Peter Troch1, Ravindra Dwivedi1, Rodrigo Valdés-Pineda1, Antonio Alves Meira-Neto1, Tirthankar Roy1, Tao Liu1, Matej Durcik2, Saúl Arciniega-Esparza3, and José Agustín Breña-Naranjo3
1Department of Hydrology and Atmospheric Sciences, The University of Arizona, Tucson, Arizona
2Biosphere 2, The University of Arizona, Tucson, Arizona
3Instituto de Ingeniería, Universidad Nacional Autónoma de México, Mexico City, Mexico
It is often difficult, if not impossible, to predict groundwater recharge at the catchment scale, but such estimates often require the use of complex numerical groundwater models whose results can be biased due to large uncertainty in model parameters at the large scale. Here we present an alternative and a much simpler method for estimating large-scale groundwater recharge for gauged and ungauged catchments, assuming that long-term median (50th percentile) streamflow can be used to approximate groundwater recharge. We employed catchment-scale water balance & baseflow separation methods for estimating maximum total and deep storages from water years 1980-2002 for 247 Model Parameter Estimation Experiment (MOPEX) catchments across the conterminous USA. We then compute a vegetation water use efficiency metric, the Horton index (HI; the ratio of catchment-scale vaporization to wetting) for the same dataset. Our results show that the HI is strongly (negatively) correlated (R2=0.77) to deep storage and to the 50th percentile flow in the flow duration curve so that it can be used for estimating long-term groundwater recharge at the catchment scale. Furthermore, we found that our HI-based groundwater recharge estimates are similar to the groundwater recharge map published by the USGS (R2=0.69) for the conterminous USA. Finally, our method suggests that the vegetation water use efficiency is a sound predictor for groundwater recharge in gauges and ungauged catchments located across different climates.