| Summary: | Study focus: Future climate impacts on streamflow are of critical concern due to its importance for water and energy supplies. However, rigorous studies of such impacts involve complicated modeling chains that can be time-consuming and costly to implement. We examine an alternative approach by developing a method to predict the 21st century peak March-May (MAM) streamflow response to warming from historical data alone, specifically, past observations of the Snow-to-Precipitation ratio divided by basin Area (SPA). Study region: Using the Colorado River Basin (CRB) as a proving ground, we test this empirical relationship over 23 basins ranging in size from 9728 to 639,806 km2 within the CRB. Data was derived from hydrology model simulations forced by a suite of climate models, run under two emissions scenarios. New hydrological insights: Our proof-of-concept study revealed useful predictive capability (r2 = 0.58) and additionally identified three response types: rain- (basin SPA < 1.2 × 10−3 km-2), snow-rain- (1.4 × 10−3 to 3 × 10−3 km-2), and snow-dominant (>3 × 10−3 km-2). The most dramatic changes occurred in snow-dominant basins where streamflow increased up to 320 % in MAM, but decreased by up to 60 % in the critical high demand months of July-September (JAS). By quantifying the relationship between historical data and future streamflow projections, our findings suggest this empirical relationship is useful to quickly and cheaply determine where and when large changes in hydrology will occur from future warming. Keywords: Streamflow forecasting, Snow, Climate variability, Modeling, Water resources management
|
|---|