Quantifying climate and pumping contributions to aquifer depletion using a highly parameterised groundwater model: Uley South Basin (SA)
Author(s): M. J. Knowling, A. D. Werner, D. Herckenrath
Date: February 2015
The relative contributions of climate and human stresses to aquifer depletion in real-world settings are rarely quantified, particularly where complex patterns of depletion arise from the spatial and temporal variability in aquifer stresses. These impacts can be assessed using calibration-constrained model predictions of disturbed (i.e., subject to human activity) and undisturbed (i.e., natural) conditions. Prior investigations that adopt this approach employ lumped-parameter or one-dimensional models. Here, we extend previous studies by using a highly parameterised, spatially distributed groundwater model to investigate the relative impacts of climate variability and pumping on aquifer depletion. The Uley South Basin (USB), South Australia, where there is conjecture surrounding the cause of declining groundwater levels, serves as a case study. The relative contributions of climate variability and pumping to USB depletion are shown to be highly variable in time and space. Temporal trends reflect variability in rainfall and pumping, as expected. Spatial trends are primarily dependent on the proximity to both the coastal boundary and pumping wells, and to the distribution of recharge and hydraulic properties. Results show that pumping impacts exceed those of climate between 1978 and 2012, and over the majority of the spatial extent of USB. The contribution of pumping to aquifer depletion is shown to be 2.9 and 1.4 times that of climate in terms of the time-averaged and maximum-in-time basin-scale water budget, respectively. Confidence in model predictions is enhanced by the outcomes of a linear predictive uncertainty analysis, which indicates that predictive uncertainty is lower than climatic and pumping impacts. This study demonstrates the application of a relatively simple analysis that can be used in combination with highly parameterised, spatially distributed groundwater models to differentiate causal factors of aquifer depletion.