Major natural hazards often induce cascading impacts changing geo-environment and human society. Understanding and assessing such cascading impacts require a systematical integration of causality, uncertainty, empirical physical domain knowledge, and multi-modal multi-fidelity data captured by satellites, in-situ sensors, and people in the disaster zone. This project provides a novel probabilistic AI framework to jointly model cascading disaster impacts through causality and enable automatic and flexible multi-hazard probabilistic inference from multi-modal, multi-sourced, and multi-resolution data. The models are deployed in a near-real-time disaster information platform that automatically retrieve remote sensing, social media, and in-situ sensing data, and provide large-scale disaster impact assessment and mapping immediately after disaster occurs global wise.
Rapid Cascading Disaster Multi-Impact Modeling and Mapping
Summary
Related Publications
- Li, X., Gao, S., Gao, R., & Xu, S. (2025). Causal spatially heterogeneous Bayesian networks with GPs and normalizing flows for seismic multi-hazard estimation. npj Natural Hazards, 2(1), 69.
- Li, X., and Xu, S. (2025). Scalable Variational Learning for Noisy-OR Bayesian Networks with Normalizing Flows for Complex Cascading Disaster Systems. npj Natural Hazards, 2(1), 30.
- Li, X., Yu, X., Burgi, P.B., Wald, D.J., Hu, X., and Xu, S. (2025). Rapid Building Damage Estimates from the M7.8 Turkey Earthquake sequence via Causality-informed Bayesian Inference from Satellite Imagery. Earthquake Spectra. p.87552930241290501
- Wang, C., Liu, Y., Zhang, X., Li, X., Paramygin, V., Sheng, P., Zhao, X. and Xu, S., (2024). Scalable and rapid building damage detection after hurricane Ian using causal Bayesian networks and InSAR imagery. International Journal of Disaster Risk Reduction, p.104371.
- Wang, C., Engler, D., Li, X., Hou, J., Wald, D.J., Jaiswal, K. and Xu, S., (2024). Near-real-time earthquake-induced fatality estimation using crowdsourced data and large-language models. International Journal of Disaster Risk Reduction, 111, p.104680.
- Yu, X., Song, Y., Li, X., Song, X., Fan, X., Wang, F., Xu, S. and Hu, X., (2024). Intelligent assessment of building damage of 2023 Turkey-Syria Earthquake by multiple remote sensing approaches. npj Natural Hazards, 1(1), p.3.
- Li, X., Bürgi, P.M., Ma, W., Noh, H.Y., Wald, D.J. and Xu, S., 2023. DisasterNet: Causal Bayesian Networks with Normalizing Flows for Cascading Hazards Estimation from Satellite Imagery. In: 29th ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD ’23)
- Li, X., Dimasaka, J., Zhang, X., Yu, X., Wang, C., Noh, H. Y.; Hu, X., Zhao, X. and Xu, S. (2023) “M7.8 Turkey-Syria Earthquake Impact Estimates from Near-real-time Crowdsourced and Remote Sensing Data.” DesignSafe-CI.
https://doi.org/10.17603/ds2-vnsc-y870 v2 - Xu, S., Dimasaka, J., Wald, D.J. and Noh, H.Y., (2022). Seismic Multi-hazard and Impact Estimation via Causal Inference from Satellite Imagery. Nature Communications, 13(1), pp.1-13.
- Xu, S. and Noh, H.Y., (2021). PhyMDAN: Physics-informed knowledge transfer between buildings for seismic damage diagnosis through adversarial learning. Mechanical Systems and Signal Processing, 151, p.107374
Funding Agencies
Collaborating Agencies
