Abstract: With the vigorous growth of the shipping industry,ensuring navigational safety and improving operational efficiency have imposed higher demands on accurate wave simulations under extreme sea states. However,the applicability of the third-generation spectral wave model WAVEWATCH Ⅲ(WW3) and the identification of optimal simulation configurations during the rapid intensification stage of super typhoons remain insufficient,limiting its ability to address two key challenges in typhoon-related maritime early warning: accurate simulation of typhoon-generated waves during the rapid intensification stage and precise classification of wave-induced disaster risk. To tackle these issues,a multi-dimensional sensitivity analysis framework is employed. First,typhoon characteristics and ship-accident information are compiled using the Northwest Pacific tropical cyclone best-track dataset provided by the China Meteorological Administration(CMA) and vessel accident investigation reports from the Guangdong Maritime Safety Administration. Second,high-resolution typhoon-wave simulation experiments are established using WW3. For source-term parameterizations,three widely used WW3 schemes(ST2,ST4,and ST6) are adopted. For atmospheric forcing,wind fields from the Weather Research and Forecasting(WRF) model,the fifth-generation ECMWF global atmospheric reanalysis(ERA5),and the NCEP Final Analysis(NCEP-FNL) are used to examine WW3 performance and to determine the optimal simulation configuration.Combined with observations from National Oceanic Administration buoys and standard statistical validation metrics,a complete technical workflow is developed. Under rapid intensification conditions of super typhoons,WW3 experiments with different source-term schemes and wind forcing fields are conducted and evaluated against buoy measurements. The results indicate that simulations driven by WRF winds and using the ST2 scheme achieve superior performance near the typhoon center,whereas the ST4 and ST6 schemes perform better in other regions.