Spatio-temporal patterns of disaster impact and recovery in YouTube content

This study investigates the utilization of YouTube as a data source in crisis management, focusing on human behavior during a tornado event in the Southern Moravia Region, Czech Republic, in 2021. Utilizing YouTube's Data API, we harvested video metadata and content to explore spatio-temporal patterns in disaster impact and recovery. Our analysis reveals how YouTube content reflects the immediate and long-term responses to the disaster, offering insights into community reactions, damage extent, and recovery processes. By integrating automated and manual content analysis, we uniquely demonstrate YouTube's potential in enhancing spatial analyses and visualizations for disaster management. This research underscores the value of social media data in providing multiple perspectives on human behavior during and after disaster, contributing to a nuanced understanding of disaster impact and community resilience.