H33I-1740
Sensitivity of streamflow to climate change in California

Wednesday, 16 December 2015
Poster Hall (Moscone South)
Ted Grantham, USGS Fort Collins Science Center, Fort Collins, CO, United States, Daren Carlisle, USGS, Reston, VA, United States, David Wolock, USGS, Lawrence, KS, United States, Greg McCabe, USGS, Denver, CO, United States, Michael Wieczorek, USGS, Baltimore, MD, United States and Jeanette Howard, The Nature Conservancy, San Francisco, CA, United States
Abstract:
Trends of decreasing snowpack and increasing risk of drought are looming challenges for California water resource management. Increasing vulnerability of the state’s natural water supplies threatens California’s social-economic vitality and the health of its freshwater ecosystems. Despite growing awareness of potential climate change impacts, robust management adaptation has been hindered by substantial uncertainty in future climate predictions for the region. Down-scaled global climate model (GCM) projections uniformly suggest future warming of the region, but projections are highly variable with respect to the direction and magnitude of change in regional precipitation. Here we examine the sensitivity of California surface water supplies to climate variation independently of GCMs. We use a statistical approach to construct predictive models of monthly streamflow based on historical climate and river basin features. We then propagate an ensemble of synthetic climate simulations through the models to assess potential streamflow responses to changes in temperature and precipitation in different months and regions of the state. We also consider the range of streamflow change predicted by bias-corrected downscaled GCMs. Our results indicate that the streamflow in the xeric and coastal mountain regions of California is more sensitive to changes in precipitation than temperature, whereas streamflow in the interior mountain region responds strongly to changes in both temperature and precipitation. Mean climate projections for 2025-2075 from GCM ensembles are highly variable, indicating streamflow changes of -50% to +150% relative to baseline (1980-2010) for most months and regions. By quantifying the sensitivity of streamflow to climate change, rather than attempting to predict future hydrologic conditions based on uncertain GCM projections, these results should be more informative to water managers seeking to assess, and potentially reduce, the vulnerability of surface water supplies to climate change in California.