A Multi-decadal Analysis of U.S. and Canadian Wind and Solar Energy Droughts

Wind and solar energy droughts are evaluated over the contiguous United States and most of Canada using a bias-corrected version of the ERA5 reanalysis data set for the time period of 1959-2020. In the first step of this analysis wind speed and irradiance variables are compared from the original ERA5 reanalysis and from multiple observational data sets.* The ERA5 and observed meteorological variables are then converted into wind power and solar power, and also compared. These comparisons demonstrate that significant errors in the ERA5 reanalysis exist limiting its direct applicability for a wind and solar energy drought analysis. Corrections to the ERA5 are then derived that greatly reduce these errors, while also largely avoiding potential over-inflation of the reanalysis variability resulting from differences between point-measurements and the temporally and spatially smoother reanalysis values. Wind and solar droughts are then analyzed in terms of their intensity, duration, and spatial extent. Questions that will be addressed are:

• How significant are wind and solar energy temporal trends, and how do these trends vary across the analysis domain?
• Which have larger amplitude, wind energy or solar energy droughts?
• How do the intensities of wind and solar energy droughts vary with duration?
• How do the characteristics of wind and solar energy droughts vary regionally?
• Do wind and solar droughts vary seasonally?
• Are wind and solar energy droughts correlated in time?
• Are wind and solar energy droughts correlated with temperature extremes (when electric system loads are largest)?
• How frequently do the most intense droughts reoccur?

*For solar, the NOAA SURFRAD and SOLRAD networks, and the DOE ARM Southern Great Plains (SGP) array; for wind, the WFIP1 and WFIP2 tall tower, lidar, and sodar arrays, the DOE ARM SGP lidar array, the New York Mesonet lidar network, and additional tower and lidar observations both on and offshore.