UC Berkeley

In the coastal regions of the western United States, atmospheric rivers (ARs) are associated with the largest precipitation generating storms and contribute up to half of annual precipitation, but the impact of ARs on the integrated hydrologic cycle, specifically on groundwater storage and hydrodynamics, is largely unknown. To better explore the hydrologic behavior of AR versus non-AR event precipitation, we present a novel combination of two water tracking methods (one in the atmosphere and one in the subsurface) to explicitly track the full lifecycle of water parcels generated by ARs. Simulations of northern California's Cosumnes River watershed during the record wet 2017 water year are performed via the coupling of a high-resolution regional climate model and a land surface-groundwater model accounting for lateral groundwater flow. Despite ARs contributing more precipitation than non-AR storms, we find less AR water is preferentially stored in aquifers by year end. Fractionally, ARs result in 300% less snow derived groundwater-recharged compared to non-AR precipitation. Rain-on-snow (RoS) plays an important role in AR-driven discharge, where over 50% of total discharge from ARs snow is from RoS events. Finally, despite record-breaking annual precipitation, simulated groundwater depletion occurs by year end due to estimates of groundwater pumping activities. The results from these simulations serve as a partial analogue of future hydrologic conditions where ARs are expected to intensify and provide a greater fraction of annual precipitation due to climate change.