 0.0625 -
ρw - - 998 - kg m-3
ρs - - 2650 - kg m-3
cw - - 4183 - J kg-1 °C-1
cs 690 1270 980 145 J kg-1 °C-1
Table 3.1: Literature values used in the Monte Carlo analysis. Sources: Schön [1996],
Shanafield et al. [2011], and Thermophysical Properties of Fluid Systems, Chemistry
WebBook, National Institute of Standards and Technology, 2011, available at
http://webbook.nist.gov/chemistry/fluid/.
3.3.3 Field example
To illustrate the method, data from Maules Creek Catchment, Murray Darling
Basin (Australia), were collected and interpreted. A temperature array was designed
consisting of temperature sensors (Onset HOBO Pro v2) at 0.05 and 0.23 m depth
within a 30 mm PVC pipe. The sensors were separated by insulating spacers. At
each measuring depth, the pipe was perforated to allow rapid thermal equilibrium.
The array was installed vertically into the riverbed adjacent to a field site where the
river and shallow aquifer levels (in a piezometer 56 m from the river) were monitored.
Temperatures were logged every 15 minutes from October 2009 to July 2010. The
47
resulting temperature time series were filtered, using a forward-backwards filter,

in order to create time series of amplitude ratio and phase shift data; the two-
directional filter is required to prevent the introduction of errors into the phase
[Hatch et al., 2006]. For filtering procedures, refer to Rau et al. [2010]. The
amplitude ratio and phase shift data were then used to estimate the Darcy velocity
and effective thermal diffusivity using Equations 3.11 and 3.12 (with 3.13 and 3.14).
For the purpose of comparing methods, the data were also interpreted using
Equations 3.9 and 3.10 (with 3.13 and 3.14). Effective thermal diffusivity values
used were from the literature [i.e. Shanafield et al., 2011]. Furthermore, previously
published data from another site within the same catchment were revisited. Rau
et al. [2010] had interpreted the data using Equations 3.9 and 3.10 (with 3.13 and
3.14). However, they observed that the velocity estimates from the two equations
were inconsistent with one another in some cases. In this chapter, the data are
reinterpreted using the proposed method, i.e., using Equations 3.11 and 3.12 (with
3.13 and 3.14).
3.4 Results and Discussion
3.4.1 Theory
Shown in Figures 3.2a and 3.2b are the solutions to Equations 3.9 and 3.10 (for ∆z
= 0.18 m). These illustrate that there is a unique value of thermal front velocity for
each Ar (Figure 3.2a), while there are two possible values of thermal front velocity for
each ∆Φ, symmetrical about 0 m/d (Figure 3.2b). Importantly, from these figures it
can be seen that if Ar and ∆Φ are used independently to estimate the thermal front
velocity, the uncertainty in effective thermal diffusivity will cause uncertainty in the
velocity calculation. For instance, Ar = 0.3 corresponds to a range of velocities from
-0.3 m/d (i.e., losing) to +0.4 m/d (i.e., gaining) (see Figure 3.2a). Also, ∆Φ = 0.14
to 0.36 d all correspond to 0 m/d (i.e., neutral) (see Figure 3.2b).
48
Wat
Surface water er tab
le
T1 A1
Water flow
Temperature
A2
t1 t2
T2
Time
Figure 3.1: Schematic of a temperature array installed in a riverbed. The concept of

amplitude ratio and phase shift is illustrated with the two temperature time series
at different depths. The temperature signal is dampened and time lagged as it
propagates downwards.
Plotting the Ar and ∆Φ results of Figures 3.2a and 3.2b against each other
(which can be done since both are a function of thermal front velocity) leads to the
conclusion that Ar and ∆Φ are not independent variables but that a relationship
exists between them (Figure 3.2c). The implications of this are significant: if the
values of Ar and ∆Φ are known, the value of effective thermal diffusivity can
be known also, and hence the value of thermal front can be determined. The
uncertainty in thermal front velocity normally associated with the uncertainty in
effective thermal diffusivity is thus eliminated.
The solution to Equation 3.11 is shown in Figure 3.2d as a surface of thermal
front velocities for different combinations of Ar and ∆Φ. Combinations of Ar and
∆Φ corresponding to reasonable values of effective thermal diffusivity (i.e., those
shown in Figure 3.2c) have been used. Viewing this surface from either the Ar
versus v or ∆Φ versus v perspectives, demonstrates that Equations 3.9 and 3.10
(see Figure 3.2a and 3.2b) are sub-solutions of Equation 3.11 (see Figure 3.2d).
49
1 0.4
a b
Dmin Dmin
0.8 Davg 0.3 Davg
Dmax Dmax
0.6
[d]
0.2
Ar
0.4
0.1
0.2
0 0
0 1 2 0 1 2
v [m/d] v [m/d]
1
c
Dmin
0.8
Davg
Dmax 2
0.6 d
v [m/d]
Ar
0.4 0
0
0.2
0.5
0.4
0 0.2
0 0.1 0.2 0.3 0.4 0 1 Ar
[d] [d]
Figure 3.2: (a) Amplitude ratio and corresponding thermal front velocity for
different values of effective thermal diffusivity. (b) Phase shift and corresponding
thermal front velocity for different values of effective thermal diffusivity. (c)
Amplitude ratio versus phase shift for different values of effective thermal diffusivity.
(d) Values of thermal front velocity for different combinations of amplitude ratio and
phase shift. Note that Dmin = 0.02, Davg = 0.075, and Dmax = 0.13 m2 /d. For all
plots ∆z = 0.18 m.
3.4.2 Interactions in Maules Creek Catchment
Collected data from the Namoi River for an 8 month period is shown in Figure
3.3. During the monitoring period there were two major flow events, consisting of
a dam release (approximately 1 m change in river level) in early December 2009,
followed by a natural flood (over 6 m increase in river level) in early January 2010
(Figure 3.3a, black line). These fluctuating river levels were reflected in the aquifer
level data, although in a dampened and lagged manner (Figure 3.3a, grey line).
The recorded temperature data ranged from approximately 10°C (winter) to 30°C
50
(summer) (Figure 3.3b). During summer the upper temperature sensor recorded
higher temperatures than the lower sensor, while this reversed during the winter.
Also, in the summer months the daily fluctuations were larger as compared to the
winter months.
226 a River
Aquifer
Level [m AHD]
224
222
220
218
30 b Upper
Temperature [oC]
Lower
25
20
15
10
11 12 01 02 03 04 05 06
Month (2009/2010)
Figure 3.3: (a) River (black line) and aquifer (grey line) level in meters above
Australian Height Datum (m AHD). (b) Recorded temperature at upper (5 cm;
black line) and lower (23 cm; grey line) sensors within the riverbed.
The interpreted data is shown in Figure 3.4. The river-aquifer hydraulic gradient
data (Figure 3.4a), demonstrate that during low flows prior to the major flow events,
there was potential for slightly gaining conditions. This reversed to losing conditions
due to the increased river stage during the events. This was then followed by a period
of gaining conditions which likely reflects a return of bank storage to the river.
In agreement with the gradient data, the Darcy velocities estimated using Equa-
tion 3.11 (Figure 3.4b), show that during low flows there were slightly losing to
slightly gaining conditions (-0.04 to 0.09 m/d), while during high flows this reversed
51
to losing conditions (to -0.58 m/d), followed by a period of gaining conditions (to
0.17 m/d). The Monte Carlo analysis showed that the uncertainty in γ is small (γ10
= 0.68; γ50 = 0.77; γ90 = 0.84) thereby leading to a narrow band of possible Darcy
velocities (Figure 3.4b, dark grey shading).
0.02
a
0
0.2 b
Effec. thermal diff. [m 2/d] Darcy velocity [m/d]
0.4 c
0.3
0.2
Dmax
0.1 Davg
Dmin
0
11 12 01 02 03 04 05 06
Month (2009/2010)
Figure 3.4: (a) Near-river hydraulic gradient in the aquifer. (b) Riverbed Darcy
velocity estimated using Equation 3.11. Dark grey shading indicates the range of
possible Darcy velocities given the uncertainty in γ. (c) Riverbed effective thermal
diffusivity estimated using Equation 3.12. Note that Dmin = 0.02, Davg = 0.075,
and Dmax = 0.13 m2 /d. The vertical grey shaded zone indicates a period where the
method is not quantitatively reliable.
52
The values for effective thermal diffusivity estimated using Equation 3.12 (Figure
3.4c), generally fall within the expected literature range, and are slightly higher
than the literature average, although it should be noted that these literature values
assume β is negligible in Equation 3.3. The effective thermal diffusivity results
computed are inclusive of the effect of thermal dispersivity which increases with
velocity. The estimated thermal diffusivities during low flows were slightly different
before and after the major flow events. This may indicate that scouring and sedi-
mentation/colmation processes are occurring. This may be either through reworking
of the riverbed or movement of suspended material into the pores, resulting in
changes in grain size and effective porosity.
During the major flow events, however, the values for effective thermal diffusivity
changed rapidly, and increased to unrealistic values (vertical grey shaded zone in
Figure 3.4c). This was due to the transient conditions during the major flow events
with velocities changing significantly on a time scale smaller than that used for
individual Ar and ∆Φ calculations. This violates a fundamental assumption of
the method, as the analytical solution to the convection-conduction equation (i.e.,
Equation 3.7) requires that fluid flow is at steady-state within a period of diel
temperature fluctuations. As the major departures in effective thermal diffusivity
occur during the major flow events, which is likely to be the period of greatest SW-
GW interaction, this is an inherent limitation of the proposed method, and of the
existing methods [e.g. Hatch et al., 2006; Keery et al., 2007].
3.4.3 Method comparison
The same data as that plotted in Figure 3.4b but reinterpreted using Equations 3.9
and 3.10 (with 3.13 and 3.14) are shown in Figure 3.5. For this reinterpretation,
literature values for effective thermal diffusivity were used [i.e. Shanafield et al.,
2011]. As the value of effective thermal diffusivity used increases, the velocities vary
more and tend toward gaining conditions.
53
Figure 3.5: Riverbed Darcy velocity estimated using (a) amplitude ratio data with
Equation 3.9 and (b) phase shift data with Equation 3.10 with literature values
for effective thermal diffusivity. No solution exists for the phase shift data when
effective thermal diffusivity equals Dmax . Dark grey shading indicates the range of
possible Darcy velocities given the uncertainty in γ. Note that Dmin = 0.02, Davg
= 0.075, and Dmax = 0.13 m2 /d.
Since Equations 3.9 and 3.10 are two solutions of the one equation (i.e., Equation
3.7), they should give the same results. However, this is not the case and the results
differ markedly depending on whether the Ar or ∆Φ approach was used (Figure
3.5a versus Figure 3.5b). Furthermore, it was not possible to calculate the Darcy
velocity with the ∆Φ approach when using the maximum effective thermal diffusivity
value (i.e., Dmax = 0.13 m2 /d) as solutions exist for only a narrow range of phase
shift values (see Figure 3.2b). From these results, it is not possible to give reliable
velocity estimates for the SW-GW interaction. Consequently, it is not even possible
to conclude if the river is gaining or losing water. The deviations indicate, given
that the assumptions behind Equatio n3.7 are valid, that incorrect values for effective
54
thermal diffusivity are being used in the analysis; the effective thermal diffusivity is
implicit in the obtained Ar and ∆Φ data (see Figure 3.2c).
The previously published data by Rau et al. [2010] but here reinterpreted using
Equations 3.11 and 3.12 (with 3.13 and 3.14) are shown in Figure 3.6. Previously, the
Darcy velocity results for Horsearm Creek (HC) derived from the Ar data differed
from those derived from the ∆Φ data (Figure 3.6a, dotted/dashed lines). Similarly,
the Darcy velocity results for Downstream Elfin Crossing (DEC) differed depending
on the data used (Figure 3.6c, dotted/dashed lines). However, when both the Ar
and ∆Φ data were used in 3.11, a single velocity result was obtained (Figure 3.6a
and 3.6c, solid lines). As there is uncertainty in the value of γ, bounds were added to
the reinterpreted Darcy velocities. The two interpretations of the data are different
and would lead to somewhat different conclusions about the system. For example,
at HC the new interpretation suggests the systems is neutral to slightly losing, while
the old interpretation suggested that perhaps the system was at times losing at rates
of more than -0.5 m/d.
55
0.2
Darcy velocity [m/d]

qA
a r
0 qΔΦ
q
−0.2
−0.4
−0.6
Effec. ther. diff. [m2/d] Darcy velocity [m/d] Effec. ther. diff. [m2/d]
DA
b r
0.1 DΔΦ
D
0.05
0
qA
c r
−0.2 qΔΦ
q
−0.4
−0.6
−0.8
DA
d r
0.1 DΔΦ
D
0.05
0
05/09 15/09 25/09 05/10 15/10 25/10
Year 2007
Figure 3.6: (a) Darcy velocities for Horsearm Creek estimated by Rau et al. [2010]
using amplitude ratio data (dotted line) and phase shift data (dashed line), and
estimated by Equation 3.11 (solid line) using both amplitude ratio and phase shift
data. Dark grey shading around solid line indicates the range of possible Darcy
velocities given the uncertainty in γ. (b) Effective thermal diffusivity values for
Horsearm Creek estimated using Equation 3.3 with thermal parameters adopted
by Rau et al. [2010] and the velocities derived from amplitude ratio data (dotted
line) and phase shift data (dashed line), and estimated using Equation 3.12 (solid
line) with both amplitude ratio and phase shift data. (c) As for (a) but for the
Downstream Elfin Crossing location. (d) As for (b) but for the Downstream Elfin
Crossing location.
56
The corresponding values of D (i.e., from Equation 3.12) to the derived values
of v (i.e., from Equation 3.11) are shown in Figures 3.6b and 3.6d. The effective
thermal diffusivity data agree well for both locations and appear to be increasing
with time. This could be due to a reduction in porosity over the monitoring period;
Rau et al. [2010] reported that colmation was taking place. Alternatively, it could
illustrate that the assumptions behind Equation 3.7 are being violated. Further
testing of the proposed method for its reliability under different scenarios (e.g.,
lateral flow conditions) is required.
Having proposed a method (i.e., Figure 3.2) for interpreting field data (i.e.,
Figures 3.3 and 3.4), and applied the existing methods to the data (i.e., Figure 3.5),
as well as reinterpreting previously published data from another site (i.e., Figure
3.6), it is now possible to make the following observations.
(1) When Equations 3.9 and 3.10 are used, a value for the effective thermal
diffusivity has to be chosen. Shanafield et al. [2011] have shown that uncertainties
in the effective thermal diffusivity lead to uncertainties in estimated thermal front
velocities. This issue is therefore highly significant. To overcome this problem,
literature values of the relevant thermal parameters are frequently adopted [e.g.
Fanelli and Lautz, 2008; Rau et al., 2010; Lautz et al., 2010; Jensen and Engesgaard,
2011]. Lapham [1989] is commonly referenced to justify this approach. However,
this research reported a range of thermal parameters for fine-grained and coarse-
grained sediment. The method proposed here by-passes this issue, and removes the
uncertainty in thermal front velocity normally associated with uncertainty in the
effective thermal diffusivity.
(2) When Equations 3.9 and 3.10 are used the issue of what the value for thermal
dispersivity should be is also raised. There is debate in the literature as to whether
dispersivity should be included within the effective thermal diffusivity term, and if
so, what the velocity relationship should be [e.g. De Marsily, 1986; Anderson, 2005].
Keery et al. [2007] suggested that the disagreements should promote research into
57
the matter. This has since been undertaken by Rau et al. [2012]; see Equation 3.3.
Again, the proposed method bypasses this issue as it does not explicitly or implicitly
set the dispersivity term to zero. Rather, the method computes the thermal front
velocity using Ar and ∆Φ data (i.e., Equation 3.11) where the effective thermal
diffusivity (inclusive of dispersivity) is implicitly included in the computation.
(3) It is not always clear when using Equations 3.9 and 3.10 which of the two
should be used, or why they often do not give the same results. Hatch et al. [2006]
notes that the two equations have differing sensitivities to velocity as well as to other
input parameters (e.g., sensor spacing). In practice, the velocity results derived from
Ar data often differ from those derived from ∆Φ data [e.g. Rau et al., 2010; Swanson
and Cardenas, 2010; Munz et al., 2011]. Furthermore, a trend is developing in the
literature which focusses on the Ar data only [e.g. Fanelli and Lautz, 2008; Vogt
et al., 2010; Schmidt et al., 2011; Gordon et al., 2012; Briggs et al., 2012]. As
outlined in the methodology, when the assumptions behind Equation 3.7 are valid,
which was the original basis for applying Equations 3.9 and 3.10 to field data, there
should not be a discrepancy in the velocity results. Or in other words, a value of
effective thermal diffusivity exists such that the velocity results will be equal. This
effective thermal diffusivity value is estimated using Equation 3.12.
(4) Since it is reasonable to assume that the effective thermal diffusivity does
not change rapidly or to values outside a reasonable range, the results can indicate
periods where the method assumptions are being violated. This provides a quality
check on the thermal front velocity results. While a rapid change in the value of
effective thermal diffusivity indicates that the method is invalid, it should be noted,
however, that steadiness of estimated thermal diffusivity does not necessarily indic-
ate times where all the model assumptions are valid. The steadiness of calculated
effective thermal diffusivity implies steadiness of thermal front propagation but does
not imply that fluid fluxes are one-dimensional or that the thermal properties of the
porous medium are constant in space and time. Nevertheless, the proposed method
58
provides a new metric for detection of violations. This is particularly helpful for
field data where transient conditions may exist.
(5) Finally, when using Equations 3.9 and 3.10, iterative routines are required
to solve for thermal front velocity. This has most recently been automated by
Swanson and Cardenas [2010] and Gordon et al. [2012]. However, the interactive
routines can be numerically intensive. In addition, in the case of Equation 3.10,
they are potentially unstable, although limits can be imposed to help with stability.
The method outlined in this chapter circumvents this issue by solving for thermal
front velocity as an explicit function of Ar and ∆Φ without iteration.
In summary, the proposed method overcomes five major issues by computing the
thermal front velocity using both Ar and ∆Φ data without the need to specify a
value for effective thermal diffusivity or the need for iteration (i.e., Equation 3.11).
In addition, the method allows for the estimation of the effective thermal diffusivity
(i.e., Equation 3.12). It should be noted though, in common with all methods
based on Stallman [1965], the proposed method has inherent limitations due to the
assumptions behind the model, namely, (1) fluid flow is in the vertical direction only,
(2) fluid flow is steady state, (3) fluid and solid properties are constant in both space
and time, and (4) fluid and solid temperatures at any particular point in space are
equal at all times.
3.5 Conclusions
A novel method to estimate the SW-GW interaction (i.e., Darcy velocity), and
effective thermal diffusivity, from pairs of temperature time series recorded at dif-
ferent depths at the SW-GW interface has been presented. The method does not
require that the effective thermal diffusivity is specified, nor does it use iteration
routines. Applying the method to data collected from Maules Creek Catchment
in the Murray Darling Basin (Australia) demonstrated its usefulness in revealing
SW-GW processes. In addition, the implementation of the method to data showed
59
that while the riverbed Darcy velocities were reliably estimated at low river flows,
the method had limitations under transient flows as seen in the estimated values of
effective thermal diffusivity. This was due to a violation of the underlying model
assumptions (i.e., that fluid flow is at steady-state within a period of one day). The
proposed method therefore also offers the ability to show periods when the model
assumptions are violated (although not where the model assumptions are valid). It
is expected that the method will be of wide usefulness in field studies of SW-GW
interactions, particularly where the anticipated Darcy velocities are low and the
thermal parameters are uncertain.
60
Chapter 4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Estimating interactions using dif-

ferential river gauging
4.1 Summary
In semi-arid and arid environments, leakage from rivers is a major source of recharge
to underlying unconfined aquifers. Differential river gauging is widely used to
estimate the recharge. However, the methods commonly applied are limited in
that the temporal resolution is event-scale or longer. In this chapter, a novel
method is presented for quantifying both the total recharge volume for an event,
and variation in recharge rate during an event from hydrographs recorded at the
upstream and downstream ends of a river reach. The proposed method is applied to
river hydrographs to illustrate the method steps and investigate recharge processes
occurring in a sub-catchment of the Murray Darling Basin (Australia). Interestingly,
although it is the large flood events which are commonly assumed to be the main
source of recharge to an aquifer, our analysis revealed that the smaller flow events
were more important in providing recharge.
61
Chapter 4. Estimating interactions using differential river gauging
4.2 Introduction
Recharge estimates are critical for the management of water resources in semi-
arid and arid environments [de Vries and Simmers, 2002; Scanlon et al., 2006].
In such environments, leakage from rivers (as well as channels and wadis) is a major
source of recharge to the underlying unconfined aquifers [Rushton, 1997]. Differential
river gauging, where the difference in river flow between successive cross sections is
calculated, is widely used to estimate the recharge from rivers [Scanlon et al., 2002;
Kalbus et al., 2006]. Many studies have made use of this method with good results
[e.g. Ruehl et al., 2006; Harte and Kiah, 2009], and when compared to other methods
(e.g., hydrometric, seepage meters) it has been found to give the best estimate of
recharge to unconfined aquifers from rivers [Cey et al., 1998; Kaleris, 1998]. As
the only data required are recorded river hydrographs at two locations, which in
most countries worldwide are readily available, the method is a powerful tool to the
hydro(geo)logist.
As the timing of peak flow at the upstream gauge is different from that at
the downstream gauge, the differential river gauging method can provide the total
recharge volume for an event but not the variation in recharge rate during an event.
The temporal resolution of the method is consequently assumed to be event-scale
or longer [Scanlon et al., 2002]. Therefore, while the average recharge rate can
be computed if the event duration is known, the variation in recharge rate during
an event cannot be determined. From the perspective of catchment-scale water
resource management this is acceptable, however, when the temporal aspect of
recharge processes is of concern, or when comparison with continuous measurements
of recharge is required (e.g., from hydrometric methods), the limitation that the
temporal resolution is event-scale or longer becomes important.
To our knowledge, only one study has used a differential river gauging method to
estimate the variation in recharge rate during an event in addition to total recharge
62
volume. Opsahl et al. [2007] established the relationship for all flow events between
transit time (i.e., the lag between peak flow at the upstream and downstream gauges)
and peak flow at the upstream gauge. This relationship was then used to adjust
the downstream data to an equivalent data set which could be directly compared
with the upstream data for each flow event. However, there was uncertainty in the
established relationship, which would likely increase if the method were applied to
regulated reaches in semi-arid or arid environments where flow regimes are highly
erratic and episodic.
In this chapter, a novel method is presented which allows for estimating the total
recharge volume for an event, as well as the variation in recharge rate during an event
from hydrographs recorded at upstream and downstream ends of a river reach. The
proposed method is applied to river hydrographs from the Namoi River to illustrate
the method steps and complete a preliminary investigation of the recharge from the
river to the aquifer in the Maules Creek Catchment, a sub-catchment of the Murray
Darling Basin (Australia).
4.3 Methodology
4.3.1 Theoretical background
The mass balance for a river reach is defined as (based on Lerner et al., 1990):
∆S
Qu + Qi + Qf = Qd + Qo + Ea + (4.1)
∆t
where Qu is flow at the upstream end of the reach, Qi is flow into the reach
(e.g., tributaries), Qf is river-aquifer flux (n.b., a positive value indicates a mass
gain to river; conversely, a negative value indicates a mass loss from river), Qd is
flow at the downstream end of the reach, Qo is flow out of the reach (e.g., surface
water diversions), Ea is evapotranspiration from the reach, and ∆S
∆t
is the change
in channel storage with time. All components of the mass balance have dimensions
63
L3 /T.
If ∆S
∆t
is assumed negligible, which is a reasonable assumption for analysis at the
event-scale or longer time-scales, Equation 4.1 simplifies to
Qf = Qd − Qu + Qo − Qi + Ea (4.2)
Equation 4.2 is the commonly used equation in the differential river gauging
method for estimating the river-aquifer flux (i.e., recharge). It has the limitation
that the temporal resolution is event-scale or longer, and so the recharge rate for
shorter time-scales cannot be determined. An alternative approach is thus required.
4.3.2 Proposed method
Starting from the same mass balance approach (i.e., Equation 4.1), the recharge rate
at shorter time-scales can be estimated, without making assumptions about ∆S
∆t
, from
hydrographs recorded at the upstream and downstream ends of a river reach. The
physical basis behind the proposed method is that although a flow event spreads
out with time as it travels between the upstream and downstream gauges, the mass
remains the same if no recharge from the river occurs. It is therefore possible to
create equivalent hydrographs by time-shifting the recorded hydrographs. These
can then be compared to check for mass conservation. This can be done by shifting
the downstream hydrograph backward in time, the upstream hydrograph forward in
time, or some combination. For the sake of illustrating the method, Ea , Qi and Qo
have been assumed to equal zero.
The proposed method consists of eight steps as outlined below. These were
scripted in Matlab (the variable names are in parentheses).
(i) Select flow event to be analysed (between start and end date/time) for
upstream and downstream hydrographs (i.e., Qu,d ).
(ii) Create cumulative hydrographs by integrating the hydrographs (i.e.,
CumQu,d ).
64
(iii) Create normalised cumulative hydrographs by dividing by the maximum

values in the cumulative hydrographs (i.e., N orCumQu,d ).
Note: if scripting these steps recreate normalised cumulative hydrographs by
discretising on the y-axis (i.e., between 0 to 1) rather than on the x-axis (i.e., between
start and end date/time).
(iv) For each value of the cumulative normalised flow curve on the y-axis, cal-
culate the value at a chosen point between CumQu and CumQd on the x-axis
(e.g., midpoint). This gives a time-shifted normalised cumulative hydrograph (i.e.,
N orCumQt ).
Note: if scripting these steps recreate the time-shifted normalised cumulative
hydrograph by discretising on the x axis (i.e., between start and end date/time)
rather than on the y axis (i.e., between 0 and 1).
(v) Create time-shifted cumulative hydrographs by multiplying by the maximum
values in the cumulative hydrographs (i.e., CumQtu,d ).
(vi) Create time-shifted upstream and downstream hydrographs by differentiat-
ing the cumulative hydrographs (Qtu,d ).
(vii) Create a time series of river-aquifer flux (n.b., a negative value indicates
recharge from river to aquifer) by subtracting the upstream from dowstream hydro-
graph
Qtf = Qtd − Qtu (4.3)
(viii) Repeat Steps iv to vii with different time-shifted normalised cumulative

hydrographs between the limits of shifting the downstream hydrograph backward
in time and shifting the upstream hydrograph forward in time to give a range of
possible time series of river-aquifer flux.
4.3.3 Field example
A reach of the Namoi River in New South Wales (Australia) (Figure 4.1) was invest-
igated using the method. The Namoi River is the main river in the Namoi Valley,
65
and is a tributary to the environmentally and politically sensitive Murray Darling

Basin. The alluvial aquifers along the river have one of the highest groundwater
abstraction levels in Australia.
A’
A
Figure 4.1: Location of the Namoi Catchment (main figure) within the Murray
Darling Basin and Australia (inset figures). Also shown are Maules Creek Catchment
and the upstream (Boggabri) and downstream (Turrawan) gauging stations. The
arrows adjacent to the rivers show the flow directions. Also shown is the location of
the geological cross section A-A’ shown in Figure 4.2.
A A’
Figure 4.2: Schematic geological cross section for Maules Creek Catchment (A-
A’ in Figure 4.1). The red dashed vertical lines show where lithological borehole
information is available. Elevation is shown in meters above the Australian Height
Datum (m AHD).
66
a b
Namoi
River
Figure 4.3: Photograph of (a) flood plain and Namoi River looking north and (b)
Namoi River and river banks at low-flow conditions looking south.
The selected reach of the Namoi River runs from south to north for 34 km
through the semi-arid Maules Creek Catchment. Beneath and to the east of the
river is a palaeochannel up to 120 m deep, filled with alluvial clays and permeable
sands and gravels (Figure 4.2). The gently sloping plains of the catchment, which
consist mostly of Holocene clay and silt rich vertosols, are 200 to 250 m above sea
level (Figure 4.3a). The Namoi River is incised into this flood plain up to a depth
of about 8 m (Figure 4.3b). Further geological details can be found in [McCallum
et al., 2013b].
Owing to groundwater abstraction for flood irrigation within the catchment,
the groundwater levels have been lowered such that the reach is now predominately
losing during both low- and high-flow conditions [McCallum et al., 2013b]. Recharge
from the river is now a major source of water for the ongoing groundwater abstraction
[Giambastiani et al., 2012]. Understanding and quantifying the recharge to the
unconfined aquifer is thus an important step in the management of the water
resource.
Government-operated gauging stations at the upstream (Boggabri) and down-
stream (Turrawan) end of the reach record river stage which is post-processed to
give river flow [DNR, 2011] (see Figure 4.1). For this chapter, hourly data from
the stations were used. The river is regulated at Keepit Dam, approximately 50
67
km south-east of the upstream gauging station. Between the two stations, there
is a major tributary, Maules Creek. While this creek has perennial pools about
10 km upstream from its confluence with the Namoi River [Rau et al., 2010], the
creek is ephemeral and rarely flows between the pools and the confluence [Andersen
and Acworth, 2009]. For this reach of the Namoi River, evapotranspiration can be
assumed negligible compared to the magnitude of the other components [McCallum
et al., 2013b].
The proposed method was applied to the gauging station data in two stages.
First, a flow event from 13 December to 26 December 2008 (i.e., approx. 14 days)
was selected to illustrate the method step-by-step. Second, high-flow events (i.e.,
> 1.5 GL/d) between the water years 2000 and 2010 were analysed (n.b., GL is
Gigalitre; e.g., 1.5 GL/d = 1.5 x 106 m3 /d; the water year runs from October to
September). In this way statistics of flow in the Namoi River and recharge in the
Maules Creek Catchment were created for each flow event (i.e., cumulative flow
volume at the upstream gauge, event duration, total recharge volume, and variation
in recharge rate during event). The start and end date/time was defined as the
date/time corresponding to the lowest flow at the upstream gauge in the 7 days
prior to/following a recorded flow of 1.5 GL/d. As the data set was large (i.e., a
decade of hourly data at two gauges), the analysis was automated using a Matlab
script.
4.4 Results and Discussion
4.4.1 Steps
The eight steps outlined in the methodology are illustrated in Figure 4.4 and dis-
cussed below:
(i) Hydrographs for the upstream and downstream gauges for the selected flow
event show that the downstream hydrograph (grey line) has a smaller and lagged
68
peak flow as compared to the upstream hydrograph (black line). The upstream
hydrograph peaked at 26.4 GL/d while the downstream hydrograph peaked at 22.5
GL/d. There was a lag of 18 hours between the peaks.
(ii) The cumulative flow for the upstream gauge was 85.7 GL while for the
downstream gauge this was 74.8 GL, that is, during the analysed flow event, 85.7
GL entered the catchment and 74.8 GL left the catchment. The difference between
these curves at the final time step (i.e., 10.9 GL) is equal to the total recharge
volume for the event.
(iii, iv) When the cumulative hydrographs are normalised, they still follow the
s-shape but become dimensionless, varying between 0 and 1 (solid lines). The time-
shifted normalised cumulative hydrograph also follows the s-shape, is dimensionless,
and varies between 0 and 1 (dotted line). The dotted line is midway between the
solid lines (i.e., shifting both the downstream hydrograph backward in time and the
upstream hydrograph forward in time).
(v) The time-shifted cumulative hydrographs for upstream (black line) and down-
stream (grey line) gauges have the same final values as the originals (i.e., 85.7 and
74.8 GL) but are shifted in time.
(vi) The difference between the curves at the final time step is equal to the
total recharge volume for the event (i.e., 10.9 GL). The time-shifted downstream
hydrograph (grey line) has a smaller peak flow as compared to the time-shifted
upstream hydrograph (black line). The peak flows are no longer lagged. By time-
shifting the recorded hydrographs, the hygrographs are therefore now comparable.
(vii, viii) From the time-shifted hydrographs, a timeline of the recharge can be
estimated (dotted line). Since different time-shifted normalised cumulative hydro-
graphs are possible (i.e., Steps iv and viii above), the timing of recharge cannot
be determined exactly (see dotted line and solid lines). However, the variation in
recharge rate can be bounded by the upstream and downstream limits (solid lines).
69
Figure 4.4: (i) Flow event hydrographs for upstream (black line) and downstream
(grey line) gauges. (ii) Cumulative hydrographs for upstream (black line) and
downstream (grey line) gauges. (iii, iv) Normalised cumulative hydrographs for
upstream (black line) and downstream (grey line) gauges. Also shown is the
time-shifted normalised cumulative hydrograph (dotted line). (v) Time-shifted
cumulative hydrographs for upstream (black line) and downstream (grey line)
gauges. (vi) Time-shifted flow event hydrographs for upstream (black line) and
downstream (grey line) gauges. (vii, viii) Recharge using time-shifted normalised
cumulative hydrograph (dotted line). Also shown are the upper and lower bounds
of the recharge (black and grey lines).
70
4.4.2 Recharge in Maules Creek Catchment
The results of applying the method to field data are shown in Figure 4.5 . Each
mark on the figure represents a recharge event that occurred sometime during the
water years 2000 to 2010. As would be expected, the total recharge volume over
an event increases with increasing cumulative flow volume at the upstream gauge
(Figure 4.5a). The total recharge volume ranges from 0.3 GL to 34.7 GL, while
the cumulative flows range from 3.5 GL to 188.9 GL. The correlation between total
recharge volume and cumulative flow is high (i.e., r = 0.97). The slope of the line
of best fit (approximately 0.2) is the total recharge volume expressed as a ratio of
cumulative flow. The scatter about the line indicates that the total recharge volume
is dependent on other variables (e.g., antecedent groundwater level), in addition to
the cumulative volume of the flow event.
71
Figure 4.5: (a) Total recharge volume versus cumulative flow. (b) Total recharge
volume versus duration of event. (c) Total recharge volume versus maximum rate of
recharge. Large natural flow events shown as circles and small natural flow events
and dam-release events shown as crosses. The illustrative flow event in Figure 4.4 is
plotted in as a single data point, with a total recharge of -10.9 GL, cumulative flow
of 85.7 GL, duration of 14.5 days, and maximum rate of recharge of -3.3 G/d.
The cumulative flow is dependent on two variables: the duration of the flow event
and the variation in recharge rate during the event. There is a clear relationship
between the total recharge volume for the event and the event duration (Figure
4.5b). The duration of flow events range from 4 to 71 days. While the correlation
is high (i.e., r = 0.97) between the total recharge volume and event duration, the
amount of scatter for events with durations less than 20 days is significant, indicating
72
that other important factors are at play.

When total recharge volume is plotted against maximum rate of recharge during
the event the data group into two distinct populations (Figure 4.5c). On the one
hand, there are data which lead to total recharge volumes of approximately 10GL
or less but have varying and large recharge rates of more than 3GL/d (shown as
circles). On the other, there are data which lead to total recharge volumes of upwards
of 30GL but have recharge rates of less than 1 GL/d (shown as crosses).
Recharge from the Namoi River in the Maules Creek Catchment during high-
flow conditions is therefore due to two distinct processes. Comparing the two data
populations with the raw hydrographs reveals that the circles correspond to large
natural flow events while the crosses correspond to small natural flow events and
dam-release events. The threshold between these two populations is approximately
4 GL/d. The large natural flow events with large rates of recharge do not necessarily
lead to large total recharge volumes for individual events. On the other hand, the
small natural flow events and dam-release events, while having relatively small flow
rates and relatively small rates of recharge, have potential to lead to large volumes
of recharge.
Over the analysed decade, twice as much recharge occurred due to small natural
flow events and dam-release events than the large natural flow events (193 GL versus
82 GL). This interesting fact has implications for water resource management. There
is a widespread perception that because of the natural occurring cycles of drought
and flood within semi-arid environments, large flow events are predominately re-
sponsible for replenishment of groundwater resources via recharge from rivers to
unconfined aquifers. For the studied decade in the studied catchment, this was
not the case. The analysis presented here suggests that the smaller flows are more
significant. This could have implications on the design of dam release events so as
to optimise recharge to the groundwater. This is presently an under-researched area
of water management [see Zammouri and Feki, 2005].
73
McCallum et al. [2013b] present data which show the annual groundwater ab-
straction for Maules Creek Catchment between 1985 and 2005 varied from 5.4 to
17.9 GL/yr with an average annual abstraction of 10.9 GL/yr. The recharge of 10.9
GL recorded by the differential river gauging in the illustrative example therefore
represents approximately the volume abstracted during the average irrigation year.
In the case of this event, recharge from the river from a single flow event was
sufficient to meet the average irrigation demand. However, as a consequence, the
downstream river flow would be reduced by the recharged volume, thereby impacting
the downstream users of surface water as well as the environment. Furthermore, for
the analysed decade, only six flow events had estimated total recharge volumes equal
to or greater than 10.9 GL (see Figure 4.5a), while the remaining forty six events
had an average total recharge volume of only 2.8 GL, indicating that the irrigation
demand would on average exceed the recharge from the river.
4.4.3 Method limitations
There are a number of limitations associated with the proposed method. First,
gauging data can have inherent uncertainties in accuracy. Oftentimes, due to the
difficulty of creating accurate rating curves, the data quality can be poor for very
low- and very high-flows. How this uncertainty is propagated into the results should
be assessed on a site-by-site basis. Understanding the uncertainties involved is
imperative to properly interpreting the results [Schmadel et al., 2010].
Second, following on from the previous limitation, the range of recharge volumes
and rates that can be calculated directly depends on the uncertainty and recording
frequency in the gauging data [Scanlon et al., 2002]. This creates a constraint on
the spatial and temporal resolution of the method. As the recharge volume must
be significantly higher than the uncertainties associated with the gauging station
measurements, the distance between the gauging stations must be sufficiently large
to overcome this issue [Kaleris, 1998]. The lag between hydrograph peaks that
74
can be resolved depends on the frequency of the collected data. In the illustrative
example, the lag of 18 hours was able to be resolved as hourly data were available.
This may pose a limitation for different catchments. Again, these issues should be
considered on a site-by-site basis.
Third, the differential river gauging method therefore gives an estimation of
fluxes over a selected reach length and is thus not sensitive to small-scale het-
erogeneities in the recharge [Kalbus et al., 2006]. This, however, is often not a
significant limitation as large-scale measurements often provide a better estimation
of the recharge than point-scale measurements [Cey et al., 1998]. This could thus be
considered an advantage rather than a limitation. The relationship between at-point
measurements and channel-reach measurements needs further investigation in any
case [de Vries and Simmers, 2002].
Fourth, as the differential river gauging method gives the net exchange of water
between the river and aquifer between gauging stations, it is conceivable that small-
scale inflows and outflows can occur simultaneously, leading to no net exchange of
water between river and aquifer [McCallum et al., 2012b]. Where needed, the in-
flows, outflows, and net exchange can be distinguished by combining the differential
river gauging method with tracer tests [e.g., Ruehl et al., 2006].
Finally, and perhaps most importantly, while Step vii refers to the “river-aquifer
flux”, this needs to be interpreted with an understanding of the specific river hydro-
logy. Conceptualising the interaction as simply being between two reservoirs (i.e.,
river and aquifer) may be too simplistic. River-aquifer interactions occur at different
spatial and temporal scales, which superimpose on one another, causing dynamic
and complex patterns of interaction [Angermann et al., in press]. Furthermore, there
are often various intermediate stores and flows which can impact on the results as
they operate on time-scales of weeks to months after the peak event flow (e.g., return
flow from bank storage, interflow within the unsaturated zone, flow from draining
pools on the floodplain, and so on). These need careful consideration on a site-
75
by-site basis. For example, the computed total recharge volume may not represent
actual recharge but potential recharge only due to the presence of perched aquifers
[Lerner et al., 1990] or bank storage [Lambs, 2004].
These five limitations are in common with differential river gauging methods
generally. Further work is required in applying the proposed method to a variety
of hydrogeological settings, as well as comparing the estimations of total recharge
volume and variation in recharge rate with other techniques of estimating recharge
from rivers to unconfined aquifers (e.g., seepage meters, heat as a natural tracer).
4.5 Conclusions
By applying this method to hydrographs recorded at the upstream and downstream

ends of a river reach, one can estimate the total recharge volume for a flow event, as
well as the variation in recharge during the event. The proposed method overcomes
the significant limitation in previously used differential river gauging methods that
assume the temporal resolution must be event-scale or longer.
Applying the method to field data allowed for statistics of flow in the Namoi River
and recharge in the Maules Creek Catchment to be generated for each flow event
(i.e., cumulative flow volume, event duration, total recharge volume, and variation
in recharge rate). The statistical data were useful in revealing that two recharge
processes are occurring during high-flows: (1) large natural flow events with larger
rates of recharge, and (2) small natural flow events and dam-release events with
smaller rates of recharge but with the potential for larger total recharge volumes.
The method can be readily transferred to other catchments worldwide where
river flows are being monitored.
76
Chapter 5 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Spatial-temporal variation in inter-

actions
5.1 Summary
A critical hydrological process is the interaction between rivers and aquifers. How-
ever, determining the interactions from one method alone is difficult. At a point,
the water exchange through the riverbed can be determined using temperature
variations over depth. Over the river reach, differential gauging can be used to
determine losses or gains. This chapter combines these two methods, and applies
them to a 34 km river in semi-arid eastern Australia. It is found that high and
low river flows translate into high and low riverbed Darcy velocities, and that
these are strongly losing during high-flows, and slightly losing or gaining for low-
flows. The spatial variability in riverbed Darcy velocities may be explained by
riverbed heterogeneity, with higher variability at greater spatial scales. Although
the river-aquifer gradient is the main driver of riverbed Darcy velocity at high-flows,
considerable uncertainty in both the velocity magnitude and direction estimates
was found for low-flows. The reach scale results demonstrate that high-flow events
can account for 69% of the groundwater recharge. By examining the relationship
between total flow volume, river stage and duration, it was shown that the channel
geomorphology is of importance in controlling the volume of recharge, with smaller
77
Chapter 5. Spatial-temporal variation in interactions
flows potentially leading to greater recharge efficiencies than larger flows of similar
duration. Permeability contrasts between riverbed/riverbanks may explain these
observations. This work also suggests that up-scaling river-aquifer interactions is
problematic, especially during high-flows, and has implications for the modelling of
connected water resources.
5.2 Introduction
Industry, agriculture and growing urban populations are placing water resources
under severe stress [Gilbert, 2012]. In semi-arid and arid environments, groundwater
has become of fundamental importance to meet water requirements [de Vries and
Simmers, 2002]. As water is essential to sustainable global development [UNESCO,
2009], it is necessary that the profound changes occurring to natural hydrological
processes are understood and quantified.
A critical hydrological process is the interaction between rivers and aquifers
[Jones and Mulholland, 2000]. Where groundwater levels are naturally or artificially
lower than the riverbed, the flux of water from the river to the aquifer will be con-
trolled by three factors: the height of the river water, heterogeneity of the riverbed
hydraulic conductivity, and whether an area of saturation can be maintained between
the river and aquifer or whether decoupling will occur [Desilets et al., 2008; Brunner
et al., 2009a]. Since rivers experience large variations in flow in both space and time,
and the hydraulic properties of the riverbed also vary, determining the river-aquifer
interaction processes can be difficult from one field method alone. Moreover, many
field measurements are often taken at one location and for short periods of time
(i.e., campaigns), therefore the ability to extrapolate riverbed exchange over larger
spatial and temporal scales is empirically constrained [e.g., Kennedy et al., 2008].
The interaction between rivers and aquifers can be measured at a point within
the riverbed, inferred from hydraulic gradients, estimated from riverbed temperat-
ure profiles, assessed using geochemical tracers, derived from analytical solutions
78
describing recharge from the flood wave, computed over a river reach from the
mass balance of surface flow, and from many other methods [e.g., Cook, 2012;
Anderson, 2005; Hatch et al., 2006; Constantz, 2008; Barlow et al., 2000; Ruehl et al.,
2006; USGS , 2008] (for a thorough review of all methods see Kalbus et al. [2006]).
Generally, the field methods are of two types: those that attempt to measure the
flux directly through the river-aquifer interface, and those that measure volumetric
flow changes along the river [Becker et al., 2004]. Using the first methodology gives
estimates of the river-aquifer interactions at the point of measurement, while using
the second gives spatially-averaged estimates. A combination of both approaches can
therefore yield insights into the water exchange process operating within catchments.
Despite the general theory of river-aquifer interactions being well developed [e.g.,
Winter et al., 1998], few field studies have measured both the spatial and temporal
variation of water fluxes from multiple methods [e.g., Su et al., 2004; Constantz
et al., 2003].
This chapter examines the water flux characteristics along a 34 km reach of
a low gradient alluvial meandering river in semi-arid eastern Australia. Heat is
used as a tracer to measure temporal changes in riverbed temperatures to derive
riverbed Darcy velocities at two spatial scales: high density single pool deployment
and evenly spaced reach deployment, for a variety of flow conditions. On a larger
scale, these results are then complemented with an analysis of the water balance
between two gauging stations over the entire 34 km reach in which our point
measurements were made. Importantly this allows for a qualitative assessment of the
processes controlling the temporal variations in the water balance and river-aquifer
interactions in general. This is intended to: (1) identify the processes controlling the
variation in the timing and magnitude of riverbed Darcy velocities, and (2) compare
these processes with the dynamics of the water balance over the whole river reach
as determined from differential gauging.
79
5.3 Study area
The Murray Darling Basin (MDB) is Australia’s largest river basin (Figure 5.1),
covering 14% of the continent’s surface area, and generates about 40% of the national
income derived from agricultural production. Within the basin, agricultural issues
and sustainable water management are environmentally and politically sensitive
[MDBA, 2010; ABS , 2012].
The Namoi River Catchment is a major catchment in the MDB, with a surface
area of about 42,000 m2 . The Namoi River is over 350 km long, and ranges in
elevation from over 1,500 m in its headwater tributaries in the east, to 100 m in
the west. It has one of the highest groundwater abstraction levels in Australia.
Sharing of water between competing users, as well as the environment, has led to
the establishment of rules as to how water can be used and traded in the catchment
[Burrell et al., 2011].
Figure 5.1: (a) Location of the Murray-Darling Basin in Australia. (b) Location
of the Namoi River Catchment. (c) Maules Creek Catchment in New South Wales,
Australia. Also shown are the upstream (Boggabri) and downstream (Turrawan)
gauging stations.
The study focuses on a 34 km reach of the Namoi River where it runs south to
north through a semi-arid tributary catchment, Maules Creek. The alluvial plains
80
of the catchment are situated 200 to 250 m above sea level. To the north and east is
the Nandewar Range, which has peaks rising to 1,500 m. Below and to the east of
the Namoi River is a 120 m deep palaeochannel filled with clay, sands and gravels.
Groundwater is abstracted for flood irrigation as well as for stock and domestic use.
Some surface water is also diverted for irrigation purposes, although this volume
is small and comprises only about 7% of the total irrigation usage [Andersen and
Acworth, 2009]. Maules Creek is the main tributary, flowing from the mountains
in the east, and is generally ephemeral although it does have perennial flow in
its mid-reach [Rau et al., 2010]. Due to its intermittent nature, Maules Creek only
contributes flow to the Namoi River during infrequent floods [Andersen and Acworth,
2009].
Groundwater recharges at the mountain front in the east of the catchment and
in the past it discharged into the Namoi River in the west. However, due to
groundwater abstraction which started in the 1980s, the groundwater levels have
been lowered such that the river reach has changed from being predominately gaining
to predominantly losing during non-flood periods McCallum et al. [2013b].
Further background information about the catchment, including details of the
hydrological and geological context, can be found in Giambastiani et al. [2012] and
McCallum et al. [2013b].
5.4 Methodology
5.4.1 Point measurements from temperature time series
Theory
The convection-conduction equation for one-dimensional fully saturated conditions

can be formulated as
∂T ∂ 2T ∂T
=D 2 −v (5.1)
∂t ∂z ∂z
81
where T is temperature, t is time, z is depth, v is thermal front velocity and D

is effective thermal diffusivity [Suzuki, 1960].
For the situation of a semi-infinite half space with a sinusoidally varying tem-
perature at the upper surface, McCallum et al. [2012a] presented an equation which
uses the dampening and phase shift in recorded temperature over depth to estimate
the thermal front velocity, i.e.,
∆z (P 2 ln2 Ar − 4π 2 ∆Φ2 )
v= √ (5.2)
∆Φ 16π 4 ∆Φ4 + 8P 2 π 2 ∆Φ2 ln2 Ar + P 4 ln4 Ar
where is P is period, Ar is amplitude ratio and ∆Φ is phase shift.

The thermal front velocity and Darcy velocity are then related as
q = γv (5.3)
nρw cw + (1 − n) ρs cs
γ= (5.4)
ρ w cw
where q is Darcy velocity, n is porosity, ρw and cw are the density and heat
capacity of water, and ρs and cs are the density and heat capacity of solids.
The solution of McCallum et al. [2012a] for thermal front velocity eliminates the
need to specify a value for effective thermal diffusivity (i.e., v is a function of Ar
and ∆Φ but not D).
Field measurements
Two types of temperature array designs were used for field deployment. The first
consisted of temperature sensors (Onset HOBO Pro v2) at 0.00, 0.15 and 0.30 m
depth within a 32 mm diameter PVC pipe. The sensors were separated by insulating
spacers. At each measuring depth the pipe was perforated to allow rapid thermal
equilibrium. The second design consisted of pressure/temperature sensors (Solinst
Levelogger Gold 3001) at 0.00 and 1.00 m depth and temperature sensors (Onset
82
HOBO Pro v2) at 0.18 and 0.34 m depth. The other features remained the same
as in the first design. The arrays were installed vertically into the riverbed and the
locations were determined using DGPS.
Temperature arrays were installed in the riverbed in two separate deployments:
within a single pool and then evenly spaced along the 34 km reach. For each
deployment six arrays were installed (Figure 5.2a and 5.2b). It should be noted
that the two deployments spanned different periods: the pool arrays were deployed
from November 2007 to April 2008 (i.e., 164 days) and the reach arrays from October
2009 to December 2010 (i.e., 415 days). The pool is monitored for river stage and
is adjacent to the site where the groundwater level is measured (~ 40 m from the
river). Both were logged every 15 minutes.
83
a
River edge
N
Aquifer
P2 P3
P5 P6
P1 P4 River flow
River
R2
50 m
T
b
N
R1
R2
Namoi
River Maules
R3 Creek
R4
R5
5 km
R6
River flow
B
Figure 5.2: Temperature arrays (shown as circles) in two different deployments:

(a) Pool (i.e., 10s meters) and (b) Reach (i.e., 1000s meters). Also shown are the
gauging stations (shown as triangles).
84
Data processing and interpretation
The raw temperature time series were filtered (i.e., two-pass forward and backward
bandwidth filtering using a Tukey window with all-pass frequencies of 0.9 to 1.1
cycles per day) to extract time series of temperature amplitudes and phases [see
Rau et al., 2010]. Two pairs of temperature time series were used from each array:
the upper and second sensors and the second and third sensors. From these the
temperature amplitude ratios, phase shifts and time series of Darcy velocity (q,
positive = gaining; negative = losing) were calculated for each array using Equations
5.2 to 5.4.
Basic statistics (i.e., minimum, maximum, average, standard deviation) were
computed for all point measurement separately for both low-flows and high-flows.
The threshold between low-flow and high-flow for all analyses within this chapter is
defined 1.5 GL/d, based on a previous analysis of flow duration curves [McCallum
et al., 2013b].
The correlation between the riverbed Darcy velocities obtained from the first and
second thermistors and that from the second and third thermistors was calculated
for each array to check if the 1D vertical flow assumption is valid. The correlation
between the river stage and the riverbed Darcy velocity was also computed for each
array, to determine the degree to which variation in river stage drives the changes
in the riverbed Darcy velocity.
5.4.2 Reach measurements from differential gauging
Gauging stations are located at the upstream (Boggabri) and downstream (Tur-
rawan) ends of the reach separated by 34 km (Figures 5.1 and 5.2b). These stations
record the river stage each hour which is then post-processed to give flow using a
calibrated rating curve [DNR, 2011].
Hourly data for the period 2007-2012 (water years) were used to calculate the
total cumulative flow and cumulative low-flow at both gauging stations. This data
85
set covers the period of the temperature array deployments, and also overlaps with
that presented in McCallum et al. [2013a] (i.e., water years 2000 to 2010).
Each flow event within this 5-year period was then extracted and analysed
using the method of McCallum et al. [2013a], whereby hydrographs recorded at
the upstream and downstream ends of a river reach are used to estimate the total
recharge volume for a flow event, as well as the variation of recharge during the
event.
The starting point of the method is the mass balance for a river reach:
∆S
Qu + Qi + Qf = Qd + Qo + Ea + (5.5)
∆t
where Qu is flow at the upstream end of the reach, Qi is flow into the reach
(e.g., tributaries), Qf is river-aquifer flux (n.b., a positive value indicates a mass
gain to river; conversely, a negative value indicates a mass loss from river), Qd is
flow at the downstream end of the reach, Qo is flow out of the reach (e.g., surface
water diversions), Ea is evapotranspiration from the reach, and ∆S
∆t
is the change
in channel storage with time. All components of the mass balance have dimensions
L3 /T.
The total recharge volume can be computed by:
TR = (5.6)
X X
Qd − Qu
The method then simplifies further calculations by introducing the concepts of

time-shifted upstream and downstream hydrographs (i.e., Qtu and Qtd ). The recharge
at any point in time during an event can then be estimated using:
R = Qtd − Qtu (5.7)
while the recharge efficiency can be computed by:
86
Qd − Qu
P P
RE = P (5.8)
Qu
For these calculations, the time intervals are equal for each component in the
original mass balance, and the integration is over the event duration. The start/end
time for each flow event was defined as the time corresponding to the lowest flow
at the upstream gauge in the 7 days prior to/following a recorded flow of 1.5 GL/d.
For the variation of recharge (R) and total recharge volume (T R), a negative value
indicates a loss (recharge) from the river to the aquifer.
For each flow event the event duration, maximum upstream stage and flow, total
recharge (i.e., using Equation 5.6), maximum and average recharge rates (i.e., using
Equation 5.7), and recharge efficiency (i.e., using Equation 5.8) were calculated.
5.5 Results
5.5.1 Point measurements from temperature time series
A representative 14 day period of temperature data collected during a flood event

is shown in Figure 5.3. The measured temperature time series show strong diel
heat patterns as well as longer-term heat trends caused by changes in the weather
(i.e., storm) and noise (Figure 5.3b). The filtered time series more clearly reveals
the dampening of amplitude and shift in phase with depth at each location (Figure
5.3c).
87
Water level [m AHD]

226 a
224
222
220
b
Temperature [oC]
28
26
24
1 c
Temperature [oC]
0.5
−0.5
−1
M T W T F S S M T W T F S S M
Figure 5.3: (a) River hydrograph for 14 days. (b) Unprocessed temperature data
from temperature array P1 for depths 0, 15 and 30 cm, same period. (c) Temperature
data filtered for frequencies of 0.9 to 1.1 cycles per day using bandwidth filtering.
During low-flows the temperature data for all arrays in the pool gave Darcy
velocity results that indicated approximately neutral conditions (~ 0.0 m/d), while
during high-flows they showed losing conditions (up to ~ -0.3 m/d) (Figure 5.4c;
Table 5.1). In some cases slightly gaining conditions (~ 0.1 m/d) were observed
following flow events. Over depth, the correlation between velocities estimated from
probes 1 and 2 (v1−2 ) and probes 2 and 3 (v2−3 ) was significant (r >0.65) at three
locations within the pool (i.e., arrays P2, P4 and P6) (Table 5.1). At these same
locations the correlation between stage and flux was also significant (Table 5.1).
88
Water level [m AHD]

226 a
River
Aquifer
224
222
220
0.02
b
R-A gradient [−]
−0.02
−0.04
−0.06
c
−0.5 P1
P2
P3
P4
P5
P6
−1
Dec Jan Feb Mar Apr
2007/2008
Figure 5.4: Results for the pool deployment. (a) River and aquifer hydrographs. (b)
River-aquifer (R-A) gradient (negative value indicates potentially losing conditions).
(c) Estimated Darcy velocities from riverbed temperature data.
Two main flow events were recorded during the reach monitoring period (Figure
5.5a, solid line). One was a large event caused by rainfall (~ 6 m stage increase) and
the other was a smaller event caused by a dam release (~ 1 m stage increase) which
preceded the rainfall event. As for the pool data, the groundwater levels responded
in a damped fashion to each of the flow events (Figure 5.5a, dashed line). The
corresponding river-aquifer gradient in Figure 5.5b shows losing conditions during
89
Array Low-flows High-flows All-flows
Data Min Max Avg (µ) Std Dev (δ) Data Min Max Avg (µ) Std Dev (δ) Data Cor. 1 Data Cor. 2
[-] [m/d] [m/d] [m/d] [m/d] [-] [m/d] [m/d] [m/d] [m/d] [-] [-] [-] [-]
Pool Data
P1 50 -0.07 0.02 -0.01 0.03 45 -0.25 0.18 -0.04 0.06 150 0.14 95 -0.43
P2 58 -0.13 0.04 0.02 0.03 62 -0.26 0.20 -0.08 0.08 93 0.73 120 -0.76
P3 58 -0.23 0.06 0.00 0.04 56 -0.28 0.06 -0.04 0.07 115 0.33 114 -0.41
P4 57 -0.13 0.03 -0.01 0.02 65 -0.19 0.04 -0.08 0.05 141 0.84 122 -0.65
P5 35 -0.06 0.03 0.00 0.02 12 -0.11 0.07 -0.06 0.05 44 0.54 47 -0.62
P6 43 -0.04 0.02 0.01 0.02 25 -0.21 0.01 -0.07 0.07 37 0.97 68 -0.86
90
Reach Data
R1 233 -0.58 0.09 -0.05 0.07 31 -0.58 0.07 -0.17 0.13 264 0.31 264 -0.57
R2 201 -0.07 0.09 0.01 0.02 61 -0.84 0.31 -0.10 0.23 262 0.79 262 -0.66
R3 32 -0.09 -0.02 -0.05 0.02 23 -0.15 -0.02 -0.08 0.04 55 0.68 55 -0.42
R4 34 -0.15 0.02 -0.06 0.05 24 -0.58 0.01 -0.34 0.16 58 0.81 58 -0.80
R5 195 -0.99 -0.13 -0.23 0.13 46 -1.20 -0.17 -0.54 0.26 12 0.92 241 -0.52
R6 207 -0.27 0.04 -0.05 0.03 53 -0.55 -0.05 -0.18 0.14 257 0.95 260 -0.78
Table 5.1: River-aquifer statistics (see Figures 5.4 and 5.5). “Cor. 1” refers to correlations between v1−2 and v2−3 and “Cor. 2” refers
to correlations between stage and flux (v1−2 ). Bold indicates |Correlation coefficient| ≥ 0.65.
the high flood period with hydraulic gradients varying from -0.02 to -0.06. Following
the flow events, gaining conditions with hydraulic gradients up to 0.01 were observed
to last for a period of almost 3 months.
During low-flows the Darcy velocities from the reach data indicated neutral
to slightly losing conditions (+0.05 to -0.10 m/d) with the exception of array R5
which showed overall losing conditions (approx. -0.15 m/d) (Figure 5.5c; Table 5.1).
During high-flows, the Darcy velocities showed losing conditions (up to -1.2 m/d)
(Figure 5.5c; Table 5.1). Immediately after the high-flow event there were gaining
conditions (>0.1 m/d) at array R2 which diminished over one and a half weeks to
neutral conditions (Figure 5.5c). The correlation between v1−2 and v2−3 was greater
than 0.65 at all locations along the reach except for one (array R1) (Table 5.1). At
three locations the correlation between river stage and flux was greater than 0.65
(i.e., arrays R2, R4 and R6) (Table 5.1).
5.5.2 Reach measurements from differential gauging
The hydrographs for Boggabri and Turrawan (Figure 5.6a; black and grey lines)
show that river flows are episodic and vary from 0 to 79 GL/d for the period 2007 to
2012. Low-flows, defined as flows below 1.5 GL/d (Figure 5.6a and 5.6b, horizontal
line), occur 80% of the time and high-flows account for the remaining 20%.
The cumulative hydrographs for Boggabri and Turrawan (Figure 5.6c; black and
grey lines) show that over the five year period, 2,031 GL entered the catchment
while 1,730 GL left the catchment, which corresponds to a total loss of 301 GL. In
terms of low-flows, the cumulative hydrographs for Boggabri and Turrawan (Figure
5.6d; black and grey lines) show that over the five year period, 411 GL entered the
catchment while 318 GL left the catchment as low-flows. This corresponds to a total
low-flow loss of 93 GL. When the losses during low-flows are divided by the total
losses, low-flows can account for 31%, and high-flows 69%, of the total loss.
91
Water level [m AHD]
226 a
River
Aquifer
224
222
220
0.02 b
R−A gradient [−]
−0.02
−0.04
−0.06
c
−0.5 R1
R2
R3
R4
R5
R6
−1
Jan Apr Jul
2009/2010
Figure 5.5: Results for the reach deployment. (a) River and aquifer hydrographs. (b)
River-aquifer (R-A) gradient (negative value indicates potentially losing conditions).
(c) Estimated Darcy velocities from riverbed temperature data.
92
80 a
Flow [GL/d] 60
40
20
0
b
10
Flow [GL/d]
0
2000
c
Cum. Flow [GL]
1500
1000
500
0
400 d
Cum. LF [GL]
300
200
100
0
07 08 09 10 11 12
Year
Figure 5.6: (a) River hydrographs. (b) River hydrographs showing flows less than 1.5
GL/d. (c) Cumulative hydrographs for up- and downstreams flows. (d) Cumulative
hydrographs for low-flows (LF). In (a) to (d) the black line is upstream flow (at
Boggabri) and grey line is downstream flow (at Turrawan).
For the 2007-2012 period, high-flow events (i.e., flows greater than 1.5 GL/d)
were analysed (see Table 5.2 and Figures 5.7 and 5.8). The durations varied from
1 to 57 days, maximum stage from 1.2 to 7.6 m, maximum river flow from 1.2 to
93
761 GL/d, total recharge from -0.2 to -81.4 GL, maximum rate of recharge from
-0.2 to -10.4 GL/d, average rate of recharge from -0.1 to -1.7 GL/d, and recharge
efficiencies from 8 to 48%. Excluded from these reported ranges is one event which
had positive recharge results (i.e., gain of water to river) (see event 12 in Table 5.2).
The reason for this gain is unclear, but it is possible that surface flow contributions
from Maules Creek masked any groundwater recharge for this event.
94
Event No. Date Duration Stage Flow TR Max R Avg R RE

[d] [m] [GL/d] [GL] [GL/d] [GL/d] [%]
1 10/12/06 16 1.49 2.1 -6.2 -0.6 -0.38 29.37
2 24/12/06 13 1.33 1.8 -4.2 -0.4 -0.33 22.96
3 30/12/06 6 1.24 1.6 -1.4 -0.3 -0.25 19.60
4 04/03/07 13 2.99 6.5 -2.0 -1.6 -0.15 25.58
5 13/06/07 14 3.57 8.7 -2.8 -1.6 -0.20 18.19
6 23/08/07 15 4.44 11.2 -4.4 -2.0 -0.28 16.62
7 27/12/07 14 4.01 12.4 -4.0 -0.9 -0.24 9.18
8 13/02/08 14 7.09 10.5 -3.3 -10.3 -1.48 21.63
9 2/12/08 20 6.42 49.1 -29.1 -3.3 -0.75 12.76
10 19/12/08 15 1.37 26.4 -10.9 -0.4 -0.3 21.53
11 16/12/09 8 1.28 1.8 -2.4 -0.2 -0.13 12.37
12 24/12/09 8 2.47 1.6 -1.0 0.02 0.14 -7.74
13 18/01/10 8 1.82 2.8 -0.9 -0.4 -0.11 14.26
14 19/07/10 14 2.75 5.6 -2.4 -0.8 -0.17 16.45
15 18/08/10 38 6.49 28.1 -18.1 -2.2 -0.48 8.04
16 05/09/10 1 1.25 1.5 -0.2 -0.3 -0.31 19.90
17 21/10/10 6 1.63 2.4 -3.5 -1.1 -0.55 48.19
18 29/10/10 11 1.47 2.0 -5.8 -0.9 -0.51 46.20
19 20/11/10 19 5.30 16.8 -32.8 -3.3 -1.75 27.43
20 28/12/10 57 7.62 79.4 -81.4 -7.2 -1.44 10.69
21 03/02/11 11 1.42 1.9 -2.9 -0.3 -0.25 17.19
22 10/02/11 4 1.31 1.6 -0.8 -0.2 -0.20 14.90
23 14/02/11 12 1.35 1.7 -1.9 -0.2 -0.15 12.53
24 25/06/11 17 2.37 4.4 -4.1 -0.6 -0.25 13.38
25 14/09/11 14 1.35 1.7 -2.2 -0.3 -0.16 16.33
Table 5.2: Analysis of high-flow events from upstream and downstream gauging
stations (see Figures 5.7 and 5.8).
95
a 8
6
River stage (m)
0
0 20 40 60 80 100
River flow (GL/d)
b 8
6
River stage (m)
0
0 -20 -40 -60 -80 -100
Recharge (GL)
Figure 5.7: For all high-flow events (i.e., flows larger than 1.5 GL/d) for the analysed
period (i.e., 2007-2012), (a) the observed relationship between river stage and river
flow for upstream (black crosses) and downstream (grey crosses) gauging stations;
(b) the observed relationship between river stage and recharge volume. For (b) the
triangles represent events with peak flows less than 4GL/d, the circles represents
between 4 GL/d and 25 GL/d, and the squares represents more than 25 GL/d (i.e.,
overbank flows).
96
a -8
Recharge (GL) -6
-4
-2
0
0 2 4 6
River stage (m)
b -8
-6
Recharge (GL)
-4
-2
0
0 5 10 15 20
Event duration (d)
c -8
-6
Recharge (GL)
-4
-2
0
0 10 20 30 40
Total river flow (GL)
Figure 5.8: Dependence of recharge volumes for within-channel high-flow events

(i.e., flows larger than 1.5 GL/d but less than 25 GL/d), on (a) river stage, (b)
event duration, and (c) total river flow. In all plots, the triangles represent events
with peak flows less than 4GL/d and the circles represents more than 4 GL/d.
97
5.6 Discussion
5.6.1 Influence of river flow events on groundwater recharge
All the point measurements of Darcy velocities within the riverbed show similar
responses to flow events, however, they have divergent responses during recession
and low-flows (Figures 5.4c and 5.5c; Table 5.1). During low-flow conditions, the
point scale measurements indicate slightly gaining to slightly losing velocities. Dur-
ing a flow event, however, these velocities uniformly respond by becoming higher
magnitude losses. During flow recession, one point measurement (R5) indicates
an immediate reversal to gaining conditions (in both riverbed Darcy velocity and
gradient data), while the rest of the measurements return to a low slightly gaining
to slightly losing Darcy velocity.
The most obvious explanation for the temporal variability of riverbed Darcy
velocities and their magnitude is the coeval variation in river stage (Figures 5.4a
and 5.5a), which drives the change in hydraulic gradient throughout the reach
(Figures 5.4b and 5.5b). Given the smaller variation in the groundwater level over
time (Figures 5.4a and 5.5a), it is clear that the dynamic variation of river height,
particularly the flood wave, is the primary driver of changes in the Darcy velocity
through the riverbed (Figures 5.4c and 5.5c). The temporal match between the
change in hydraulic gradient and the independent measurements of riverbed Darcy
velocities gives confidence to the temperature based method.
These point scale observations are complemented by an analysis of the differential
river gauging data for the river reach (Figures 5.6). Under low-flow conditions, which
have occurred 80% of the time in the 5-year record examined here, groundwater
recharge through the riverbed can only account for 31% of the total losses in the
river water balance. Although slightly losing conditions are maintained for the
majority of the time, the Darcy velocities (calculated from the temperature data)
are very low and result in low overall river losses to the aquifer. In contrast, high-flow
98
conditions, which have occurred only 20% of the time, account for the remaining
69% of the losses to the overall river water balance as groundwater recharge. This
supports the point scale results in highlighting the dominant role of high-flow events
in shallow alluvial aquifer recharge in semi-arid and arid systems, a finding similar
to that of previous studies [Shentsis et al., 1999; Dahan et al., 2007].
5.6.2 Influence of channel geomorphology on groundwater
recharge
Additional insight into the influence of river flows on groundwater recharge is gained
by further analysis of the flow events (see Table 5.2 and Figures 5.7 and 5.7).
Through rainfall-runoff processes and dam-releases, the Namoi River experiences
a wide range of magnitude in the flows (Figure 5.7a). The routing of this water
along the channel from south to north leads to an increase in the river stage of
up to almost 8 m at the upstream gauge (Figure 5.7a). The upstream relationship
between stage and discharge is roughly linear for the full range of bankfull flows (up
to ~25 GL/d). For the 2007-2012 period analysed, <1% of flows are overbank at
this section, while none are overbank at the downstream gauge as a result of the
increased bankfull capacity downstream (> 50 GL/d; i.e., water re-enters channel
along the reach). The low-frequency of overbank flows, however, does not diminish
their significance for the catchment water balance. This is illustrated in Figure 5.7b,
where overbank events generate the largest recharge volumes (between 10 to 80 GL
per event) of all the examined flow events. This is intuitive as a much larger part
of the landscape (i.e., surface area) is included in the recharge process for overbank
flows. In contrast, the largest within-channel event has a recharge volume of only
6.2 GL in the period analysed (Figure 5.7b). From this it can be concluded that,
although infrequent, overbank recharge can be a significant source of water to the
underlying aquifer for this section of the Namoi River.
When focusing on the within-channel flows (Figure 5.8a) the relationship between
99
the groundwater recharge and the river stage is surprisingly weak. Thus, while in
general terms an increase in stage leads to increased recharge as shown by the
temperature data (Figures 5.4 and 5.5), there must be additional factors controlling
the recharge. Further scrutiny of Figure 5.8a shows that the data fall into two
populations: one with low river stages (i.e., 1.2 to 1.8 m) but with a large range
in recharge volumes (0.2 to 6.2 GL) (the triangles in Figure 5.8); and another with
larger river stages (i.e., 2.4 to 4.4 m) but without proportionally larger recharge
volumes (3 to 4.1 GL) (the circles in Figure 5.8a). As some events with smaller stages
clearly lead to greater recharge volumes, it might be supposed that these events may
have a longer duration, allowing more time for recharge to occur. However, this is
found not to be the case, since low-stage high-recharge events (triangles) have the
same duration as the high-stage events with moderate recharge (circles) (Figure
5.8b). Therefore, the efficiency of water loss (recharge) is clearly different as events
proceed beyond the 2m stage range, suggesting other factors may be responsible.
These two populations based on stage range are also apparent when recharge
is considered as a function of total river flow (Figure 5.8c). The ratio of the river
loss (i.e., recharge) to cumulative flow (i.e., total river flow) is indicative of recharge
efficiency (Equation 8), with the slope of the trends representing the overall recharge
efficiency for these two populations. For the smaller events (i.e., the triangles, which
have stages < ~2 m) this efficiency is ~25%, much higher than the 6% efficiency
for the larger events (i.e., the circles, which have stages > ~2 m) (see efficiency
results for individual events in Table 5.2). One plausible explanation for this is that
as the river stage increases, there is proportionally more water flowing along the
river due to increased hydraulic efficiency, than is lost to the aquifer as recharge.
However, this does not explain the data presented here as seen from Figure 5.7a as
the relationship between river stage and river flow is approximately linear for river
flows below about 25 GL/d. Another possible explanation is that the riverbed has
a higher permeability than the riverbanks. Thus, the smaller flows predominantly
100
recharge via the higher permeability riverbed, whereas the larger flows also have
recharge via the lower permeability riverbanks. This explanation is backed by field
observations that the upper banks are generally of clayey or loamy texture along
the river reach.
Based on some field observations which have been made in the literature, the
opposite case is often assumed. That is that the channel bottom sediments have
low permeability due to the existence of a clogging layer [e.g, Brunner et al., 2009a].
For instance, Lange [2005] found that small and medium flows led to low recharge
volumes while large flows led to high recharge volumes. He speculated that the
low recharge volumes were due to the presence of clogging layers on and within
the channel alluvium, whereas the high recharge volumes were due to enhanced
water infiltration in overbank areas. Contrary to this, a field-based study by Jolly
et al. [1994] actually concluded that diffuse vertical recharge of floodwaters was
an insignificant recharge pathway for aquifers the MDB. This is also supported by
Doble et al. [2012], who in a numerical study of overbank recharge, concluded that
the permeability of a thin veneer of sediments across the floodplain was of critical
importance in controlling the amount of recharge. The present study does suggest
that large magnitude flow events may well lead to greater recharge volumes (see
Figure 5.7b), but it also suggest that the hydraulic conductivity of the channel is
much higher than that in upper banks and possibly the floodplain and that the
riverbed in this case is a more direct route of recharge than are the riverbanks.
Thus the present study shows that observations on the permeability distribution
cannot be generalised and have to be assessed individually for each river system and
possibly even for individual river reaches.
101
5.6.3 Role of riverbed sediment heterogeneity and hypo-
rheic flow
The discussion above draws conclusions about processes on the catchment scale
and indicate some general observable effects caused by large scale geomorphological
heterogeneity, however, considerable heterogeneity is to be expected on smaller scales
too [Buffington and Tonina, 2009; Fleckenstein et al., 2006; Genereux et al., 2008].
From the temperature data significant spatial variability can be observed in the
estimated Darcy velocities, both within and between the pool and reach data (see
values of µ and δ in Table 5.1). Darcy velocities are similar during low-flow periods
and become increasingly divergent during higher flow conditions. This divergence in
velocity is further amplified as the distance between arrays increases: i.e., the widely
spaced reach measurements show much greater variation in peak Darcy velocities
that those within the pool.
This variability in Darcy velocity for identical river stage fluctuations can be
explained in two ways. First, a variable distribution in riverbed hydraulic conduct-
ivity due to the natural variability in riverbed sediments will lead to variation in
flux spatially. This heterogeneity in conductivity is a well-documented phenomenon
of riverbeds [e.g., Storey et al., 2003; Cardenas et al., 2004; Buffington and Tonina,
2009]. Second, departures from the 1D flow field. Topographic features (e.g., ripples,
bars, meanders) influence Darcy flow fields, drive hyporheic exchange, and thus
have an effect on the subsurface heat flow [Cardenas and Wilson, 2007]. Signific-
ant hyporheic exchange would therefore affect the estimates of the Darcy velocity
depending on the probe locations within the actual 3D flow field. Recently, Hyun
et al. [2011] found that point and area-averaged estimates of flux differed from each
other and hypothesised that two processes were occurring simultaneously: hyporheic
exchange superimposed onto regional groundwater discharge into the river. In a
separate study, Ward et al. [2012] found that local in-river hydraulic gradients
did not necessarily reflect the regional gradients. Thus, the cause of this spatial
102
heterogeneity in fluxes for low-flow conditions cannot be known with certainty based
on the temperature data alone. Under strongly losing or gaining conditions, however,
hyporheic exchange is expected to have a diminished effect on the more regional flow
field [Cardenas, 2009]. Hyporheic velocities are therefore more likely to contribute
greater uncertainty to the flux estimates under low-flow conditions than during high-
flows. On the other hand, during high-flows the observed variability can be largely
attributed to the effects of riverbed heterogeneity.
Assuming that the variability in Darcy velocities observed in the pool scale is
representative for other pools along the reach, this uncertainty can be propagated
to all reach results under low-flow conditions. This allows the estimation of 10th
and 90th percentiles (i.e., µ +/- 1.24 δ) for the Darcy velocities, thereby giving
a realistic range to the reach results, without minimising the complications of the
heterogeneity and hyporheic exchange. This approach provides some guidance as to
whether the estimated velocities at each reach location are actually losing or gaining,
or are a result of the temperature probe being located in a local (hyporheic) zone
of up- or down-welling.
When the uncertainty is added to the results, the point measurement velocities
from the reach deployment during low-flow conditions are comparable to a reach
averaged low-flow estimate from the differential gauging. The Darcy velocities
derived from the temperature time series along the reach are neutral to slightly
losing, with average 10th and 90th percentile values of -0.10 and -0.04 m/d. Taking
the river loss during low-flows for the analysed period, and by using an estimated
reach length of 34,000 m and width of 20 m, the reach-scale five-year average river-
aquifer Darcy velocity was estimated at -0.07 m/d. This compares favourably,
though not exactly, with the results from the temperature time series. From this it
can be tentatively concluded that the reach is slightly losing overall during low-flow
periods.
During high-flow events the divergence between point measurements can be large
103
(-0.5 m/d to -1.5 m/d). This suggests differing degrees of river-aquifer connectiv-
ity along the reach. Other researchers have found that recharge through higher
conductivity “windows” in semi-arid and arid riverbeds is a significant recharge
pathway (e.g., Shentsis and Rosenthal, 2003). These exchanges must be considered
when accounting for the recharge volumes.
In addition to the spatial variability in riverbed sediments and hydraulic conduct-
ivity there is also some evidence to suggest that these vary over time which would also
impact the interactions between river and aquifer. During high-flows there was direct
evidence of bed sediment scouring and deposition occurring. In the case of array
R3, the bed scour during the event was sufficient to remove the riverbed installation,
while in the case of array R4, the same event caused deposition of sediment leaving
the array buried within the riverbed to such a depth that the temperature sensors
were no longer sensitive to diel variations. The effect of this sediment transport
on hydraulic conductivity over time is difficult to quantify directly, however there
are a small but growing number of field studies which have documented temporal
variability in riverbed hydraulic conductivity [e.g., Springer et al., 1999; Genereux
et al., 2008; Rau et al., 2010; Mutiti and Levy, 2010; Hatch et al., 2010].
5.6.4 Implications for numerical modelling and resource
management
Based on surface water hydrograph analysis, river aquifer gradients and modelling
of the catchment water balance, previous studies [i.e., Giambastiani et al., 2012;
McCallum et al., 2013b] suggested that the Namoi River should have become overall
losing in recent years due to groundwater abstraction. The present results support
the suggestion that the river is indeed losing at low flows by the direct measurements
of the Darcy flow in the riverbed. To our knowledge this is the first time this has been
achieved for prolonged periods (beyond the duration of a typical field campaign).
The large variability in riverbed velocities observed at different locations, and
104
over time, raises questions as to the validity of inferring large scale processes from
point measurements. Whilst this variation is small during low-flow conditions,
even within 10s of meters within the pool the results spanned gaining, neutral
and losing conditions. Based on this finding, it would not be justifiable for the
purpose of catchment scale water exchange calculations to classify the river-aquifer
interaction based on the temperature time series from a single array. However,
provided variation at small spatial distances is first acknowledged and sufficiently
understood, it may be possible to link point measurements of river-acquirer exchange
to reach estimates for the low-flow conditions as done for this study. On the other
hand, during high-flow events the majority of the exchange by volume appears to
occur through higher connectivity “windows”, and so up scaling point measurements
without a sufficient understanding of these processes and without a significant spatial
coverage of the reach will lead to a potentially erroneous estimation of the water
balance.
Finally, generalisations such as assuming a relatively lower permeability in the
riverbed (due to clogging layers) relative to the riverbanks should be avoided and
replaced by concepts based on field observations and incorporated into numerical
models accordingly.
5.7 Conclusions
Factors affecting river-aquifer interactions along a reach of the Namoi River in

semi-arid eastern Australia were investigated using temperature derived riverbed
Darcy velocities and recharge losses from differential river gauging. The results
show slightly gaining to slightly losing conditions during low-flows with a temporal
shift, to strongly losing conditions (i.e., increasing riverbed Darcy velocities) driven
primarily by increases in the river stage. The water balance of the reach, determined
by differential river gauging, reveals that the increase in riverbed flux during high-
flows accounts for 69% of the total river derived groundwater recharge despite only
105
occurring 20% of the time.

A second factor that may influence variations in the riverbed fluxes examined
here is hydraulic conductivity. Although variations in river stage can account for
most of the velocity variations, large differences in peak losses were observed during
an event between the arrays. This variability was found within the pool data, and
particularly within the reach data. It is suggested that this spatial variation in
Darcy riverbed velocity magnitude is driven by variations in hydraulic conductivity
throughout the reach.
Using the differential river gauging data the relationships between the total flow
volume, river stage, duration and recharge were examined. It was found that not
only are the total flow volume, river stage and duration important in determining
the volume of recharge but the channel geomorphology also plays a significant role.
This study has implications for the conceptual understanding of river-aquifer
interactions. Given the large spatial and temporal variations observed in riverbed
fluxes for both low and high-flow conditions, it is reasonable to question any up-
scaling of point measurements to reach estimates, especially during high-flows, which
are not based upon a wider understanding of the catchment. The need to use field
observations to drive conceptual generalisations made in numerical models was also
seen.
106
Chapter 6 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Conclusion
To conclude the thesis, the significance of the research for the modelling and man-
agement of water resources is discussed. This was previously touched upon in the
Introduction and Chapters 2 and 5, but here some further links are pointed out
between the research undertaken and the difficult day-to-day realities which confront
those who seek to manage our rivers and aquifers as a single resource.
6.1 Modelling
Determination of water volumes and fluxes in semi-arid environments is fraught with

difficulty [e.g. de Vries and Simmers, 2002]. Consequently, groundwater models are
often used to synthesise all the available data. Numerical models have been used
in variety of settings, from investigating the spatial-temporal variation in river-
aquifer interaction in an urban environment [e.g. Martinez-Santos et al., 2010],
to determining the influence of groundwater abstraction on river-aquifer interac-
tion at the catchment-scale [e.g. Werner et al., 2006; Sanz et al., 2011]. As the
river losses/gains are closely related to the groundwater levels in connected reaches
[Braaten and Gates, 2003], the link between rivers and aquifers, for both losing
and gaining conditions, is typically included in numerical models via a “riverbed
conductance” term [Rushton, 2007].
This thesis has implications for the use of a simple conductance term. The
107
Chapter 6. Conclusion
surmised changes in riverbed conductivity and consequently in the exchange flux

due to flood events is an important finding. Furthermore, the recharge “windows”
observed, and the flow rate threshold in recharge efficiency, both show that current
models with a conceptual understanding that riverbed hydraulic conductivity is
constant over time and space are inadequate. This research demonstrates that
using a conductance term is too simplistic and does not represent the physical
reality, particularly in semi-arid environments which are characterised by episodic
flow events. Consequently, new conceptual methods need to be developed [Doppler
et al., 2007].
Extending this insight to the trendy area of climate change modelling, it is clear
that the correct conceptualisation and implementation of river-aquifer interactions
will be particularly important. At present, even though groundwater is the main
water resource in many parts of the world, most hydrological studies considering
climate change simplify or neglect it [Scibek et al., 2008; Goderniaux et al., 2009].
Impacts from changes in climate can be of the same order of magnitude as those due
to human factors which complicates the analysis of river-aquifer interactions (e.g.,
river regulation, groundwater abstraction) [Mayer and Congdon, 2008]. Given the
difficulties inherent in these modelling exercises, the danger of simplifying assump-
tions in the hydrological system ought to be highlighted, otherwise discrepancies in
future projections will occur. As has been the case in different contexts over the
last century, it may well turn out that the “missing link” in these models is the
seemingly easily forgotten interactions between rivers and aquifers.
6.2 Management
There exists in the hydrological community the idea that sustainable pumping must
not exceed natural recharge [Bredehoeft, 2002; Sophocleous, 2004]. This comes
from the apparently impeccable logic that if an aquifer is pumped more than it
is recharged it will eventually run out of water (e.g., see how this idea is assumed in
108
Chapter 6. Conclusion
de Vries and Simmers [2002], Lange [2005], and Wada et al. [2012]). Historically, this
has meant, water resources have been managed using a policy termed “safe yield”.
This approach allows water users to pump no more groundwater than is naturally
replenished by recharge [Sophocleous, 1997]. However, this policy has been used to
the detriment of semi-arid environments worldwide.
In an undeveloped groundwater system, the long term average natural recharge
is equal to the average natural discharge and the system is in equilibrium. With
the introduction of groundwater abstraction the system will tend towards a new
equilibrium. In the transition to this state, abstraction must be balanced by an
increase in recharge, a decrease in discharge, a loss of storage, or some combination
of these [Theis, 1940]. Significantly, a predevelopment groundwater budget does
not provide information on where the abstracted groundwater will be sourced from
[Alley et al., 1999].
However, the yield of an aquifer must be considerably lower than the natural
recharge if adequate amounts of water are to be available to sustain both the quantity
and quality of surface waters [Sophocleous, 2000]. Consequently, alternatives to the
safe yield policy have been developed. Some authors now argue that safe yield should
be related to the amount of captured discharge rather than recharge estimates [Alley
and Leake, 2004].
Significantly, the “missing link” in the safe yield policy was that it had ignored
the interactions between rivers and aquifers [Alley et al., 1999]. In order to inform
water resource management, studies are requried which seek to understand and
quantify the interactions between rivers and aquifers when stressed by groundwater
abstraction. This thesis is one small contribution towards that end.
109
Bibliography
Abbott, M. (1969), Petrology of the nandewar volcano, nsw, australia, Contribution

to Mineralogy and Petrology, 20, 115–134.
ABS (2012), 4628.0.55.001 - completing the picture - environmental accounting in

practice, Tech. rep., Australian Bureau of Statistics.
Acworth, R. I. (2009), Surface water and groundwater: understanding the

importance of their connections introduction, Australian Journal of Earth
Sciences, 56 (1), 1–2.
Alley, M., and S. A. Leake (2004), The journey from safe yield to sustainability,
Ground Water, 42 (1), 12–16.
Alley, W. M., T. E. Reilly, and O. L. Frank (1999), Sustainability of ground-water

resources, U.S. Geological Survey Circular 1186.
Andersen, M. S., and I. R. Acworth (2009), Stream aquifer interactions in the maules
creek catchment, namoi valley, nsw, australia, Hydrogeology Journal, 17, 2005–
2021, doi: 10.1007/s10040-009-0500-9.
Anderson, M. P. (2005), Heat as a ground water tracer, Ground Water, 43 (6), 951–
968, doi: 10.1111/j.1745-6584.2005.00052.x.
Anderson, M. P., and W. W. Woessner (1992), Applied groundwater modeling:

simulation of flow and advective transport.
Angermann, L., J. Lewandowski, J. H. Fleckenstein, and G. Nützmann (in press),

A 3d analysis algorithm to improve interpretation of heat pulse sensor results for
the determination of small-scale flow directions and velocities in the hyporheic
zone, Journal of Hydrology, (0), doi: 10.1016/j.jhydrol.2012.06.050.
Banks, E. W., C. T. Simmons, A. J. Love, and P. Shand (2011), Assessing spatial

and temporal connectivity between surface water and groundwater in a regional
catchment: Implications for regional scale water quantity and quality, Journal of
Hydrology, 404 (1-2), 30–49, doi: 10.1016/j.jhydrol.2011.04.017.
Barlow, P. M., L. A. DeSimone, and A. F. Moench (2000), Aquifer response to

stream-stage and recharge variations. ii. convolution method and applications,
Journal of Hydrology, 230 (3-4), 211–229.
111
Bibliography
Becker, M. W., T. Georgian, H. Ambrose, J. Siniscalchi, and K. Fredrick

(2004), Estimating flow and flux of ground water discharge using water
temperature and velocity, Journal of Hydrology, 296 (1-4), 221–233, doi:
10.1016/j.jhyrol.2004.03.025.
Bencala, K. E. (1993), A perspective on stream-catchment connections, Journal of

the North American Benthological Society, 12 (1), 44–47, doi: 10.2307/1467684.
Biemans, H., I. Haddeland, P. Kabat, F. Ludwig, R. W. A. Hutjes, J. Heinke,

W. von Bloh, and D. Gerten (2011), Impact of reservoirs on river discharge and
irrigation water supply during the 20th century, Water Resources Research, 47,
doi: 10.1029/2009wr008929.
BOM (2009), Precipitation and evapotranspiration databases, Bureau of Meteoro-

logy, NSW.
Braaten, R., and G. Gates (2003), Groundwater-surface water interaction in inland

new south wales: a scoping study, Water Science and Technology, 48 (7), 215–224.
Bredehoeft, J. D. (2002), The water budget myth revisited: Why hydrogeologists

model, Ground Water, 40 (4), 340–345.
Briggs, M. A., L. K. Lautz, J. M. McKenzie, R. P. Gordon, and D. K. Hare (2012),

Using high-resolution distributed temperature sensing to quantify spatial and
temporal variability in vertical hyporheic flux, Water Resources Research, 48 (2),
W02,527, doi: 10.1029/2011WR011227.
Brunke, M., and T. Gonser (1997), The ecological significance of exchange

processes between rivers and groundwater, Freshwater Biology, 37 (1), 1–33, doi:
10.1046/j.1365-2427.1997.00143.x.
Brunner, P., P. G. Cook, and C. T. Simmons (2009a), Hydrogeologic controls on

disconnection between surface water and groundwater, Water Resources Research,
45.
Brunner, P., C. T. Simmons, and P. G. Cook (2009b), Spatial and temporal aspects
of the transition from connection to disconnection between rivers, lakes and
groundwater, Journal of Hydrology, 376 (1-2), 159–169.
Brunner, P., P. G. Cook, and C. T. Simmons (2011), Disconnected surface water

and groundwater: From theory to practice, Ground Water, 49 (4), 460–467, doi:
10.1111/j.1745-6584.2010.00752.x.
Buffington, J. M., and D. Tonina (2009), Hyporheic exchange in mountain rivers

ii: Effects of channel morphology on mechanics, scales, and rates of exchange,
Geography Compass, 3 (3), 1038–1062, doi: 10.1111/j.1749-8198.2009.00225.x.
Buntebarth, G., and J. R. Schopper (1998), Experimental and theoretical

investigations on the influence of fluids, solids and interactions between them on
thermal properties of porous rocks, Physics and Chemistry of The Earth, 23 (9-
10), 1141–1146, doi: 10.1016/S0079-1946(98)00142-6.
112
Bibliography
Burrell, M., P. Moss, D. Green, A. Ali, and J. Petrovic (2011), General purpose
water accounting report 2009–2010: Namoi catchment, Tech. rep., NSW Office of
Water, Sydney.
Cardenas, M. B. (2008), The effect of river bend morphology on flow and timescales
of surface water-groundwater exchange across pointbars, Journal of Hydrology,
362 (1-2), 134–141.
Cardenas, M. B. (2009), Stream-aquifer interactions and hyporheic exchange in

gaining and losing sinuous streams, Water Resources Research, 45.
Cardenas, M. B., and J. L. Wilson (2007), Effects of current-bed form induced

fluid flow on the thermal regime of sediments, Water Resources Research, 43 (8),
W08,431, doi: 10.1029/2006WR005343.
Cardenas, M. B., J. L. Wilson, and V. A. Zlotnik (2004), Impact of heterogeneity,

bed forms, and stream curvature on subchannel hyporheic exchange, Water
Resources Research, 40 (8).
Cey, E. E., D. L. Rudolph, G. W. Parkin, and R. Aravena (1998), Quantifying

groundwater discharge to a small perennial stream in southern ontario, canada,
Journal of Hydrology, 210 (1-4), 21–37, doi: 10.1016/s0022-1694(98)00172-3.
Chen, X. H. (2003), Analysis of pumping-induced stream-aquifer interactions for

gaining streams, Journal of Hydrology, 275 (1-2), 1–11.
Constantz, J. (2008), Heat as a tracer to determine streambed water exchanges,

Water Resources Research, 44, doi: 10.1029/2008wr006996.
Constantz, J., S. W. Tyler, and E. Kwicklis (2003), Temperature-profile methods

for estimating percolation rates in arid environments, Vadose Zone Journal, 2 (1),
12–24, doi: 10.2113/2.1.12.
Cook, P. G. (2012), Estimating groundwater discharge to rivers from river chemistry

surveys, Hydrological Processes.
Dahan, O., Y. Shani, Y. Enzel, Y. Yechieli, and A. Yakirevich (2007), Direct

measurements of floodwater infiltration into shallow alluvial aquifers, Journal of
Hydrology, 344 (3-4), 157–170, doi: 10.1016/j.jhydrol.2007.06.033.
De Marsily, G. (1986), Quantitative hydrogeology: groundwater hydrology for

engineers, xix, 440 p. pp., Academic Press, Orlando, FL.
de Vries, J. J., and I. Simmers (2002), Groundwater recharge: an overview of

processes and challenges, Hydrogeology Journal, 10 (1), 5–17.
Desilets, S. L. E., T. P. A. Ferre, and P. A. Troch (2008), Effects of stream-aquifer

disconnection on local flow patterns, Water Resources Research, 44 (9).
DMR (1998), Gunnedah coalfield north regional geology (1:100,000 map), Geological
survey of NSW, Department of Mineral Resources, Sydney.
113
Bibliography
DNR (2009), Surface water hydrograph, groundwater hydrograph, and groundwater

abstraction databases, Department of Natural Resources, NSW.
DNR (2011), Surface water hydrograph database, Department of Natural Resources,
NSW.
Doble, R. C., R. S. Crosbie, B. D. Smerdon, L. Peeters, and F. J. Cook (2012),
Groundwater recharge from overbank floods, Water Resources Research, 48 (9).
Domenico, P., and F. Schwartz (1990), Physical and chemical hydrogeology, vol. 824,
Wiley New York.
Doppler, T., H. J. H. Franssen, H. P. Kaiser, U. Kuhlman, and F. Stauffer
(2007), Field evidence of a dynamic leakage coefficient for modelling
river-aquifer interactions, Journal of Hydrology, 347 (1-2), 177–187, doi:
10.1016/j.jhydrot.2007.09.017.
Doussan, C., G. Poitevin, E. Ledoux, and M. Detay (1997), River bank filtration:
Modelling of the changes in water chemistry with emphasis on nitrogen species,
Journal of Contaminant Hydrology, 25 (1-2), 129–156.
Fanelli, R. M., and L. K. Lautz (2008), Patterns of water, heat, and solute flux
through streambeds around small dams, Ground Water, 46 (5), 671–687, doi:
10.1111/j.1745-6584.2008.00461.x.
Fleckenstein, J. H., and G. E. Fogg (2008), Efficient upscaling of hydraulic
conductivity in heterogeneous alluvial aquifers, Hydrogeology Journal, 16 (7),
1239–1250.
Fleckenstein, J. H., R. G. Niswonger, and G. E. Fogg (2006), River-aquifer
interactions, geologic heterogeneity, and low-flow management, Ground Water,
44 (6), 837–852.
Fleckenstein, J. H., S. Krause, D. M. Hannah, and F. Boano (2010), Groundwater-
surface water interactions: New methods and models to improve understanding
of processes and dynamics, Advances in Water Resources, 33 (11), 1291–1295, doi:
10.1016/j.advwatres.2010.09.011.
Fox, G. A. (2004), Evaluation of a stream aquifer analysis test using analytical
solutions and field data, Journal of the American Water Resources Association,
40 (3), 755–763.
Fox, G. A., and D. S. Durnford (2003), Unsaturated hyporheic zone flow in
stream/aquifer conjunctive systems, Advances in Water Resources, 26 (9), 989–
1000.
Fullagar, I., R. Brodie, B. Sundaram, S. Hostetler, and P. Baker (2006), Managing
connected surface water and groundwater resources, Tech. rep., Bureau of Rural
Sciences, Canberra, http://www.brs.gov.au.
Gates, G. (1980), The hydrology of the unconsolidated sediments in the mooki river
valley, new south wales, MSc Thesis, University of New South Wales, Sydney.
114
Bibliography
Gates, G. W. B., and J. B. Ross (1980), Cainozoic alleviation and hydrogeology of

the namoi river basin, nsw, In: The Cainozoic Evolution of Continental Southeast
Australia, BMR record 1980/67.
Genereux, D. P., S. Leahy, H. Mitasova, C. D. Kennedy, and D. R. Corbett (2008),

Spatial and temporal variability of streambed hydraulic conductivity in west bear
creek, north carolina, usa, Journal of Hydrology, 358 (3-4), 332–353.
Giambastiani, B. M. S., A. M. McCallum, M. S. Andersen, B. F. J. Kelly, and R. I.

Acworth (2012), Understanding groundwater processes by representing aquifer
heterogeneity in the maules creek catchment, namoi valley (new south wales,
australia), Hydrogeol J.
Gilbert, N. (2012), Water under pressure, Nature, 483, 256–257.
Glover, R., and C. Balmer (1954), River depletion from pumping a well near a river,
American Geophysical Union Transactions, 35(3), 468–470.
Goderniaux, P., S. Brouyere, H. J. Fowler, S. Blenkinsop, R. Therrien, P. Orban, and

A. Dassargues (2009), Large scale surface-subsurface hydrological model to assess
climate change impacts on groundwater reserves, Journal of Hydrology, 373 (1-2),
122–138.
Gordon, R. P., L. K. Lautz, M. A. Briggs, and J. M. McKenzie (2012), Automated

calculation of vertical pore-water flux from field temperature time series using the
vflux method and computer program, Journal of Hydrology, 420–421 (0), 142 –
158, doi: 10.1016/j.jhydrol.2011.11.053.
Goto, S., M. Yamano, and M. Kinoshita (2005), Thermal response of sediment with
vertical fluid flow to periodic temperature variation at the surface, Journal of
Geophysical Research-Solid Earth, 110 (B1), doi: 10.1029/2004jb003419.
Halford, K. J., and G. C. Mayer (2000), Problems associated with estimating ground
water discharge and recharge from stream-discharge records, Ground Water,
38 (3), 331–342.
Hancock, P. J. (2002), Human impacts on the stream-groundwater exchange zone,

Environmental Management, 29 (6), 763–781.
Hantush, M. (1965), Wells near streams with semipervious beds, Journal of

Geophysical Research, 70(12), 2829–2838.
Harte, P. T., and R. G. Kiah (2009), Measured river leakages using conventional
streamflow techniques: the case of souhegan river, new hampshire, usa,
Hydrogeology Journal, 17 (2), 409–424.
Hatch, C. E., A. T. Fisher, J. S. Revenaugh, J. Constantz, and C. Ruehl (2006),

Quantifying surface water-groundwater interactions using time series analysis of
streambed thermal records: Method development, Water Resources Research,
42 (10).
115
Bibliography
Hatch, C. E., A. T. Fisher, C. R. Ruehl, and G. Stemler (2010), Spatial

and temporal variations in streambed hydraulic conductivity quantified with
time-series thermal methods, Journal of Hydrology, 389 (3-4), 276–288, doi:
10.1016/j.jhydrol.2010.05.046.
Hoppe, H., and G. Kiely (1999), Precipitation over ireland - observed change
since 1940, Physics and Chemistry of the Earth Part B-Hydrology Oceans and
Atmosphere, 24 (1-2), 91–96.
Humphreys, W. F. (2009), Hydrogeology and groundwater ecology: Does each
inform the other?, Hydrogeology Journal, 17 (1), 5–21.
Hunt, B. (1999), Unsteady stream depletion from ground water pumping, Ground
Water, 37 (1), 98–102.
Hyun, Y., H. Kim, S.-S. Lee, and K.-K. Lee (2011), Characterizing streambed
water fluxes using temperature and head data on multiple spatial scales in
munsan stream, south korea, Journal of Hydrology, 402 (3â€“4), 377–387, doi:
http://dx.doi.org/10.1016/j.jhydrol.2011.03.032.
Ivkovic, K. M., R. A. Letcher, and B. F. W. Croke (2009), Use of a simple surface-
groundwater interaction model to inform water management, Australian Journal
of Earth Sciences, 56 (1), 71–80.
Jensen, J. K., and P. Engesgaard (2011), Nonuniform groundwater discharge across
a streambed: Heat as a tracer, Vadose Zone Journal, 10 (1), 98–109, doi:
10.2136/vzj2010.0005.
Johansen, O. (1975), Thermal conductivity of soils, Ph.D. thesis, University of
Trondheim.
Jolly, I. D., G. R. Walker, and K. A. Narayan (1994), Floodwater recharge processes
in the chowilla anabranch system, south australia, Soil Research, 32 (3), 417–435.
Jones, J. B., and P. J. Mulholland (2000), Streams and Groundwater, Academic,
San Diego, Calif.
Kalbus, E., F. Reinstorf, and M. Schirmer (2006), Measuring methods for
groundwater - surface water interactions: a review, Hydrology and Earth System
Sciences, 10 (6), 873–887, doi: 10.5194/hess-10-873-2006.
Kalbus, E., C. Schmidt, J. W. Molson, F. Reinstorf, and M. Schirmer
(2009), Influence of aquifer and streambed heterogeneity on the distribution of
groundwater discharge, Hydrology and Earth System Sciences, 13 (1), 69–77.
Kaleris, V. (1998), Quantifying the exchange rate between groundwater and small
streams, Journal of Hydraulic Research, 36 (6), 913–932.
Keery, J., A. Binley, N. Crook, and J. W. N. Smith (2007), Temporal and spatial
variability of groundwater-surface water fluxes: Development and application of
an analytical method using temperature time series, Journal of Hydrology, 336 (1-
2), 1–16, doi: 10.1016/j.jhydrol.2006.12.003.
116
Bibliography
Kennedy, C. D., D. P. Genereux, H. Mitasova, D. R. Corbett, and S. Leahy (2008),

Effect of sampling density and design on estimation of streambed attributes,
Journal of Hydrology, 355 (1), 164–180.
Kollet, S. J., and V. A. Zlotnik (2003), Stream depletion predictions using pumping
test data from a heterogeneous stream-aquifer system (a case study from the great
plains, usa), Journal of Hydrology, 281 (1-2), 96–114.
Lambs, L. (2004), Interactions between groundwater and surface water at river

banks and the confluence of rivers, Journal of Hydrology, 288 (3-4), 312–326, doi:
10.1016/j.jhydrol.2003.10.013, times Cited: 23 Lambs, L.
Lange, J. (2005), Dynamics of transmission losses in a large and stream channel,

Langhoff, J. H., K. R. Rasmussen, and S. Christensen (2006), Quantification and

regionalization of groundwater-surface water interaction along an alluvial stream,
Lapham, W. W. (1989), Use of temperature profiles beneath streams to determine

rates of vertical ground-water flow and vertical hydraulic conductivity, Water-
Supply Paper 2337, U.S. Geological Survey.
Lautz, L. K., N. T. Kranes, and D. I. Siegel (2010), Heat tracing of heterogeneous

hyporheic exchange adjacent to in-stream geomorphic features, Hydrological
Processes, 24 (21), 3074–3086, doi: 10.1002/hyp.7723.
Lerner, D. N., A. S. Issar, and I. Simmers (1990), Groundwater recharge - A guide

to understanding and estimating natural recharge, International Association of
Hydrogeologists, Kenilworth.
Mair, A., and A. Fares (2010), Influence of groundwater pumping and rainfall spatio-
temporal variation on streamflow, Journal of Hydrology, 393 (3-4), 287–308.
Martin, H. (1994), The stratigraphic palynology of the namoi river valley, baan baa
to boggabri, northern new south wales, Proceedings of the Linnean Society, New
South Wales, 114(1), 45–58.
Martinez-Santos, P., P. E. Martinez-Alfaro, E. Sanz, and A. Galindo (2010), Daily

scale modelling of aquifer-river connectivity in the urban alluvial aquifer in
langreo, spain, Hydrogeology Journal, 18 (6), 1525–1537.
Massmann, G., A. Pekdeger, and C. Merz (2004), Redox processes in the oderbruch
polder groundwater flow system in germany, Applied Geochemistry, 19 (6), 863–
886.
Mayer, T. D., and R. D. Congdon (2008), Evaluating climate variability and

pumping effects in statistical analyses, Ground Water, 46 (2), 212–227.
117
Bibliography
McCallum, A. M., M. S. Andersen, G. C. Rau, and R. I. Acworth (2012a), A

1â€?d analytical method for estimating surface waterâ€“groundwater interactions
and effective thermal diffusivity using temperature time series, Water Resources
Research, 48 (11).
McCallum, A. M., M. S. Andersen, and R. I. Acworth (2013a), A new method

for estimating recharge to unconfined aquifers using differential river gauging,
Groundwater.
McCallum, A. M., M. S. Andersen, B. M. S. Giambastiani, B. F. J. Kelly, and

R. Ian Acworth (2013b), Riverâ€“aquifer interactions in a semi-arid environment
stressed by groundwater abstraction, Hydrological Processes, 27 (7), 1072–1085,
doi: 10.1002/hyp.9229.
McCallum, J. L., P. G. Cook, D. Berhane, C. Rumpf, and G. A. McMahon

(2012b), Quantifying groundwater flows to streams using differential flow
gaugings and water chemistry, Journal of Hydrology, 416, 118–132, doi:
10.1016/j.jhydrol.2011.11.040.
MDBA (2010), Guide to the proposed basin plan: Technical background, Tech. rep.,
Murray–Darling Basin Authority, Canberra.
Munz, M., S. E. Oswald, and C. Schmidt (2011), Sand box experiments to evaluate
the influence of subsurface temperature probe design on temperature based
water flux calculation, Hydrol. Earth Syst. Sci. Discuss., 8 (3), 6155–6197, doi:
10.5194/hessd-8-6155-2011.
Mutiti, S., and J. Levy (2010), Using temperature modeling to investigate the
temporal variability of riverbed hydraulic conductivity during storm events,
Journal of Hydrology, 388 (3-4), 321 – 334, doi: 10.1016/j.jhydrol.2010.05.011.
Nemeth, M. S., and H. M. Solo-Gabriele (2003), Evaluation of the use of reach

transmissivity to quantify exchange between groundwater and surface water,
Journal of Hydrology, 274 (1-4), 145–159, doi: 10.1016/s0022-1694(02)00419-5.
Opsahl, S. P., S. E. Chapal, D. W. Hicks, and C. K. Wheeler (2007), Evaluation of

ground-water and surface-water exchanges using streamflow difference analyses,
Journal of the American Water Resources Association, 43 (5), 1132–1141.
Postma, D., F. Larsen, N. T. M. Hue, M. T. Duc, P. H. Viet, P. Q. Nhan, and

S. Jessen (2007), Arsenic in groundwater of the red river floodplain, vietnam:
Controlling geochemical processes and reactive transport modeling, Geochimica
Et Cosmochimica Acta, 71 (21), 5054–5071.
Rau, G. C., M. S. Andersen, A. M. McCallum, and R. I. Acworth (2010), Analytical

methods that use natural heat as a tracer to quantify surface water-groundwater
exchange, evaluated using field temperature records, Hydrogeology Journal, 18 (5),
1093–1110.
118
Bibliography
Rau, G. C., M. S. Andersen, and R. I. Acworth (2012), Experimental investigation

of the thermal dispersivity term and its significance in the heat transport equation
for flow in sediments, Water Resour. Res., 48, W03,511.
Roberts, O. R. F. M., J. (2006), Carboniferous to lower permian stratigraphy of

the southern tamworth belt, southern new england orogen, australia: Boundary
sequences of the werrie and rouchel blocks, Australian Journal of Earth Sciences,
53(2), 249–284.
Ruehl, C., A. T. Fisher, C. Hatch, M. Los Huertos, G. Stemler, and

C. Shennan (2006), Differential gauging and tracer tests resolve seepage fluxes
in a strongly-losing stream, Journal of Hydrology, 330 (1-2), 235–248, doi:
10.1016/j.jhydrol.2006.03.025.
Rushton, K. (1997), Recharge from permanent water bodies, in Simmers I (ed)

Recharge of phreatic aquifers in (semi)arid areas, AA Balkema, Rotterdam.
Rushton, K. (2007), Representation in regional models of saturated river-aquifer

interaction for gaining/losing rivers, Journal of Hydrology, 334 (1-2), 262–281.
Rushton, K. R. (2002), Will reductions in groundwater abstractions improve low

river flows?, Geological Society, London, Special Publications, 193 (1), 199–210,
doi: 10.1144/gsl.sp.2002.193.01.15.
Sanz, D., S. Castano, E. Cassiraga, A. Sahuquillo, J. J. Gomez-Alday, S. Pena,

and A. Calera (2011), Modeling aquifer-river interactions under the influence of
groundwater abstraction in the mancha oriental system (se spain), Hydrogeology
Journal, 19 (2), 475–487.
Scanlon, B. R., R. W. Healy, and P. G. Cook (2002), Choosing appropriate

techniques for quantifying groundwater recharge, Hydrogeology Journal, 10 (1),
18–39, doi: 10.1007/210040-00101762-2.
Scanlon, B. R., K. E. Keese, A. L. Flint, L. E. Flint, C. B. Gaye, W. M. Edmunds,

and I. Simmers (2006), Global synthesis of groundwater recharge in semiarid and
arid regions, Hydrological Processes, 20 (15), 3335–3370.
Schmadel, N. M., B. T. Neilson, and D. K. Stevens (2010), Approaches to estimate

uncertainty in longitudinal channel water balances, Journal of Hydrology, 394 (3-
4), 357–369, doi: 10.1016/j.jhydrol.2010.09.011.
Schmidt, C., M. Martienssen, and E. Kalbus (2011), Influence of water flux and
redox conditions on chlorobenzene concentrations in a contaminated streambed,
Hydrol. Process., 25(2), 234–245.
Schön, J. H. (1996), Physical Properties of Rocks: Fundamentals and Principles of

Petrophysics, 583 pp., Pergamon.
Scibek, J., D. M. Allen, and P. H. Whitfield (2008), Quantifying the impacts of

climate change on groundwater in an unconfined aquifer that is strongly influenced
by surface water, Geological Society Special Publication, vol. 288, pp. 79–98.
119
Bibliography
Shanafield, M., G. Pohll, and R. Susfalk (2010), Use of heat-based vertical fluxes
to approximate total flux in simple channels, Water Resources Research, 46,
W03,508, doi: 10.1029/2009wr007956.
Shanafield, M., C. Hatch, and G. Pohll (2011), Uncertainty in thermal time

series analysis estimates of streambed water flux, Water Resources Research, 47,
W03,504, doi: 10.1029/2010wr009574.
Shentsis, I., L. Meirovich, A. Benâ€?Zvi, and E. Rosenthal (1999), Assessment of

transmission losses and groundwater recharge from runoff events in a wadi under
shortage of data on lateral inflow, negev, israel, Hydrological Processes, 13 (11),
1649–1663.
Sinclair, P., C. Barrett, and W. R. M. (2005), Impact of groundwater extraction

on maules creek: Upper namoi valley, nsw, australia, in Acworth RI, Merrick N,
Macky G (eds) Where waters meet. Proceedings of the NZHS-IAH-NZSSS 2005
Conference, Auckland, 29 November–1 December.
Smakhtin, V. U. (2001), Low flow hydrology: a review, Journal of Hydrology, 240 (3-
4), 147–186.
Sophocleous, M. (1997), Managing water resources systems: Why ”safe yield” is not
sustainable, Ground Water, 35 (4), 561–561.
Sophocleous, M. (2000), From safe yield to sustainable development of water

resources - the kansas experience, Journal of Hydrology, 235 (1-2), 27–43.
Sophocleous, M. (2002), Interactions between groundwater and surface water: the

state of the science (vol 10, pg 52, 2002), Hydrogeology Journal, 10 (2), 348–348,
doi: 10.1007/s10040-002-0204-x.
Sophocleous, M. (2004), "the water budget myth revisited: Why hydrogeologists

model," by john d. bredehoeft, july-august 2002 issue, v. 40, no.4 : 340-345 -
discussion, Ground Water, 42 (4), 618–618.
Sophocleous, M., A. Koussis, J. L. Martin, and S. P. Perkins (1995), Evaluation of

simplified stream-aquifer depletion models for water rights administration, Ground
Water, 33 (4), 579–588.
Springer, A. E., W. D. Petroutson, and B. A. Semmens (1999), Spatial and temporal

variability of hydraulic conductivity in active reattachment bars of the colorado
river, grand canyon, Ground Water, 37 (3), 338–344.
Stallman, R. W. (1965), Steady 1-dimensional fluid flow in a semi-infinite porous

medium with sinusoidal surface temperature, Journal of Geophysical Research,
70 (12), 2821–&, doi: 10.1029/JZ070i012p02821.
Storey, R. G., K. W. F. Howard, and D. D. Williams (2003), Factors controlling riffle-

scale hyporheic exchange flows and their seasonal changes in a gaining stream: A
three-dimensional groundwater flow model, Water Resour. Res., 39 (2), 1034, doi:
10.1029/2002wr001367.
120
Bibliography
Su, G. W., J. Jasperse, D. Seymour, and J. Constantz (2004), Estimation of

hydraulic conductivity in an alluvial system using temperatures, Ground Water,
42 (6-7), 890–901.
Suzuki, S. (1960), Percolation measurements based on heat flow through soil with
special reference to paddy fields, Journal of Geophysical Research, 65 (9), 2883–
2885, doi: 10.1029/JZ065i009p02883.
Swanson, T. E., and M. B. Cardenas (2010), Diel heat transport within the hyporheic
zone of a pool-riffle-pool sequence of a losing stream and evaluation of models for
fluid flux estimation using heat, Limnology and Oceanography, 55 (4), 1741–1754,
doi: 10.4319/lo.2010.55.4.1741.
Tao, H., M. Gemmer, Y. G. Bai, B. D. Su, and W. Y. Mao (2011), Trends of

streamflow in the tarim river basin during the past 50 years: Human impact or
climate change?, Journal of Hydrology, 400 (1-2), 1–9.
Theis, C. V. (1940), The source of water derived from wells: Essential factors
controlling the response of an aquifer to develoment, Civil Engineer, 10, 277–280.
Theis, C. V. (1941), The effect of a well on the flow of a nearby stream, Transactions-
American Geophysical Union, 22, 734–738.
Toth, J. (1962), A theory of groundwater motion in small drainage basins in

central alberta, canada, Journal of Geophysical Research, 67 (11), 4375–&, doi:
10.1029/JZ067i011p04375.
Toth, J. (1963), A theoretical analysis of groundwater flow in small drainage basins,

Journal of Geophysical Research, 68 (16), 4795–&.
UNESCO (2009), World water assessment programme. 2009. the united nations
world water development report 3: Water in a changing world, Tech. rep.
USGS (2008), Field techniques for estimating water fluxes between surface water
and ground water, Tech. rep., Edited by Donald O. Rosenberry and James W.
LaBaugh, p 17.
Vogel, R. M., and N. M. Fennessey (1994), Flow-duration curves .2. new

interpretation and confidence-intervals, Journal of Water Resources Planning and
Management-Asce, 120 (4), 485–504.
Vogt, T., P. Schneider, L. Hahn-Woernle, and O. A. Cirpka (2010), Estimation

of seepage rates in a losing stream by means of fiber-optic high-resolution
vertical temperature profiling, Journal of Hydrology, 380 (1-2), 154–164, doi:
10.1016/j.jhydrol.2009.10.033.
Wada, Y., L. P. H. van Beek, and M. F. P. Bierkens (2012), Nonsustainable

groundwater sustaining irrigation: A global assessment, Water Resources
Research, 48, doi: 10.1029/2011wr010562.
121
Bibliography
Ward, A. S., M. Fitzgerald, M. N. Gooseff, T. J. Voltz, A. M. Binley, and K. Singha

(2012), Hydrologic and geomorphic controls on hyporheic exchange during base
flow recession in a headwater mountain stream, Water Resour. Res., 48 (4),
W04,513, doi: 10.1029/2011wr011461.
Werner, A. D., M. R. Gallagher, and S. W. Weeks (2006), Regional-scale, fully

coupled modelling of stream-aquifer interaction in a tropical catchment, Journal
of Hydrology, 328 (3-4), 497–510, doi: 10.1016/j.jhydrol.2005.12.034.
WFD (2008), Establishing a framework for community action in the field of water
policy., Directive 2008/105/ED of the European Parliment and of the Council.
Brussels: EU.
Williams, R. (1997), The cainozoic geology, hydrogeology and hydrochemistry of

the unconsolidated sediments associated with the namoi river in the lower namoi
valley, nsw, NSW Department of Land and Water Conservation, Sydney, NSW.
Winter, T. C. (1999), Relation of streams, lakes, and wetlands to groundwater flow

systems, Hydrogeology Journal, 7 (1), 28–45, doi: 10.1007/s100400050178.
Winter, T. C., J. W. Harvey, O. L. Franke, and W. M. Alley (1998), Ground water

and surface water - a single resource, Circular 1139, U.S. Geological Survey.
Woessner, W. W. (2000), Stream and fluvial plain ground water interactions:

Rescaling hydrogeologic thought, Ground Water, 38 (3), 423–429.
Woodside, W., and J. H. Messmer (1961), Thermal conductivity of porous media

.1. unconsolidated sands, Journal of Applied Physics, 32 (9), 1688–&, doi:
10.1063/1.1728419.
Xi, H. Y., Q. Feng, J. H. Si, Z. Q. Chang, and S. K. Cao (2010), Impacts of river
recharge on groundwater level and hydrochemistry in the lower reaches of heihe
river watershed, northwestern china, Hydrogeology Journal, 18 (3), 791–801.
Young, R. W., A. R. M. Young, D. M. Price, and R. A. L. Wray (2002),

Geomorphology of the namoi alluvial plain, northwestern new south wales,
Australian Journal of Earth Sciences, 49 (3), 509–523.
Zammouri, M., and H. Feki (2005), Managing releases from small upland reservoirs
for downstream recharge in semi-arid basins (northeast of tunisia), Journal of
Hydrology, 314 (1-4), 125–138.
Zlotnik, V. A. (2004), A concept of maximum stream depletion rate for leaky aquifers
in alluvial valleys, Water Resources Research, 40 (6).
Zume, J., and A. Tarhule (2008), Simulating the impacts of groundwater pumping
on stream-aquifer dynamics in semiarid northwestern oklahoma, usa, Hydrogeology
Journal, 16 (4), 797–810.
122

whole

Uploaded by

Copyright:

Available Formats

You might also like

whole

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

whole

Uploaded by

Copyright:

Available Formats

River-aquifer interactions in a semi-arid environment

stressed by groundwater abstraction

Downloaded from http://hdl.handle.net/1959.4/52606 in https://

Andrew Murray McCallum

A Thesis submitted for the Degree of

School of Civil and Environmental Engineering

3 September 2012 · Sydney · Australia

Surname or Family name: McCallum

First name: Andrew Other name/s: Murray

Abbreviation for degree as given in the University calendar: PhD

School: Civil and Environmental Engineering Faculty: Engineering

Title: River-aquifer interactions in a semi-arid environment

Abstract 350 words maximum: (PLEASE TYPE)

methods oﬀer several advantages over previously developed approaches.

Declaration relating to disposition of project thesis/dissertation

…………………………………………………………… ……………………………………..……………… ……….……………………...…….…

FOR OFFICE USE ONLY Date of completion of requirements for Award:

THIS SHEET IS TO BE GLUED TO THE INSIDE FRONT COVER OF THE THESIS

2 Groundwater abstraction and interactions 7

3 Estimating interactions using riverbed temperature time series 41

3.3.2 Proposed method . . . . . . . . . . . . . . . . . . . . . . . . . 46

4 Estimating interactions using differential river gauging 61

5 Spatial-temporal variation in interactions 77

2.1 Maules Creek study area . . . . . . . . . . . . . . . . . . . . . . . . . 11

3.1 Schematic of a temperature array installed in a riverbed . . . . . . . 49

4.1 Namoi Catchment study area . . . . . . . . . . . . . . . . . . . . . . 66

5.1 Study area . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80

2.1 Precipitation statistics . . . . . . . . . . . . . . . . . . . . . . . . . . 19

3.1 Monte Carlo analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . 47

5.1 River-aquifer statistics . . . . . . . . . . . . . . . . . . . . . . . . . . 90

– Chapter 2: Understand the links between groundwater abstraction and river-

– Chapter 3: Develop a new methodology for using heat as a tracer to investigate

– Chapter 4: Develop a new methodology for using river hydrographs to invest-

– Chapter 5: Understand the spatial-temporal variation in river-aquifer interac-

Groundwater abstraction and in-

and groundwater allocations sustainable?’, largely remain unanswered [Ivkovic et al.,

2.3 Study area

Throughout the catchment there is a network of government monitoring infra-

2.4.1 Hydrological stresses

2.4.2 Hydrological impacts

Analysis of groundwater levels

Analysis of river flows

for each period.

2.4.3 Linking groundwater gradients and river low-flows

2.5.1 Hydrological stresses

Table 2.2: Evapotranspiration statistics for pre-development (before October 1984)

Figure 2.4: Spatial distribution of cumulative groundwater abstraction since the

2.5.2 Hydrological impacts

Declines in groundwater levels

contours in Figure 2.4).

Figure 2.5: Location of selected representative observation bore hydrographs within

Losses in river flows

2.5.3 Linking groundwater gradients and river low-flows

2.6.1 Groundwater abstraction and changes in groundwater

2.6.2 Changes in river flows

2.6.3 Relationship of groundwater declines to river flow

2.6.4 Summary of processes

The processes involved in river-aquifer interactions in a semi-arid environment

2.6.5 Significance for resource management