<p>Typhoon-induced storm surges and extreme waves are important marine hazards in coastal regions, and their evolution is highly sensitive to wind-field forcing. Differences in wind-field construction methods can significantly affect storm-surge and wave responses during typhoon events, but the mechanisms by which multi-source wind-field fusion influences regional responses remain insufficiently understood. Taking Typhoon In-fa (2021) as a case study, this study carried out comparative experiments based on the ADCIRC-SWAN coupled model using ERA5, WRF, Holland, and two fused wind fields. The model performance was comprehensively validated using tide-gauge water-level observations, storm-surge series derived by subtracting the corresponding astronomical tide from measured water-level data, scatterometer wind speeds, and satellite altimeter wave-height data. On this basis, difference fields, spatial clustering, correlation analysis, and EOF decomposition were employed to systematically analyze, at the regional scale, the effects of different wind-field schemes on wind-field structure, wave response, and storm-surge response. The results show that the fused wind fields are closer to the parametric wind field in the typhoon core region and closer to the background wind field in the outer region, exhibiting differentiated reconstruction in different regions, which is further transmitted to wave and storm-surge responses. In addition, this study proposes and implements a multi-source wind-field fusion approach that includes typhoon-center matching, track-direction alignment, and transition treatment, and systematically reveals the propagation process of wind-field structural differences from the forcing field to the response field, thereby providing a reference for subsequent wind-field scheme selection, fusion-strategy design, parameterization improvement, and coupled simulation of storm surges and waves.</p>