Data Denial Experiments from the Wind Forecast Improvement Project (WFIP)

The U.S. Department of Energy sponsored a joint research project with NOAA and private industry to improve wind energy forecasts, called the Wind Forecast Improvement Project. The key elements of this program have been 1) a one-year deployment of extensive meteorological observing systems in two regions with significant wind energy production, from August 2011-September 2012; 2) assimilation of these observations into the research version of the hourly-updated NOAA Rapid Refresh (RAP) model and the High Resolution Rapid Refresh(HRRR) model, run nationwide at 13 and 3 km resolution; and 3) evaluation of the benefits of these improved wind forecasts on electrical utility operations, including financial savings, especially for ramp-events in the 0-6 h forecast time-frame.

The special observation data sets assimilated in each of the two study areas includes:

Northern Plains • seven 915 MHz wind profiling radars, two 449 MHz wind profiling radars, five sodars, 39 instrumented tall towers, and approximately 400 nacelle anemometers .

Texas • three 915 MHz wind profiling radars, seven sodars, and 15 instrumented tall towers .

In this presentation we will describe results from data denial experiments that have been run for limited periods within the WFIP project. The goal of the data denial experiments is to quantitatively document the precise impact that assimilation of the special WFIP data had on model accuracy, by comparing simulations from identical models run with and without the new data. NOAA/NCEP will also conduct data denial experiments with its 12 km North American Mesoscale (NAM) and 4 km NAM CONUS nest. Results from separate data denial runs will be shown for the Northern Plains study area and the Texas study area, including a meteorological analysis of ramp events.