Townley, L.R., Turner, J.V., Barr, A.D., Trefry, M.G., Wright, K.D., Gailitis, V., Harris, C.J., and Johnston, C.D. (1993), Wetlands of the Swan Coastal Plain, Volume 3, Interaction between lakes, wetlands and aquifers, Water Authority of Western Australia, 115pp.

This report summarises the results of a three-year study of the interaction between lakes, wetlands and unconfined aquifers. Apart from a scientific goal of achieving greater understanding of lake-aquifer interaction, the project had three specific practical objectives relating to (i) the identification of capture zones, (ii) management of water levels and (iii) the development of effective parameters for groundwater flow models in plan.

This report is based on a much more detailed report by Townley et al. [1993]. The latter contains a full description of the technical aspects of the groundwater modelling carried out during this study. A series of journal papers is also being prepared, and readers with specific interests should contact the senior author of this report for appropriate references.

Identification of Capture and Release Zones

This report focuses on lakes, rather than on sumplands or damplands. We have obtained evidence by modelling and field work to support the conclusion that the majority of lakes on the Swan Coastal Plain act as flow-through lakes which capture groundwater on their upgradient side and discharge lakewater on their downgradient side.

During this project, we have developed two-dimensional models in vertical section, two-dimensional models in plan and three-dimensional models in order to study the shape of capture and release zones as a function of nearby aquifer flows and net groundwater recharge. The depth of a capture zone depends mostly on the length of a lake, in the direction of average groundwater flow, relative to the thickness of the aquifer. The depth of a capture zone also depends on aquifer anisotropy, the resistance of low conductivity bottom sediments, aquifer inflows and outflows, and recharge. The width of a capture zone in plan is roughly twice the width of the lake.

The depth of a release zone is closely related to the depth of a capture zone. Groundwater seepage into and out of a flow-through lake is concentrated near the upgradient and downgradient edges of the lake.

The shape of a lake's release zone can be identified in the field by taking water samples and analysing for isotopic and hydrogeochemical concentrations. We have studied the release zones of Nowergup Lake, Mariginiup Lake, Jandabup Lake and Thomsons Lake using isotopic and hydrogeochemical tracers. In particular, isotopic and hydrogeochemical tracers have shown that outflow from Lake Pinjar (its release zone) becomes inflow to Nowergup Lake (its capture zone), a distance of 5.75 km downgradient.

The most cost-effective way to learn about a lake's release zone and hence its groundwater flow regime is to install a nest of piezometers or a multi-level piezometer at the middle of the downgradient side of a lake. Measurements of piezometric heads upgradient and downgradient of a lake can in principle give information about the geometry of capture and release zones, but are not as conclusive as isotopic and hydrogeochemical data.

The concentration of isotopes and chloride in lakewater and a lake's release zone can assist in the determination of a lake's water balance. Capture zone geometries vary seasonally as lake levels and surface areas fluctuate. Capture zones of lakes on the Swan Coastal Plain can be determined by regional scale modelling, coupled with results of idealised modelling of isolated lakes. Nitrate, phosphate, petroleum products and pesticides can all be carried by groundwater, but some are degraded or retarded, thus reducing their rate of movement through an aquifer.

Management of Water Levels

Water levels in lakes on the Swan Coastal Plain fluctuate seasonally, and some lakes dry out at the end of summer. Lake levels can be effectively maintained by pumping relatively small volumes of groundwater into the lakes for a few months each year. Artificial water level maintenance can lead to an improvement in lake water quality. In order to minimise its impact on lakes, pumping for public or private water supply should be located as far away as possible, both in space and in time.

Average groundwater and lake levels depend on long-term average recharge, whereas seasonal fluctuations depend on the deviations between fluctuating recharge and the long-term average. Long-term fluctuations in groundwater and lake levels depend on long-term fluctuations in recharge. Lake levels can fluctuate either more or less than nearby groundwater levels, depending on whether a lake is driven by surface water inflows or by groundwater inflows.

Effective Parameters for Models in Plan

Groundwater flow patterns near shallow lakes are fundamentally three-dimensional, but we have developed approximate methods for representing lakes in two-dimensional regional models of aquifer flow. We have developed guidelines for assigning large transmissivities to represent circular lakes in a one-layered model of a regional aquifer. We have also developed guidelines for assigning leakage coefficients to represent circular lakes in a two-layered model of a regional aquifer. Field data on the hydraulic conductivities of lake linings confirm that they are often low, but we have not related measured values to effective values needed to represent lakes in two-layered plan models.

Other Findings

We have reviewed the development of isotope balance equations for evaporating water bodies, and summarised previous literature in a concise unified framework. We have summarised the correct way of determining the angle between equipotentials and directions of flow in a vertically exaggerated cross-section through an anisotropic medium. We have developed the theory for a fully coupled groundwater and surface water model, which solves simultaneously for groundwater and surface water levels in a vertical section or in three dimensions. In the process of reviewing simple methods for predicting the movement of phosphate fronts, we have discovered inconsistencies and developed a new method for predicting travel distance. Finally, we have developed a method for combining measurements of all the components of a lake water balance to obtain estimates of the same components which are constrained to satisfy an exact water balance.

Management Implications

Lakes and wetlands are intimately coupled to the unconfined aquifer of the Swan Coastal Plain. The water quality of a lake depends on the quality of groundwater and surface water entering the lake, as well as on chemical and biological processes taking place within the lake. The concept of a "groundwater capture zone" for a lake has implications for management, in that it defines the shape of a region at the land surface within which any recharge will ultimately pass through the lake. The capture zone defines perhaps the largest "buffer zone" that could be required to be protected, in order to protect the quality of a surface water body in perpetuity.

There are many activities that may not be desirable within the capture zone of a lake, or at least within some short distance of the lake. These activities depend on the risk of contamination of recharge to groundwater, and on the nature of the particular contaminants. Although groundwater on the Swan Coastal Plain flows generally towards rivers or the ocean, capture zones are not restricted to the upgradient side of the lakes. During the winter season when recharge occurs, groundwater on the downgradient side of a lake may flow in a reverse direction, counter to the average regional flow. Capture zones may therefore extend some distance downgradient of lakes.

Although water level fluctuations are natural and may even be desirable, water levels can be effectively maintained by pumping groundwater into a lake for as little as a few months per year. Artificial maintenance can in principle maintain lake levels well above regional groundwater levels. Depending on the source of the groundwater, the water quality of a lake can even be improved. The source should preferably be as far away from the lake as possible, preferably from below the unconfined aquifer.

One finding of this study is that groundwater flow regimes near lakes can be either dominated by surface water or by groundwater. Natural lakes may have been groundwater dominated, but networks of drains may have changed the balance of some lakes so that they are now surface water dominated. The role of drains in modifying and controlling the behaviour of lakes and wetlands is not fully understood and needs careful investigation.

There appears to be some tendency at present for the responsibility for wetlands management to be passed on to local councils and Shires. While implementation of management plans for controlling access, vegetation and fauna may be handled effectively at this level of government, it is not possible for individual councils and Shires to manage the water balance of lakes in isolation from others. All lakes and wetlands on the Swan Coastal Plain are interconnected by a regional scale aquifer. Lake and wetlands will not continue to exist if the water balance of the unconfined aquifer of the Swan Coastal Plain is disrupted to the extent that they are no longer "wet". In our view, the water balance of the Swan Coastal Plain and maintenance of the condition of lakes and wetlands can best be managed at a regional level, by a suitably skilled interdisciplinary team of experts.

A poor quality black and white scanned copy of this report can be downloaded from the Department of Water (11.4 MB). Other reports on "Wetlands of the Swan Coastal Plain" can be found by searching for this phrase at the same website. A better quality searchable PDF will be uploaded soon to this website.

This report is a much abbreviated version of a 469-page report prepared for LWRRDC, which can be downloaded from this website.