Different Precipitation Elasticity of Runoff for Precipitation Increase and Decrease at Watershed Scale

The precipitation elasticity of runoff, defined as the percentage change in mean annual runoff for a given percentage change in mean annual precipitation, has been widely used as a simple rule of thumb for estimating runoff sensitivity to changes in precipitation. Most applications assume the same runoff sensitivity for both an increase and a decrease in precipitation. This study examines this assumption using long historical annual precipitation and streamflow data from 353 watersheds across the United States. The results show that the assumption is acceptable in humid and subhumid regions (aridity index less than 1.0). However, in arid and semiarid regions (aridity index greater than 1.0), the sensitivity of runoff to precipitation is greater for a decrease in precipitation than for an increase in precipitation. Therefore, the use of precipitation elasticity of runoff estimated using the entire data set in semiarid and arid regions will result in an overestimation of runoff increase from precipitation increase and an underestimation of runoff decrease from a decrease in precipitation. These findings suggest that multiple precipitation elasticities of runoff may be needed to estimate nonlinear responses of streamflow to precipitation change in a changing climate, especially for persistent increase and decrease in precipitation.