Knowledge of the exchange between streams and underlying aquifers is required for management in systems where both surface water and groundwater are exploited and also well connected. In northern Victoria, the Campaspe River catchment (within the Murray Darling Basin), downstream of Lake Eppalock, is one such catchment in which it is necessary to quantify the impact of changes to allocations/extractions of surface water and groundwater on stream-aquifer exchange. The primary tool for assisting such quantification is a numerical model of the system that is constrained by field observations of groundwater head and stream stage. However, quantifying stream-aquifer exchange is notoriously difficult, and so any data (type, frequency, location) that may reduce the uncertainty in model predictions is desirable. The aim of this study is to quantify the reduction in uncertainty of simulated stream-aquifer exchange rates over a variety of spatial and temporal scales through the acquisition of innovative data beyond traditional groundwater head and stream stage data. In particular, carbon-14 (C-14), radon, and electrical conductivity data that were collected within the study area are investigated in terms of their “worth”, defined here as their ability to reduce stream-aquifer exchange predictive uncertainty.
Utilising the parameter estimation (PEST) software suite, the traditional and innovative data supported the calibration of a regional-scale groundwater flow (MODFLOW-NWT) and transport (MT3D-USGS) model. The calibration-constrained model has been used to test the extent to which these innovative data provide a reduction or otherwise in uncertainty for predictions of stream-aquifer exchange. Analysis of modelled stream-aquifer exchange was carried out at reach and whole of river spatial scales, and for each of these, at annual, seasonal, and monthly temporal scales.