Abstract: The China Meteorological Administration Land Surface Data Assimilation System (CLDAS) and its high-resolution counterpart (HRCLDAS) provide gridded analysis datasets that help compensate for the sparsity of in situ marine wind observations. However, notable biases exist between their maximum wind speed estimates and observations, particularly over the Taiwan Strait, where the pronounced Venturi effect leads to systematic underestimation. This underestimation is more severe in CLDAS, limiting its applicability for meteorological services. To address the issue, this study proposes a maximum wind speed correction method based on the double fusion of in situ wind observations from automatic weather stations and buoy stations. The method employs an inverse distance weighting algorithm to interpolate wind speed from such stations to grid points within a region centered on the location of the maximum wind speed. The optimal influence radius is determined by balancing wind field errors and spatial smoothness. The method was applied to experimentally correct the maximum wind speeds in CLDAS and HRCLDAS across three subregions of the Taiwan Strait (the Mindong, Minzhong, and Minnan Fishing Grounds) during 2021–2023. Results indicate that the proposed correction method effectively mitigates the underestimations of maximum wind speeds in both datasets, with HRCLDAS showing a more pronounced improvement. Compared with CLDAS, HRCLDAS features a higher frequency of smaller optimal influence radii. After correction, the hourly mean absolute errors (MAEs) of maximum wind speed decreased by 70%–85% for CLDAS and 90%–95% for HRCLDAS. The reduction rates of MAEs exceed 60% in all months, exceeding 85% and 90% for CLDAS and HRCLDAS, respectively, from October to January. Spatially, the MAE distribution aligns with the coastline, decreasing from west to east, with MAEs near the Fujian coastal buoys reduced to below 1 m s⁻¹ after correction. The proposed correction method is effective for different types of strong wind events caused by Typhoon Doksuri (2023) and a typical cold tide from January 23 to 25, 2023.