Assessing impacts of climate change in a semi arid watershed using downscaled IPCC climate output

This presentation discusses our research aimed at helping water managers at Salt River Project (SRP), Phoenix, assess long term climate change impacts for the Salt and Verde River basins, and make informed policy decisions. Our goal was to assess the future 100 year water balance by development, application and testing of a physically based distributed hydrologic model forced by downscaled IPCC climate information. The variable infiltration capacity (VIC) model was set up to simulate historical observed streamflow at the outlet of Salt and Verde River basins using gridded observed precipitation and temperature data. The model was calibrated using the Shuffled Complex Evolution (SCE-UA) optimization algorithm. The models found to best simulate the climatology of the region, (UK-HADCM3, MPI-ECHAM5, NCAR-CCSM3) and emission scenarios (A1B, A2, B1) from the Global Climate Model’s (GCM’s) participating in the IPCC fourth assessment were obtained from the bias-corrected and spatially downscaled climate projections derived from the World Climate Research Programme's (WCRP's) Coupled Model Intercomparison Project phase 3 multi-model dataset. The data was then temporally downscaled to serve as forcing for the VIC model. This downscaled forcing dataset was used to analyze the basin scale responses to climate change. Based on stakeholder feedback two additional GCM's one that represents a wet scenario and one that represents a dry scenario, were also downscaled as mentioned above and run through the hydrologic model. All the models show a statistically significant increase in temperature over the 21st century. Due to increased winter temperatures the multi-model mean shows a significant decrease in storage capacity in the basin, viz. snow water equivalent. This decrease is already evident in observed SNOTEL records of the basin. Since these watersheds are snow dominated, the cold season multi-model mean streamflow shows a decreasing trend by the end of the century though the warm season streamflow tends to increase in response to increased summer precipitation. Increased summer streamflow does not compensate for the decrease in winter streamflow. In addition to the above analysis a synthetic study was performed to quantify the uncertainty in coupled GCM-Hydrologic model predictions.