Some hydrogeological systems are at the intersection of processes that interact in complex ways. One such example is the lower River Murray floodplain. Regional groundwater flows, river conditions, climate, land use, and environmental watering all impact potentiometric heads and groundwater salinity. This is of concern as the groundwater influences both vegetation health within the floodplain and also river water quality downstream.
Numerical groundwater modelling supports management decisions for the River Murray. South Australia’s Department for Environment and Water has built a number of groundwater models of the SA floodplain; these vary greatly in terms of the processes simulated, their spatial and temporal resolution, and approach to salt movement. The models aim to estimate groundwater flux to the river, potential risks from groundwater to vegetation, or both.
This talk presents results from comparisons between four styles of groundwater model: (i) a regional model which neglects short-terms processes, (ii) a regional model modified to include short-term processes, but not further calibrated, (iii) a regional model that simulates flooding and environmental watering, and (iv) a detailed, floodplain-scale model that simulates both short-term flow processes and solute transport.
The development, calibration and validation of these models has demonstrated that the modelled floodplain water balance is extremely sensitive to assumptions, with sometimes unexpected interactions. The findings underscore the difficulties of retro-fitting a model designed for one purpose to answer questions of other kinds. They also demonstrate how each key process – changing river levels, evapotranspiration, managed and unmanaged inundation – influences floodplain water balances during and after floods. Some essential parameters are still poorly known, due to heterogeneity or difficulties in data collection. Nevertheless, it is possible to build floodplain groundwater models that produce sound results, as demonstrated from post-audits.
As environmental management of floodplains becomes more nuanced due to water availability, the ability to achieve the best environmental outcomes for the water available becomes critical. Groundwater modelling has a part to play in this, and understanding how the model assumptions impact the outcomes is an important step to be able to use these tools in the future.