PROBABILISTIC PREDICTIONS OF GROUNDWATER RECHARGE UNDER RAIN-FED CROPLAND IN THE SOUTHERN HIGH PLAINS

Understanding how groundwater recharge changes with land-use and climate change is critical for managing water resources. Motivated by the need to prepare for the full range of possible outcomes, we present a framework for assessing recharge changes that explicitly accounts for uncertainties in soil and vegetation properties and future climate. The study focuses on a semi-arid southern High Plains (SHP) area that underwent land-use change to rain-fed cotton about 75 years ago. Negligible diffuse recharge occurred under natural vegetation, yet the shift to cropland now allows appreciable amounts. Recharge will likely further be impacted by future climate change. Using Monte Carlo simulations, we generated ensemble predictions of diffuse recharge at the site for different climate alternatives from various general circulation models. Soil and vegetation parameter realizations were conditioned on soil moisture and soil water chloride observations, which was crucial for reducing model uncertainty. A stochastic weather generator was used to represent meteorological variability from each climate alternative. For most climate alternatives, predicted changes in average recharge (spanning −75% to +35%) were larger than the corresponding changes in average precipitation (spanning −25% to +20%). This suggests that amplification of climate change impacts may occur in groundwater systems. Predictions also include varying changes in the frequency and magnitude of recharge events. The temporal distribution of precipitation change relative to the growing season explains most of the variability in predictions of recharge totals and episodic occurrence. These results show how land-use and climate change conditions can interact to impact recharge and groundwater systems.