Simulating the Impacts of Projected Climate Change on Streamflow Hydrology for the Chesapeake Bay Watershed

Abstract

A gridded model was developed to simulate the hydrology of the Chesapeake Bay Watershed, the largest estuary in the United States. CMIP3 and CMIP5 climate projections were used to drive the model to assess changes in streamflow and watershed-wide hydrology. Index of agreement values indicated good model performance. Annual average temperature is projected to increase 1.9°C to 5.4°C by 2080 to 2099, with the greatest warming occurring in summer and fall in the northern part of the watershed. Annual total precipitation is projected to increase between 5.2 percent and 15.2 percent by 2080 to 2099, with the largest increases generally occurring in winter. Average evapotranspiration and rainfall are projected to increase while snowfall, snow water storage, and snowmelt decrease. Subsurface moisture is projected to decrease during the warmer months and the time to recharge increases and, in some cases, never actually occurs. Changes in annual runoff for all 346 climate projections averaged 0 percent (2020–2039), –1.5 percent (2050–2069), and –5.1 percent (2080–2099). There is a 48 percent, 52 percent, and 60 percent chance, respectively, for the future time periods that annual runoff will be less than baseline values (1950–1999). Extreme runoff projections are overwhelmingly associated with the negative end of the distribution. Runoff increases are confined to January through March and to higher elevations. This study is novel in its use of a large number of climate models, the gridded nature of the hydrologic model, and the simulation of several hydrologic variables, all of which allowed for the assessment of both uncertainty in the projections and variation across multiple spatial and temporal scales.