Homogenization of Surface Wind Speed Over Global Land From 1973 to 2022

Abstract Although homogenization has been widely applied to historical records of surface air temperature, the homogenization of surface wind speed (SWS) data remains limited likely because the SWS is more sensitive to local environment, making it difficult to find suitable reference from nearby stations. To address this issue, we have developed a method for calculating geostrophic wind speed, which has been used as the reference to perform homogeneity detection and adjustment for the SWS at stations in China. Here, the method is applied to homogenize annual mean SWS from more than 3,600 stations over global land from 1973 to 2022. Although further confirmation from the metadata is required, the homogenization significantly reduces statistical inhomogeneity ( p = 0.01) in the raw SWS. The results show that about 35% of the global stations exhibit inhomogeneities in SWS, and this phenomenon is more pronounced over North America and Europe. The homogenization has a minimal impact on the global average trend of SWS, but has major impacts on regional trends, especially in Europe, Asia, and Oceania. Compared with the raw SWS, the trend of homogenized SWS from 1973 to 2022 changed from −0.052 to −0.027 m s −1 per decade in Europe, and from −0.066 to −0.033 m s −1 per decade in Asia. In Oceania, the homogenization decreases the SWS trend from 0.24 to 0.081 m s −1 per decade. This study emphasizes the importance of considering the homogeneity of SWS data when analyzing its changes, particularly at regional scales.