Interannual-to-Decadal Variability of Wind Power Density over the Continental US: Sensitivity to Model Resolution and Teleconnections

In order to increase the adoption of wind power throughout the world, the variability of the wind resource must be more predictable, both short and long-term. According to a study done by NREL, measured interannual variability of wind power can be as high as 38%. This large magnitude of variability, if not properly forecast, can dissuade developers if a wind farm underperforms and make decisions for balancing authorities more difficult when trying to match energy demand.

To help address the problem, this study focuses on wind power density variability in three regions of the United States known for large wind resources (Southern Great Plains, Northern Great Plains, and Southern California) at the current typical hub height of 80 meters. The variability is analyzed using a high-resolution regional climate model (WRF) and global climate models (from the CMIP5 archive) at differing resolutions over a time span of multiple decades. Preliminary results indicate that higher resolution climate modeling is better able to identify wind resource changes due to topographical features than lower resolution modeling.

Statistical analysis is carried to identify correlations between wind power variability and teleconnections such as ENSO (El Niño-Southern Oscillation) to see what influence these climate phenomena have upon the variability of the wind resource in each of the three regions. The remote influence of Atlantic sea surface temperature variability is also investigated, as it can affect the Great Plains Low Level Jet. Identifying and explaining these correlations can help wind developers plan where to site new wind farms for maximum production efficiency.