Monday, 13 January 2025: 8:30 AM
347 (New Orleans Ernest N. Morial Convention Center)
While an average of 88 people die from flooding every year in the United States, we have a poor understanding of the types of flood situations and flood-related decisions that put people at risk of dying. A key step in identifying these situations is understanding patterns of recent flood mortality and how they have changed over time. To address this gap, we compiled US flood fatality data from the Storm Events Database for the years 1996-2023. We focus on inland flooding categorized in the Storm Events Database as “Flood” and “Flash Flood” and identify what people were doing during these flood fatality events by labeling the records with additional categories such as those in vehicles versus those on foot, and whether roads were closed at time of the fatality. In this study we ask, “How are deaths changing over time and to what extent?”, “What are some key similarities and differences between flash floods and other inland flooding outcomes?”, and “What are the activities or settings where these fatalities occur?”. We found a high variation of fatalities both in time and in region, suggesting caution for those looking for short or long-term trends. In addition, the national aggregation of fatalities misses key differences in both region and season, raising questions about its usefulness as a measure. We also found that flash floods have higher fatalities than other inland flooding events and that only 78% of recorded fatality events had a known circumstance; of these, 58% were vehicle related, suggesting that focusing on preventing vehicle related deaths could have a large impact on reducing total fatalities. We provide some further recommendations and caveats for using Storm Events data for tracking and investigating flood mortality.
- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner