Radar Wind Profiler Deployments to Improve Short-Term Forecasting for Wind Plants

Successful wind plant operations require reliable information regarding the daily potential wind energy to ensure efficient energy production. Wind farm power projections influence the production of energy from more conventional power sources to maintain a stable delivery of power to the grid. Previous research (Benjamin et al., Month. Wea. Rev., 2010) has demonstrated that assimilation of wind profiling radar observations into numerical weather prediction (NWP) models improves the accuracy of wind forecasts in the troposphere. More recent results from the Wind Forecast Improvement Project, conducted in 2011-2012, have shown that assimilation of wind profiler and ancillary observations improves the skill of NWP models in forecasting wind turbine hub-height winds, particularly reducing large forecast errors and errors associated with wind ramp events.

In the contiguous U.S., the majority of large-scale weather systems propagate from the west to the east, first impacting the west coast states. A better description of the position and strength of these systems and assimilation of that information into NWP models will result in increased short-term (0-12h) model forecast skill for western states. As those weather systems propagate eastward, improved model initial states based on data from the western US are also likely to improve model skill for longer term forecasts over the U.S. Great Plains and further east. This improvement may also have a positive impact on the “next day” forecasts that are the basis for the utility unit commitment process.

Consequently, a network of three 449-MHz radar wind profilers will be deployed along the coast of Washington and Oregon, completing a “picket fence” that includes four profilers along the California coast (deployed under a separate project). These profilers will be spaced at approximately 250 km intervals, from Santa Barbara in Southern California to Forks in Washington's Olympic Peninsula. Profiles of the wind speed and direction are measured up to 6 to 8 km above ground level, as well as virtual temperature profiles up to ~2 km. Both height coverages depend on atmospheric conditions. Hourly data from these instruments will be made available to weather forecasters, wind energy producers, and other end users on-line, in near real-time, and also through a data archive. The project to procure, test, and install these profilers is currently underway, with installation anticipated during the summer of 2015.