Problem
Research questions and friction points this paper is trying to address.
sinkhole susceptibility
survival analysis
multimodal data
right-censored data
geohazard prediction
Multimodal Deep Survival Analysis for Sinkhole Susceptibility
🤖 AI Summary
This study addresses the challenges of missing negative samples and multi-source heterogeneous data integration in karst collapse prediction by proposing a multimodal Cox proportional hazards model. The framework employs dedicated encoders and a cross-modal fusion layer to process heterogeneous inputs, while modeling unreported areas as right-censored data to circumvent hard negative labeling, thereby enabling time-aware continuous risk assessment. Combined with spatial block validation, the model effectively ranks collapse risks and generates high-resolution statewide susceptibility maps in a Florida case study, accurately capturing spatial heterogeneity. This work provides a novel paradigm for dynamic early warning of geological hazards.
📝 Abstract
Sinkholes are a widespread geohazard in karst terrain. In Florida, soluble carbonate bedrock, shallow groundwater, and intense rainfall combine to make subsidence both common and spatially heterogeneous. Predicting where and when sinkholes will occur is difficult for two reasons. First, locations without reported sinkholes cannot be directly labeled or sampled as true negative locations. Second, the potential factors governing sinkhole risk span heterogeneous data modalities and therefore require careful integration within a unified modeling framework. We address both problems with our proposed model, a multimodal Cox proportional hazards framework for sinkhole susceptibility. Our contributions are threefold. First, we extend the proportional-hazards formulation to heterogeneous multimodal input through modality-specific encoders and a cross-modal fusion layer. Second, we treat unreported locations as right-censored rather than negative, avoiding hard-negative labeling and yielding continuous, time-aware susceptibility from the predicted survival function. Third, a statewide Florida case study with spatially blocked validation and ablation studies quantifies the benefit of multimodal integration. A Florida case study demonstrates that the proposed method effectively ranks sinkhole risk and produces a statewide susceptibility map that captures spatial variations in sinkhole occurrence.
Innovation
Methods, ideas, or system contributions that make the work stand out.
Multimodal Deep Survival Analysis
Cox Proportional Hazards
Cross-modal Fusion
Right-censored Data
Sinkhole Susceptibility
🔎 Similar Papers
No similar papers found.
