DisasterNet: Causal Bayesian networks with normalizing flows for cascading hazards estimation from satellite imagery
Citation
Li, X., Bürgi, P. M., Ma, W., Noh, H. Y., Wald, D. J., Xu, S. (2023). DisasterNet: Causal Bayesian networks with normalizing flows for cascading hazards estimation from satellite imagery. In Proceedings of the 29th ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD '23), pp. 4391–4403. Link to paper
Abstract
Sudden-onset hazards like earthquakes often induce cascading secondary hazards (e.g., landslides, liquefaction, debris flows) and subsequent impacts (e.g., building and infrastructure damage) that cause catastrophic human and economic losses. Rapid and accurate estimates of these hazards and impacts are critical for timely and effective post-disaster responses. However, hazards and damage often co-occur or colocate with underlying complex cascading geophysical processes, making it challenging to directly differentiate multiple hazards and impacts from satellite imagery using existing single-hazard models. This study presents DisasterNet, a novel family of causal Bayesian networks to model processes that a major hazard triggers cascading hazards and impacts. The framework integrates normalizing flows to effectively model the highly complex causal dependencies in this cascading process. A triplet loss is further designed to leverage prior geophysical knowledge to enhance the identifiability of the highly expressive Bayesian networks. A novel stochastic variational inference with normalizing flows is derived to jointly approximate posteriors of multiple unobserved hazards and impacts from noisy remote sensing observations. Integrating with the USGS Prompt Assessment of Global Earthquakes for Response (PAGER) system, the framework was evaluated in recent global earthquake events, showing that DisasterNet significantly improves multiple hazard and impact estimation compared to existing USGS products.