On the accuracy of a logarithmic extrapolation of the wind speed measured by horizontal lidar scans

Abstract

Remote sensing-based wind power forecasts are nowadays being increasingly investigated. Long-range lidar scans are hereby often performed at low heights, causing the need for a wind speed extrapolation to hub height. In this work we analysed the accuracy of the stability corrected logarithmic wind profile and its sensitivity to atmospheric stability, wind speed and extrapolation height by means of a theoretical error estimation using error propagation. Emphasis was given to analyse the contributions of the profile's individual variables but also considering the measurement campaign framework. We further used lidar measurements at the offshore wind farm Global Tech I to support the theoretical analysis. The logarithmic profile was found to be able to describe profiles during most situations, however, decreasing wind speeds with height cannot be represented. Results showed that due to the nature of the stability correction term extrapolation errors are largest during very stable atmospheric conditions. Here, stability estimation errors were dominant. Under near neutral and neutral atmospheric conditions the wind speed error contributed most to the overall error. We conclude that extrapolation errors can mainly be reduced by optimising the estimation of atmospheric stability using accurate measurement devices. Furthermore, the precise horizontal alignment of the lidar device is important.