全球变化知识资源中心

Analyzing future changes in hydrologic extremes such as floods, low flows, and soil moisture extremes is important because many impacts on ecosystems and human systems occur during extreme events. To quantify changes in hydrologic extremes, this study conducts hydrologic modeling experiments over 20 Midwestern watersheds using the Variable Infiltration Capacity (VIC) model forced by historical observed datasets and future projections from statistically downscaled Global Climate Model (GCMs) simulations. Our results show that peak daily streamflow at the 100-yr reoccurrence interval will increase ( +10-30%) in most watersheds by 2080s due to significant increases in precipitation (P) and increasing P as rainfall during winter and spring seasons. The simulations also show strong shifts towards earlier peak flow timing (up to a month), especially in strongly snowmelt-dominated watersheds. These effects are linked to strong decreasing trends in maximum Snow Water Equivalent (SWE) with warming, which are simulated over essentially the entire domain. Projected changes in 7-day extreme low flows are smaller in magnitude ( - 10-+10%) with some what larger decreases simulated at the end of century; however, the timing of extreme low flows is projected to shift from winter/spring to summer and fall in strongly snowmelt-dominated watersheds in the northernmost parts of the domain. Extreme low soil moisture increases over most of the domain in the future projections up to the 2050s, but by the 2080s there are more widespread decreases in extreme low soil moisture, especially in the northernmost parts of the domain. (C) 2018 Elsevier B.V. All rights reserved.