conceptual formulations are used to generate salt export along the drainage lines. The model is a component of a decision support framework for exploring the impact of policy and programs on irrigation futures over a 20 year time period.

Introduction

Highly saline shallow groundwater fluctuations effect agricultural productivity and stream water quality. Shallow watertable fluctuations increase the potential of waterlogging and land salinisation. This is most evident within irrigated areas of eastern and southern Australia where records show that the groundwater table has risen markedly resulting in an increase in land salinisation. Salinity is a significant cause of land degradation in south-west Victoria. Currently 40,000 ha of land is affected; resulting in annual losses of $4million to agricultural production. In order to formulate effective long term farm management practicesand options it is necessary to quantify the dominant processes effecting watertable movement.

Fluctuations in the watertable are in response to unsaturated and saturated physical processes that occur both on a local and regional scale. Mathematical models havebeen developed to gain a better understanding of the physical processes and to predict the catchment response to changed land use practices. The mathematical models must be capable of representing the spatial and temporal variability in climate, soil, hydrogeology and vegetation at the appropriate scale.

Few mathematical models are currently available that purport to simulate physical processes affecting water table fluctuations. To investigate processes at a point scale the crop water use and solute transport model called SWAGMAN Destiny has been developed by CSIRO. This model is comparable to the one-dimensional eco-hydrological model WAVES (CATCHWORD 1997). On a sub-catchment scale, SWAGSIM (also developed by CSIRO) has been widely used. This model approximates evaporative demand based on a modified Penman formulation and adopts an analytical solution to estimate net recharge to a single layer groundwater model (Prathapar et al .1995). The fully distributed physically based catchment model Mike-SHE has been applied to irrigated catchments with some success (Mudgway et al. 1997) butrequires considerable data, computational resources and simulation time.

The objectives of the current project was to formulate and integrate a simplified unsaturated module into the US Geological Survey three