Compound patterns and environmental drivers of typhoon disaster chains in southeast coastal China: a multiscale framework

Typhoons frequently trigger interacting hazards, including river flooding, urban waterlogging, landslides, and storm surge, but disaster records compiled at the event or administrative-unit level limit systematic analysis of their spatial co-occurrence and environmental differentiation. Here, we developed a multiscale framework to quantify chain-specific disaster-forming-environment (DFE) sensitivity for four predefined typhoon disaster chains: typhoon–rainstorm–urban waterlogging (TRU), typhoon–rainstorm–flooding (TRF), typhoon–rainstorm–landslide (TRL), and typhoon–wind–storm surge (TWS). The results revealed a dual spatial pattern, with widespread landslide-related sensitivity in the mountainous interior and localized flood-, waterlogging-, and storm-surge-related sensitivity in coastal areas. TRL had the largest high-sensitivity area, covering 31.6% of Fujian Province, whereas TRU had the smallest spatial extent but was strongly concentrated in coastal urban districts. TRF and TRU exhibited the strongest positive spatial dependence in urban areas (Kendall's τ = 0.834). Areas highly sensitive to at least one chain covered 42.0% of the province and comprised two broad patterns and nine specific types. Single-chain and compound types accounted for 93.1% and 6.9% of this area, respectively, and TRF + TWS was the most extensive compound type, representing 69.4% of the compound-type area. The spatial organization of these types was scale dependent: TRL became increasingly predominant from county to basin scales, indicating that broader-scale aggregation can obscure localized compound-type heterogeneity. Green-space proportion was the leading environmental discriminator for most flood- and coastal-related types, whereas elevation most strongly differentiated TRL. Types containing a TRU component were generally associated with greater impervious-surface coverage and less green space. These findings reveal contrasting inland and coastal patterns of typhoon disaster-chain sensitivity and provide a basis for multiscale hotspot identification and type-specific disaster-chain management.