CrisiSense-RAG: Crisis sensing multimodal retrieval-augmented generation for rapid disaster impact assessment
- 1Zachry Department of Civil and Environmental Engineering, Texas A&M University, Texas, United States
- 2Department of Computer Science and Engineering, Texas A&M University, Texas, United States
Abstract
Timely and spatially resolved disaster impact assessment is essential for effective emergency response. However, automated methods typically struggle with temporal asynchrony. Real-time human reports capture peak hazard conditions, while high-resolution satellite imagery is frequently acquired after peak conditions. This often reflects flood recession rather than the maximum extent. Naive fusion of these misaligned streams can yield dangerous underestimates when post-event imagery overrides documented peak flooding. We present CrisiSense-RAG, a multimodal retrieval-augmented generation framework that reframes impact assessment as evidence synthesis over heterogeneous data sources without disaster-specific fine-tuning. The system employs hybrid dense-sparse retrieval for text sources and CLIP-based retrieval for aerial imagery. A split-pipeline architecture feeds into asynchronous fusion logic that prioritizes real-time social evidence for peak flood extent while treating imagery as persistent evidence of structural damage. Evaluated on Hurricane Harvey across 110 ZIP-code queries with multimodal coverage, the full multimodal configuration achieves a flood extent MAE of 8.86% to 19.64% and damage severity MAE of 10.14% to 14.47% across four model backends. Full multimodal RAG reduces flood extent MAE by up to 8.62 percentage points relative to text-only inference. Prompt-level alignment proves critical for quantitative validity: without explicit metric grounding, models interpret damage severity as qualitative intensity rather than the statistical average fraction of building value destroyed. These results demonstrate a practical and deployable approach to rapid resilience intelligence under real-world data constraints.
Related
- Sustainability assessment using multimodal artificial intelligence agents
- Neutrosophic reliability assessment of smart grid power generation system using universal generating function under Weibull distribution
- Radiative Transfer Modelling of Ka-band Microwave Interactions with Snow-Covered Freshwater Ice: Implications for SWOT KaRIn Property Retrieval
- BRONet v1.0: A machine learning approach for BrO profile retrieval using a full-physics inverse model
- Bayesian Retrieval of L-Band Absorption by Ice using SMOS Brightness Temperatures over the Antarctic Plateau
BibTeX
@article{xiao2026crisisenserag,
title = {{CrisiSense-RAG: Crisis sensing multimodal retrieval-augmented generation for rapid disaster impact assessment}},
author = {Yiming Xiao and Kai Yin and Ali Mostafavi},
journal = {Computer-Aided Civil and Infrastructure Engineering},
year = {2026},
month = may,
doi = {10.1016/j.cacaie.2026.100096},
url = {https://doi.org/10.1016/j.cacaie.2026.100096},
note = {CC BY},
}
RIS
TY - JOUR
TI - CrisiSense-RAG: Crisis sensing multimodal retrieval-augmented generation for rapid disaster impact assessment
AU - Yiming Xiao
AU - Kai Yin
AU - Ali Mostafavi
PY - 2026
DA - 2026/05/19
JO - Computer-Aided Civil and Infrastructure Engineering
DO - 10.1016/j.cacaie.2026.100096
UR - https://doi.org/10.1016/j.cacaie.2026.100096
AB - Timely and spatially resolved disaster impact assessment is essential for effective emergency response. However, automated methods typically struggle with temporal asynchrony. Real-time human reports capture peak hazard conditions, while high-resolution satellite imagery is frequently acquired after peak conditions. This often reflects flood recession rather than the maximum extent. Naive fusion of these misaligned streams can yield dangerous underestimates when post-event imagery overrides documented peak flooding. We present CrisiSense-RAG, a multimodal retrieval-augmented generation framework that reframes impact assessment as evidence synthesis over heterogeneous data sources without disaster-specific fine-tuning. The system employs hybrid dense-sparse retrieval for text sources and CLIP-based retrieval for aerial imagery. A split-pipeline architecture feeds into asynchronous fusion logic that prioritizes real-time social evidence for peak flood extent while treating imagery as persistent evidence of structural damage. Evaluated on Hurricane Harvey across 110 ZIP-code queries with multimodal coverage, the full multimodal configuration achieves a flood extent MAE of 8.86% to 19.64% and damage severity MAE of 10.14% to 14.47% across four model backends. Full multimodal RAG reduces flood extent MAE by up to 8.62 percentage points relative to text-only inference. Prompt-level alignment proves critical for quantitative validity: without explicit metric grounding, models interpret damage severity as qualitative intensity rather than the statistical average fraction of building value destroyed. These results demonstrate a practical and deployable approach to rapid resilience intelligence under real-world data constraints.
SN - 1093-9687
SN - 1467-8667
ER -