H155-05
A Retrospective Analysis of Observed US Reservoir Operations: What Actually Happens During a Drought?

Monday, 14 December 2020: 17:46
Virtual
Jennie Steyaert, University of Arizona, Tucson, AZ, United States and Laura E Condon, University of Arizona, Hydrology and Atmospheric Sciences, Tucson, AZ, United States
Abstract:
In the contiguous United States, there are over 52,000 reservoirs that range from 0.5 to 243 meters high and collectively hold 600,000 MCM of water. These structures have dramatically impacted the hydrology of every major watershed in the country. While many studies have explored reservoir operations, existing large-scale analyses rely on derived or optimized reservoir operations as opposed to direct observations. This is primarily due to the poor accessibility of reservoir time series data. To remedy these concerns, our research has two main focuses: (1) assemble a national dataset of historical reservoir operations and time series and (2) utilize this dataset to evaluate how reservoirs have actually operated during droughts. We have assembled historical reservoir inflows, outflows and changes in storage for more than 500 major reservoirs across the US. These reservoirs cover more than half of the total storage in large reservoirs (defined as storage greater than 1 MCM and height greater than 15 m). We use these observations to evaluate the historical reservoir response to drought across the US. Meteorological drought periods were quantified using Standardized Precipitation Indices and Standardized Precipitation Evapotranspiration Indices. Hydrological drought were quantified via Standardized Streamflow Indices. All values were aggregated and calculated for USGS four-digit Hydrological Units (HUC4). Combining these hydrological and meteorological drought indices with observed reservoir releases allow us to evaluate historical responses to drought. By directly observing historical reservoir releases, our analysis provides novel insights into (1) actual large-scale management operations during drought and (2) how realistically optimized operating policies capture historical behavior.