H178-11
Evaluating Future Hydrological Drought under A Changing Climate using A Reservoir Storage Drought Index in the United States
Abstract:
To evaluate future hydrological droughts under a changing climate in the United States, we introduce a drought assessment approach by analyzing the projected values of a reservoir storage drought index. First, both the multi-year moving averages of reservoir inflow and storage during a historic period (1970-1999) are decomposed into long-term trend, seasonal variability, and noise. A linear regression relationship between the long-term trend of the inflow and storage data is established for each reservoir. Assuming that the long-term storage trend can be estimated by reservoir inflow—and that seasonal variability remains similar, and the noise term can be neglected—we can then predict future reservoir storage variations using future inflows. In this study, reservoir inflows are simulated using the Variable Infiltration Capacity (VIC) model driven by an ensemble of six statistically downscaled Coupled Model Intercomparison Project 6 (CMIP6) general circulation models (GCMs) under the Shared Socioeconomic Pathways (SSP585) scenario.
By setting a percentage threshold for reservoir storage, a drought index is used to evaluate the severity and duration of hydrological droughts for 84 reservoirs in the conterminous United States (CONUS). Compared to the historical period, the results show that the majority of reservoirs will experience longer and more severe hydrological drought events in the near future (2020-2049). At the seasonal level, the duration and frequency of droughts will be more severe in the fall and winter than in the spring and summer. The findings are significant for purposes of regional and national scale water resources management, and disaster prevention and reduction.