Cargeeg, G.C., Boughton, G.N., Townley, L.R., Smith, G.R, Appleyard, S.J., and Smith, R.A. (1987c), The Perth Urban Water Balance Study, Volume 2 - Data Collection and Analysis Techniques, Water Authority of Western Australia Report and UWA Environmental Dynamics Reference ED-87-190, 124pp.

The groundwater system of the Perth area is affected by many processes which lead to changing groundwater levels and changing groundwater quality. These processes have been assessed to enable the calculation of the Perth urban water balance under a range of practical and hypothetical circumstances.

The information contained in Volume 2 describes the techniques used in collecting data and the techniques used in the modelling analysis of the water balance using the data collected. Results and conclusions are presented in Volume 1 as part of the complete discussion of the outcome of the Study. The presentations in Volume 2 make reference to three separate research projects conducted during the Study, which are documented either independently or as an Appendix to this volume. Figure 83 depicts, diagrammatically, the relationships between the contents of each volume.

There have been several previous water balance studies of the Swan Coastal Plain, each with its own clearly defined domain and time scale. Bestow (1976) estimated each component of the water balance, excluding aquifer storage changes, for a single time increment (i.e., he evaluated the long-term average water balance) and a single spatial increment (the unconfined aquifer of the Northern Perth Area). He then calculated the water balance for two scenarios: undeveloped and developed land use.

Allen (1981) proposed a simple water balance model which included all three aquifers underlying the coastal plain and provided estimates of annual average fluxes for the Northern and Southern Perth areas. Allen¡¯s domain had two spatial increments (the Northern and Southern Perth Areas) and one time increment (again, a long-term average).

The purpose of these early studies was to show relative magnitudes of the different components of the regional water balance, and to alert the community to the need for further, more definitive, studies. Because of the magnitude of groundwater extraction (estimated to be two-thirds of average annual net recharge at that time), work was recommended to improve the estimates of evapotranspiration and recharge under different topographic and land use conditions (Allen, 1981). A study known as the Net Recharge Study, since undertaken by the Water Authority and the CSIRO Division of Groundwater Research, is nearing completion.

The Metropolitan Water Authority (1985b and 1985c) assessed the water balance in regions influenced by the Mirrabooka and Wanneroo public water supply schemes in their reviews of groundwater management in those areas. The domains for these studies covered the regions immediately influenced by the production borefields. The spatial increments of their water balance model varied from 50 m (around wetlands) to 200 m and time steps as small as one week were used. The changes in fluxes were assessed over many time steps with the aid of computer modelling techniques.

Unique definitions of domain and scale have been established for the Perth Urban Water Balance Study. This does not imply any inadequacy in the domain and scales of previous studies, but rather a difference in their specific objectives. In particular, this Study focussed on the urbanised area of the Swan Coastal Plain, where factors affecting the water balance are extremely variable in space and time.

The study area, as shown in Figure 84, covers 3150 kms with highly variable topography, hydrogeology and land use. Regional data bases with scales of 100 m and 500m, i.e. with 1 ha and 25 ha cells, were chosen so that the variability of all important parameters could be adequately represented. This resolution was considered to be adequate for this Study, as well as for other Water Authority needs in the forseeable future. No other water balance study to date, at least in Australia, has utilised spatial data bases as detailed as in this Study.

Regional variations in groundwater levels and flows were predicted at 1 km and 2 km intervals across the study area using computer modelling techniques. The detailed regional results were then integrated to give global estimates, comparable with those of Bestow (1976) and Allen (1981). Estimates of the water balance and the spatial distribution of water levels and flows for localised areas within the region were also made by analysing the detailed results over those smaller domains.

In future, output from the regional model will provide accurate estimates of boundary conditions for smaller domains. For example, management reviews of the Mirrabooka and Wanneroo Groundwater Areas (Metropolitan Water Authority, 1985b and 1985c) were faced, in their initial stages, with the selection of boundary conditions for their modelled study areas. The regional model will be available to provide these boundary conditions for future studies.

The temporal domain and scales for this Study were selected to ensure appropriate levels of accuracy. The significant long-term departure from annual average conditions during the last ten years necessitated the selection of a temporal domain covering many years and also made model calibration difficult. Historical variations in the groundwater systems since 1975 have been analysed and scenarios for predicting future variations have been developed. Since groundwater levels are seasonally variable, a monthly time increment was chosen for long-term modelling.

During the Study, data were analysed both independently of and using numerical modelling techniques. This volume contains details of data collection, and of the model developed for analysis of the Perth urban water balance and assessment of management strategies.