Abstract
Abstract Landslides, causing significant loss of life and property, are traditionally attributed to heavy rainfall and geological or topographical factors, often overlooking strong winds. In landslide-prone forest regions, typhoon-induced strong winds can uproot trees and destabilize their root systems, potentially triggering shallow landslides. In this study, a data-driven approach revealed that strong wind is one of the important factors that may trigger landslides, particularly strong wind lasting for hours. The significance of the combined rain–wind influence on landslides is examined using Mann–Whitney U test and 3D histogram, and a Random Forest machine learning model is constructed to predict the occurrences of landslides based on factors, such as heavy rain, strong winds, traditional geological conditions, and topographical factors. The models with wind factors have better performance than models without wind factors. These findings infer that both strong winds and heavy rainfall during typhoons may trigger or increase the risk of landslides.