Spatial distribution law analysis of geological disasters induced by the “25·7” extreme rainfall in the Beijing–Tianjin–Hebei region

Abstract

In July 2025, the Beijing–Tianjin–Hebei region was hit by an unprecedented extreme torrential rain, which triggered extensive clustered geological disasters. To reveal the spatial distribution pattern and dominant controlling factors of regional landslides induced by this extreme rainfall, this study establishes a landslide inventory database based on remote sensing interpretation and field investigations. On this basis, GIS spatial analysis is integrated with the random forest model to quantitatively assess the contribution of nine predisposing factors to landslide occurrence. The results show that a total of 15,308 landslides were triggered by the extreme rainstorm, exhibiting an obvious spatially clustered distribution pattern. High-landslide-susceptibility areas are mainly concentrated in zones with cumulative rainfall of 250–300 mm, elevation of 400–600 m, slope gradient of 10–15°, concave slope morphology, and areas close to river valleys. Factor importance analysis further indicates that cumulative rainfall acts as the primary dominant factor, while elevation, lithology, and slope gradient serve as critical synergistic controlling factors. This study clarifies the unique formation mechanism of extreme rainfall-induced landslides in the semiarid region of North China and can provide a scientific reference for meteorological early warning of regional geological disasters as well as disaster prevention and mitigation decision-making.