Harnessing Diverse Data for Global Disaster Prediction: A Multimodal Framework

TL;DR

This paper introduces a multimodal framework that combines weather data, satellite images, and text to improve global disaster prediction, focusing on floods and landslides amid climate change challenges.

Contribution

It proposes a novel multimodal approach integrating diverse data sources for disaster prediction, addressing class imbalance and demonstrating variable performance improvements.

Findings

Multimodal data integration enhances disaster prediction accuracy.

Performance gains vary depending on disaster type.

Addressing class imbalance improves model reliability.

Abstract

As climate change intensifies, the urgency for accurate global-scale disaster predictions grows. This research presents a novel multimodal disaster prediction framework, combining weather statistics, satellite imagery, and textual insights. We particularly focus on "flood" and "landslide" predictions, given their ties to meteorological and topographical factors. The model is meticulously crafted based on the available data and we also implement strategies to address class imbalance. While our findings suggest that integrating multiple data sources can bolster model performance, the extent of enhancement differs based on the specific nature of each disaster and their unique underlying causes.