Multi-Hazard and Loss Event Sets for the Contiguous US
This study utilizes a customizable Python tool to analyze NOAA NCEI data from 1996 to 2024, revealing that multi-hazard events and their associated human and economic losses in the contiguous US predominantly occur during spring and summer, with significant spatial hotspots in the Southwest, Central US, and coastal regions, while highlighting the sensitivity of multi-hazard definitions to the chosen temporal window.
Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer
Disasters rarely happen in isolation. A single storm can bring heavy rain, high winds, and flooding all at once, while a long period of dry weather can set the stage for a wildfire that later turns rain into a mudslide. Scientists call these overlapping threats "multi-hazards." For decades, researchers and emergency planners have relied on databases that list disasters one by one, treating a tornado or a flood as a separate event. While this approach helps track the frequency of individual dangers, it misses the complex reality where one hazard triggers another or where several strike the same place at the same time. Understanding these connections is vital because the damage caused by interacting hazards can be far greater than the sum of their parts. To move from simply counting disasters to understanding how they combine, researchers need a way to map these overlapping events and the losses they cause.
A team of researchers has now built a new tool to do exactly that for the United States. They created a customizable system that sifts through decades of weather records to find moments when different types of hazards hit the same county within a specific window of time. Using data from the National Oceanic and Atmospheric Administration's Storm Database, which contains records of weather events going back to 1950, the team focused on the period between 1996 and 2024. They cleaned up the raw data, filling in missing details where possible and standardizing how different types of storms are named, to create a reliable set of records. They then defined what counts as a "multi-hazard" event by looking for instances where two or more different hazards, such as a tornado and a flash flood, occurred in the same county within ten, thirty, or ninety days of each other. This approach allowed them to generate a detailed map of where these dangerous combinations happen and how much they cost in terms of human injury and financial loss.
The results reveal a clear pattern: most of these multi-hazard events happen in the spring and summer. This timing aligns with the seasons when severe thunderstorms, tornadoes, and tropical cyclones are most active across the country. The researchers found that the most frequent combinations involve thunderstorm winds paired with hail, flash floods, or tornadoes. These dangerous pairings are not spread evenly across the map. Instead, they cluster in specific hotspots. The Southwest, particularly the area where Arizona, Nevada, and California meet, sees a high number of these events. The central United States, including parts of Nebraska, Kansas, and Indiana, is another major zone where thunderstorms and tornadoes frequently overlap. Along the Gulf Coast and the East Coast, the pattern shifts to include tropical cyclones, which often bring a mix of storm surge, heavy rain, and high winds to the same communities.
When the team looked at the human and financial toll, the picture became even more specific. The number of deaths and injuries from these events tracks closely with where people live; the most populated urban areas see the highest number of casualties simply because more people are exposed. However, the financial damage tells a slightly different story. While property losses are high in coastal regions and along the Great Lakes, the most significant crop losses occur in the agricultural heartland of the Midwest, particularly in Nebraska, Kansas, and Indiana. The study also highlighted a limitation in how we currently understand these risks. The researchers noted that their method of looking at events within a few days or weeks captures immediate combinations well, but it might miss slower-moving chains of events. For example, a drought that lasts for years can dry out vegetation, making a later wildfire much worse, or a wildfire can strip away soil stability, leading to floods months later. These longer-term connections require a different kind of analysis than the one used here.
Despite these limitations, the new dataset offers a significant step forward. It provides the first detailed, county-level view of multi-hazard interactions in the United States that includes specific loss metrics. By showing exactly where and when these overlapping threats occur, the work gives emergency managers and financial institutions a clearer picture of risk. Instead of preparing for a single type of disaster, communities can now see where they are most likely to face a complex mix of dangers. This shift in perspective allows for better planning, ensuring that resources are directed to the places where the combination of hazards poses the greatest threat to life and property. The tool itself is open for others to use, meaning researchers can now customize the search to look at specific types of hazards or different time windows, further refining our understanding of how nature's dangers interact.
Drowning in papers in your field?
Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.