The influence of seasonal biases on long-term correction uncertainty

In this report, the challenges associated with long-term correction techniques for site assessment, when short-term (less than one year) measurement campaigns are used, is examined. Long-term wind speed measurements from nine sites located in Northern Europe and South Africa is used in combination with reference data from simulation results from the Weather Research and Forecasting (WRF) model. The dataset is used to evaluate the uncertainties associated with long term correction from short windows of observations, and how much the uncertainties can be reduced if the seasonality of the reference-data biases can be better predicted in the future. The main conclusions of the study are:
1. Short-term measurement campaigns are currently unfeasible because the mean wind speed uncertainty grows rapidly with a shortening of the measurement time-window below one year.
2. Knowing the full seasonal variations of the reference-data biases (by some future technique) allows a significant reduction of the uncertainty, upwards of 50 %, for 3 - 9-month campaigns. However, for very-short-term (1-2 months) campaigns the uncertainties remain about three times larger than a full-year campaign even when correcting for the seasonal variations of the reference data biases.