Uncertainties connected to long-term correction of wind observations

Abstract

Good knowledge about on-site wind climate is necessary in wind power assessments. Due to yearly and natural variability, a few years' measurements do not give information about wind climate. Long-term reference data can be used to correct observations to better correspond a normal year. The method doing this is commonly referred to as measure-correlate-predict (MCP). Uncertainties are introduced in predicted wind climate due to method weaknesses. In this paper uncertainties connected to length and season of on-site measurements are studied. An on-site series of 29 years, different MCP-methods and reanalysis datasets have been used. An additional high-resolution model dataset has been used to study variation within a wind farm. At least 12 months of observations and a long-term reference longer than 15 year are recommended. Large seasonal influences are seen; summer periods tend to overestimate while winter periods underestimate wind climate.