Spatial and temporal variability of recharge in the U.S.

Abstract:

Diffuse recharge of precipitation is a major component of renewable groundwater. It is also the process through which climate and climate change directly impact renewable groundwater. Due to the paucity and limitations of in situ estimates of recharge, understanding the impact of climate change on recharge and groundwater often needs to rely on numerical hydrological simulation. Land surface models (LSMs) are capable of providing spatially and temporally continuous estimates of recharge at fine spatial and temporal resolutions over continental to global domains and doing so with computational efficiency. When driven by long-term atmospheric forcing data, LSMs can yield long records of recharge estimates important for climate analyses. In this study, we evaluate diffuse recharge output from a suite of LSMs forced using a common set of meteorological input data in the U.S that span nearly 40 years. Annual recharge is evaluated using an independently derived recharge dataset and the timing of seasonal maximum recharge is evaluated using in situ soil moisture and groundwater. The evaluation reveals strong impacts of model physics of ET, runoff and snow on simulated recharge at seasonal to inter-annual scales. The study finds long-term trends in recharge correlate strongly with precipitation trends while showing no correlation with temperature trends. We discuss the source of uncertainties in simulated recharge and their implications for hydrological applications and climate analyses.