Assessment of offshore island wind energy potential in typhoon-prone regions with a KDE-based probabilistic modeling approach

Accurate assessment of offshore wind energy potential is critical for enhancing renewable power generation and ensuring the reliability of islanded energy systems in extreme weather-prone regions. In typhoon-prone coastal and island areas of East Asia, tropical cyclones can significantly alter short-term wind speed distributions, affecting turbine performance and system operation. This study develops a comprehensive wind resource assessment framework that explicitly incorporates typhoon disturbances. Building upon modelling techniques established in meteorology and wind engineering, the framework reconstructs and tailors typhoon environment simulation methods for the context of offshore wind resource evaluation, enabling the probabilistic representation of typhoon–wind interactions. A virtual typhoon trajectory generation model based on kernel density estimation (KDE) is used to simulate probable typhoon paths, which are then coupled with local background wind conditions to generate synthetic wind fields. Model accuracy is validated through spatial similarity analysis using Kullback–Leibler (KL) divergence against historical typhoon tracks. Power output is estimated for both large- and small-scale turbines under synthetic typhoon scenarios. A case study shows that typhoon winds exert a dual influence on turbine performance—slightly reducing the total generation of the large-scale Vestas V236 turbine (−0.38 %) due to frequent cut-out shutdowns, while modestly increasing that of the small-scale AH-60K turbine (+8.4 %) owing to its better adaptability under fluctuating winds. The proposed method offers a practical and transferable tool for wind farm siting, turbine type selection, and the planning of resilient offshore renewable energy systems in extreme weather-prone regions.