In the semi-arid southwestern United States, much of the water supply for seven states primarily starts as snowpack in the Upper Colorado River Basin (UCRB). With many natural (e.g. climate variability) and man-made (e.g. population increases) factors stressing these water resources, supplies are becoming more limited. Though long-term climate change can impact these water supplies, hydroclimatic variability has a much greater impact on an interannual timescale and still needs to be better understood.

Previous studies have already identified connections between the local hydrology of the western United States and Pacific Ocean sea surface temperature (SST) variability. Several studies have used this connection to create statistical models that will predict hydroclimatic variables over certain regions of the western United States on a seasonal to annual timescale. This study aims to improve these predictions for the UCRB in several ways: 1) by focusing on a sub-basin scale within the UCRB, 2) by gaining a better physical understanding of the connections between Pacific Ocean SST variability and hydroclimatic variables of the UCRB, and 3) by combining a statistical model with a hydrologic model to better predict springtime runoff on a seasonal to annual timescale.

The UCRB is comprised of eight sub-basins. Using National Weather Service COOP and Natural Resources Conservation Service SNOTEL data, monthly temperature and precipitation values are calculated for each sub-basin, for the period 1981 – 2010. These monthly values are correlated with Hadley Center gridded monthly SST data, using moving windows across the Pacific Ocean. The window with the highest correlation to a sub-basin's variable will be that sub-basin's climate predictor. Climate predictors are then used to formulate a Gaussian mixture model to forecast the most probable springtime temperature and precipitation for each sub-basin. These forecasted variables will then be used in the variable infiltration capacity (VIC) hydrologic model to simulate springtime runoff. Ideally, the coupling of a statistical model with a hydrologic model on a sub-basin scale will improve long-term springtime runoff prediction.