Neighbourhood-Scale Typhoon Risk Mapping Using Machine Learning and SAR-Derived Impact Proxies: Cross-Event Evidence from Cebu City, Philippines

Typhoon impacts in the Philippines exhibit strong neighbour hood-scale heterogeneity driven by interactions among haz ard intensity, exposure, and vulnerability, yet most opera tional assessments remain hazard-centric. This study presents a grid-based machine-learning framework for neighbour hood-scale typhoon risk mapping integrating geophysical hazards, exposure and vulnerability indicators, and satellite derived impact proxies. A Random Forest classification model is implemented at 100 m resolution for Cebu City, us ing Sentinel-1 Synthetic Aperture Radar (SAR) backscatter change as an all-weather indicator of disturbance. Binary im pact labels are generated by applying a disturbance threshold (τ = 1.5) to SAR backscatter change, converting the disturb ance signal into a supervised classification target. Because SAR backscatter changes may reflect surface wetness, vege tation disturbance, or flooding effects, the label represents a disturbance proxy rather than a direct measure of structural damage. Predictor variables include ERA5-Land wind and rainfall metrics, population and building exposures, topogra phy, land cover, and proximity to coastal and riverine fea tures. Model performance is evaluated using the Area Under the Receiver Operating Characteristic Curve (AUC) through cross-event generalisation between Typhoon Rai (2021) and Typhoon Phanfone (2019). Results show above-random dis crimination (average AUC ≈ 0.59) and stronger transferabil ity in built-up areas (AUC ≈ 0.69), supporting neighbour hood-scale risk screening under realistic data constraints.