Kim, Hyun-Goo;Kim, Jin-Young;Kim, Ha-Yang
19
Third-generation reanalysis data such as CFSR, ERA-Interim, and MERRA, which have improved spatial resolution and accuracy by assimilating satellite observation data, are widely used for the long-term correction of wind resource assessments. However, there is no obvious criterion for the selection of datasets, and the reported accuracy from actual application cases are all different. In this study, we provide basic information for estimating the uncertainty of reanalysis data selection by reviewing the characteristics of each dataset with a quantitative comparison of three kinds of reanalysis data. The wind speed and wind power density showed significant differences between the reanalysis data, but there was relatively little difference in the Weibull shape factor, which defines wind speed distribution. It was found that wind speed distribution in a low latitude band follows normal distribution rather than a Weibull shape. In conclusion, substantial uncertainty is expected depending on the reanalysis data, and further comparison analysis to establish its application guideline is anticipated.