Rain-induced landslides pose a growing threat to communities in South Africa’s high-rainfall regions, particularly in steep, densely settled and hydrologically complex catchments. The aim of this study was to assess the spatio-temporal response of slope stability to rainfall events using a coupled hydrological–slope stability model. The April 2022 floods in the Durban/Pinetown area (catchment U60F) served as a case study for evaluating the predictive capability of the factor of safety (Fs) under different thresholds. The calculated Fs values reflected the impact of rising soil moisture on slope instability, with model outputs highlighting increasing landslide susceptibility in response to cumulative rainfall. The model’s performance was assessed using confusion matrices, receiver operator characteristic curves and area under the curve values. While the traditional Fs < 1 threshold yielded a moderate specificity (0.67) but low sensitivity (0.38), an optimised threshold of Fs = 1.43 improved sensitivity to 0.78, although with a trade-off in precision and specificity. The Fs-based model captures broad spatial patterns of landslide susceptibility, but its predictive power remains moderate (area under the curve = 0.63) and is limited by the absence of infrastructural and anthropogenic triggers, such as road culverts, terracing and retaining walls, in the modelling framework. Nonetheless, when calibrated appropriately and integrated into early warning systems, Fs mapping provides valuable insights into landslide risk during extreme rainfall events.
Tol et al. (Tue,) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: