Characterizing shifts in historical streamflow extremes in the Colorado River Basin, USA

Abstract:

Global temperature increases are leading to large changes in critical watersheds worldwide. Many of these changes have substantial impacts on terrestrial hydrology including a higher incidence of extreme events such as droughts or floods, changes in vegetation, and shifts in the quantity and timing of snowmelt and associated runoff. Nowhere in the continental United States are these issues more pressing than the Colorado River Basin (CRB) due to its high rate of population growth raising the demand for water supplies, projections of accelerated warming relative to other areas, and the large number of climate-driven disturbance impacts including drought-induced vegetation mortality, wildfires, and insect infestations. Our work examines how modifications such as these are affecting historical streamflow extremes in the CRB using gages known to have minimal impacts from management. First, the Mann-Kendall trend analysis is applied to several important extreme streamflow metrics comprising of high- and low-flow percentiles, 7-day minimum and maximum flows, and the center timing of flow to quantify the magnitude and direction of changes in extremes. Statistics are evaluated at the annual, seasonal, and monthly timescale using daily historical streamflow observations with a minimum record length of 30-years (1985-2014) to a maximum of 60-years (1955-2014). We then applied the Generalized Extreme Value (GEV) theorem to determine how the tails of streamflow distributions have shifted over the same time periods. Results indicate that the largest trends with the highest level of statistical significance are occurring during the late winter to spring high flow season. Moreover, there is an overall decreasing trend in the metrics that were tested for all gages with the largest decreases occurring in the 7-day minimum and maximum flows and at higher elevations. Given the importance of the 7-day flows to ecosystem structure and function—coupled with the occurrence of more significant trends at higher elevation sites that prior to now are likely to have experienced minimal disturbance from humans—these results have great consequences for improved land management strategies that will be necessary to offset damage to the ecosystem as the climate continues to warm.