Abstract: The China Meteorological Administration Land Surface Data Assimilation System (CLDAS) and the High Resolution CLDAS (HRCLDAS) provide gridded analysis datasets that compensate for the sparsity of in situ wind observations over sea. However, bias exists between their maximum wind speed estimates and observations, particularly in the Taiwan Strait region where the venturi effect is pronounced, leading to underestimations of maximum wind speeds. Notably, CLDAS exhibits a more severe underestimation, which may not satisfy the meteorological service requirements. To address the issue, this paper proposes a maximum wind speed correcting method based on the doubled fusion of wind speed in situ observations from automatic weather stations and buoy stations. It employs the inverse distance weighting (IDW) algorithm to interpolate the wind speed observations of stations to the grid points within the region centered by the location of the maximum speed, which optimal influence radius is determined by considering both wind field errors and spatial smoothness. By using this method, an experiment was conducted to correct the maximum wind speeds of the CLDAS and HRCLDAS in different sub - regions (Mindong, Minzhong and Minnan Fishing Grounds) of the Taiwan Strait region during 2021-2023. The results indicate that the correction method can effectively mitigates the underestimations of maximum wind speeds of gridded analysis data in the Taiwan Strait region for both CLDAS and HRCLDAS, with HRCLDAS showing a more significant improvement. Compared with CLDAS, the frequency of the optimal influence radius for small values increases in HRCLDAS. After correction, the hourly mean absolute errors (MAEs) for the CLDAS and HRCLDAS are mostly decreased by 70-85% and 90-95%. The reduction rates of MAEs exceed 60% in different months, especially exceeding 85% and 90% for CLDAS and HRCLDAS from October to January, respectively. Additionally, the spatial distribution of MAEs is parallel to the coastline, decreasing from west to east, with MAEs near the buoys along the Fujian coast reduced to below 1 m s-1 after correction. Furthermore, it is demonstrated that the correcting method is effective for different kinds of strong wind caused by Typhoon Doksuri (2023) and a typical cold tide occurred from January 23rd to 25th, 2023.