In response to the U.S. Department of Energy’s (DOE’s) goal of using 20% wind energy by 2030, the Wind Integration National Dataset (WIND) Toolkit was created to provide information on wind speed, wind direction, temperature, surface air pressure, and air density on more than 126,000 locations across the United States from 2007 to 2013. The numerical weather prediction model output, which is gridded at 2 kilometers and at a 5-minute resolution, was further converted to detail the wind power production time series of existing and potential wind facility sites.
Users of the WIND toolkit need to know if the the information presented in the WIND Toolkit is accurate and that errors are known, so corrective steps can be taken. Therefore, a team of researchers from the National Renewable Energy Laboratory are providing a validation code written in R that will be made public to allow users to validate data at their own locations. Validation is based on statistical analyses of wind speed, using error metrics such as bias, root mean squared error, centered root mean squared error, mean absolute error, and percent error. Additionally, plots of diurnal cycles, annual cycles, wind roses, histograms of wind speed, and quantile-quantile plots are created to visualize how well observational data compares to model data. Ideally, validation will confirm beneficial locations to utilize wind energy and encourage regional wind integration studies using the WIND Toolkit.