Continental scale hydrogeochemistry: Groundwater system processes and prediction of groundwater chemistry through water-rock interactions across Australia

As the driest inhabited continent, Australia’s groundwater resources are essential to sustain our communities, food production, industry and environment. Groundwater hydrogeochemistry allows us to identify the processes controlling groundwater movement and quality, however, to date no integrated national spatial analysis of water-rock interactions has been conducted. As part of the Australian Government Exploring for the Future Program, this study aims to integrate geological and hydrogeochemical datasets to use water-rock interactions as a tool to identify groundwater system processes and predict groundwater composition and quality across Australia. This study uses a new spatial analysis portal (GA 2019) to combine new national hydrogeochemistry datasets (CSIRO 2019, BOM 2019) with updated solid geology, cover material and mineral thermodynamics datasets to identify dominant water-rock interactions spatially. Additionally, machine learning was used for predictive modelling of rock geochemistry (Wilford 2018) and integrated with hydrogeochemistry data for prediction of groundwater chemistry. The results of this study show groundwater chemistry across Australia is largely controlled by water-rock interactions with solid geology and cover materials. Groundwater chemistry that does not reflect geology indicates groundwater flow-paths and system dynamics. The consistency of the Spatial and hydrogeochemical relationships between groundwater and geology show that water-rock interactions can be used to predict groundwater chemistry. Machine learning applied to detailed environmental datasets has been upscaled to predict geochemistry at a national level. Integrating this predicted geochemistry with water-rock interactions allows a consistent approach for interpreting groundwater processes and predicting groundwater chemistry. This study establishes continent-wide spatial relationships between chemistry of groundwater and rocks. This demonstrates the utility of nationally consistent datasets and hydrogeochemistry to understand groundwater system processes and predict groundwater composition and quality. The applications of this work include targeting groundwater sampling of existing bores, planning of new groundwater bores and geological sampling for geochemical endmembers to benefit community water supplies, mineral exploration and environmental monitoring. Presented at the 2019 Australasian Groundwater Conference (AGC 2019)