Modern multi-megawatt wind turbines are tall and may reach heights of 200 meter. Tall wind turbines require measurements above typical measurement mast height. As tall measurement masts are expensive and cumbersome to install, wind measurements at hub height are scarce. Currently, the use of lidars in wind measurement campaigns is increasing. Lidars provide wind measurements over the full wind turbine rotor, but lidars are not always available for an extended period. This thesis investigates the possibility of applying the measure-correlate-predict (MCP) method to short-term lidar measurements in order to extrapolate wind shear statistics.

The first step was to compare the wind shear derived from lidar measurements with wind shear derived from mast measurements because the instruments have a different measurement principle. Wind data from mast-lidar pairs at Høvsøre (DK) and Breezanddijk (NL) are used for the analyses. Subsequently, a detailed study of the seasonality of wind shear, as the seasonality becomes a concern
when the measurement period is shorter than one year. The MCP model has been implemented and validated using the commercially available software WindPRO. The data from the measurement sites have been used to test the new approach of estimating the mean wind shear exponent using short-term lidar measurements and MCP. Lastly, the uncertainty in the wind shear exponent is propagated
to uncertainty in the annual energy production.

Although the wind shear statistics are subject to seasonality, this thesis shows that the proposed method has the potential to significantly reduce the error in the estimate of the mean wind shear exponent. As the error in the mean wind shear exponent is decreased, the uncertainty in the vertical extrapolation of the wind resource can be reduced. This reduction ultimately leads to a decrease of the uncertainty in the annual energy production.