SIMULATING HYDROLOGIC RESPONSES IN THE COLORADO RIVER TO CLIMATE CHANGE

The Colorado River Basin (CRB) has experienced widespread and prolonged drought with recent (2000 to 2014) precipitation up to 25% below and air temperature up to 0.8^oC warmer than 20^th century means, and Lake Mead level approaching a Tier 1 shortage. Climate models project more frequent and intense droughts and heatwaves, with major consequences for water resources in the southwestern U.S. Prior studies that have estimated streamflow changes in the CRB have employed the Variable Infiltration Capacity (VIC) model. We utilize recent advancements in the model formulation and land surface parameter specification for the VIC model, in addition to improved meteorological forcings, to produce long-range streamflow projections in the CRB. Meteorological datasets were obtained from gridded (1/16^o spatial resolution) daily observations (1950 to 2013) and downscaled historical (1950 to 2005) and future-projections (2006 to 2099) from 8 selected General Circulation Models of the Climate Model Intercomparison Project 5 that represent the historic climatology well. Daily forcings were disaggregated to hourly resolutions with the MetSim Python package, which was recently improved to approximate precipitation using monthly-averaged storm durations and peak intensities computed from observations. We use monthly-averaged land surface parameters derived from remote sensing products and updated soil parameters that were calibrated for major CRB subbasins using the current model setup. We employ the R-VIC channel routing model to simulate naturalized streamflow for major CRB subbasins. Based on this model setup, we conduct two main analyses: (1) evaluating the effects of employing the current VIC model setup by comparing results to earlier studies, and (2) quantifying spatiotemporal variability of major water balance terms from the updated simulations. For this latter analysis, mean annual and monthly anomalies of model results in historic and future periods were compared to identify changes in streamflow and their linkages to shifting precipitation and snowpack dynamics. Climate model ensemble statistics were also computed to gauge climate uncertainty. Results show a change in how different subbasins respond in terms of streamflow resulting from shifting snowpack dynamics under climate change.