Incorporation of Recharge Uncertainty in Basin-Scale Water Resource Models in Semi-Arid Environments

Quantifying the amount of recharge to a groundwater flow system and its uncertainty is paramount for water resources investigations in semi-arid basins. In this study, we investigate the impacts of recharge uncertainty on drawdown predictions for the recently completed interbasin groundwater transfer at Honey Lake Valley (HLV), Nevada. Recharge estimates for HLV are computed using 1000 sets of optimized recharge coefficients generated by the study of Epstein et al. (2009) that empirically relate mean annual precipitation in mountain block watersheds to recharge, given measurement error in precipitation, variability of independent recharge estimates and regression model uncertainty. Two conceptual approaches are used to incorporate recharge uncertainty into a water resource model of HLV. The first conceptual approach combines recharge contributions from each watershed into a single distribution for the basin. Recharge values and hydraulic conductivity of the basin fill are correlated using an empirical relationship determined during model calibration. The second conceptual approach allows recharge contributions from each watershed and hydraulic conductivity values for a subset of hydrogeologic units to vary independently. Evaluation of the conceptual models occurs through comparisons of drawdown predictions of mean, median and 95 percent confidence intervals generated by 2500 Monte Carlo realizations.