Evaluation of a Hydrologic Model Applied to a Headwater Basin in the Rio Grande (USA) Using Observed and Modelled Land Surface Fluxes and States

Researchers at the Desert Research Institute are conducting research aimed at developing andcalibrating both operational and physically based numerical models that can be used to predict the quantityand timing of runoff in semi-arid regions where the majority of runoff originates in the seasonal snow pack.Unfortunately, observations of hydrologic variables (precipitation, streamflow, evapotranspiration, snowwater equivalent, etc.) are sparse in the semi-arid regions of the western United States and, therefore, theevaluation of model accuracy (usually in terms of streamflow) is often very limited. However, comparisonsof model output with newly developed high-resolution estimates of hydrologically based land surface fluxesand states may provide insight to model accuracy in areas with little or no observed information. In thisstudy, we apply a hydrologic model to a watershed in the headwaters of the Rio Grande to simulate thestreamflow generated at the watershed outlet and several internal subwatersheds. The model simulations ofstreamflow are compared to values from long term land surface model studies and observations at streamflowsurface water stations. Additional comparisons of model snow water equivalent (SWE) estimates are madewith the SWE values from the long term land surface model studies and SWE observations at three pointlocations within the watershed.