Estimating long-term mean winds from short-term wind data

Abstract

The estimation of long-term mean winds from short-term data is especially important in the area of wind energy. It is desirable to obtain reliable estimates of the long-term wind speed from as short a period of on-site measurements as possible. This study examined seven different methods of estimating the long-term average wind speed and compared the performance of these techniques. Three linear, three weather pattern, and one eigenvector methods were compared for measurement periods ranging from 3 months to 36 months. Average errors, both relative and absolute, and the rms errors in the techniques were determined. The best technique for less than 12 months of measurement was the eigenvector method using weekly mean wind speeds. However, this method was only slightly better than the linear adjusted method. When 12 or more months of data were used, the difference in errors between techniques was found to be slight.

Keywords

This publication has 0 references indexed in Scilit: