Streamflow simulation using a water-balance model with annually-resolved inputs

Availability of long-term information on the variability of water resources in a given area is particularly important for sustainable resource management. Developing watershed simulation models that can be run using annually-resolved proxy climate data provides a way to improve reconstructions of hydrological parameters over multi-century time scales. Through the addition of a snowmelt modeling component, we enhanced a simple water-balance model to simulate streamflow at seasonal resolution. The model was calibrated to the upper Meadow Valley Wash, Nevada, USA, using USGS gage number 09417500 streamflow records. PRISM data at 2.5 arc-min resolution were used to reconstruct streamflow at the seasonal timescale (October through May) from 1896 to 2008, with and without a temperature-index snowmelt component. Best-fit model simulations had an R2 of 0.81 against stream gage observations. Average predicted seasonal streamflow during calibration was 0.81 cm with a standard deviation of 0.35 cm, compared to the observed average seasonal streamflow of 0.76 cm and standard deviation of 0.48 cm. Despite some shortcomings for this watershed, the model approach has promise for providing scenario-based estimates of hydrologic variability in semi-arid mountain environments.