Catchment-scale groundwater-flow and recharge paradox revealed from base flow analysis during the Australian Millennium Drought (Mt Lofty Ranges, South Australia)

Catchment-scale recharge and water balance estimates are commonly made for the purposes of water resource management. Few catchments have had these estimates ground-truthed. One confounding aspect is that runoff and soil-water inputs commonly occur throughout the year; however, in climates with strong dry seasons, base flow can be directly sampled. In an experimental catchment in the Mt. Lofty Ranges of South Australia, run-of-stream hydrochemical parameters were monitored. In this Mediterranean climate during the Millennium Drought (2001–2009), the stream was reduced to disconnected groundwater-fed pools. Two groundwater types were identified: (1) high-salinity type from meta-shale bedrock with thick, clayey regolith and (2) low-salinity type from meta-sandstone bedrock with sandy regolith. End-member mixing using silica and chloride concentrations and robust 87 Sr/ 86 Sr ratios reveal an apparent groundwater-flow paradox as follows. According to chloride mass balance and spatial distribution of hydrogeological units, the low-salinity groundwater type has seven times more recharge than the high-salinity type. Over the 28-year record, low-salinity groundwater contributed 25% of stream water, whereas high-salinity groundwater contributed 2–5%. During the drought year, however, annual stream flow from the high-salinity groundwater contributed 50%, whereas low-salinity groundwater contributed 18%. High-salinity groundwater dominated dry-season base flow during all years. The paradox can be resolved as follows: The meta-sandstone terrane drains quickly following wet-season recharge and therefore contributes little to dry-season base flow. Conversely, the meta-shale terrane drains slowly and therefore provides stream flow during dry seasons and drought years.