Answers are grounded in this paper and cite their sources.
Extreme rainfall can trigger widespread shallow landsliding, yet accumulated rainfall alone does not adequately explain its spatial concentration. We examined 15,622 landslides mapped following Typhoon Doksuri in Laishui County, northern China, of which 15,387 were retained for modelling. A RUSLE-derived event erosion proxy (EEP) was constructed by combining 10-day accumulated rainfall, soil erodibility, slope length and steepness, and land-cover factors. Its log-transformed form, EEP′ = ln (EEP + 1), showed a clearer positive association with landslide number density and area percentage than accumulated rainfall alone. Among four logistic regression models, the model incorporating both rainfall and EEP′ alongside conventional conditioning factors achieved the best discrimination and fit (five-fold AUC = 0.7736; AIC = 33,938.81). Spatial-block validation yielded the same model ranking, with the combined model achieving the highest AUC (0.7489). Component-wise comparisons indicated additional improvement in model fit, while random forest analysis identified EEP′ as an influential predictor. These findings support the use of EEP′ as a readily derived, interpretable indicator for characterising spatial differences in shallow-landslide occurrence and supporting regional screening after extreme rainfall. Its broader applicability remains to be tested across other events and regions.
No takes yet. Share an insight, caveat, or question.
Wu et al. (2026) studied this question.
Synapse has enriched 8 closely related papers on similar clinical questions. Consider them for comparative context: