A gray funnel cloud touching Earth’s surface
A tornado touches down in Manitoba, Canada. Credit: Justin Hobson/Wikimedia, CC BY-SA 3.0

In the past 70 years, more than 75,000 tornadoes have been recorded in the United States. Recordkeeping of these phenomena outside this region has been largely fragmented, sitting isolated in books, government databases, and research archives. But a new effort to scour as many publicly accessible records as possible is highlighting the scale of this hazard around the world.

In a new study, Malcolm Maas, an undergraduate student at the University of Maryland, College Park, and a team of tornado researchers compiled a tornado database that they hope will boost tornado research globally.

“The most developed countries have permanent, usually governmental, organizations tasked with compiling records of tornado occurrence, and thus have the most thorough datasets,” Maas said. “Datasets for other countries mostly come from independent researchers who are limited to putting together reports from newspapers and websites.”

“The tricky part is that every single dataset keeps track of this information in a slightly different way.”

The first challenge the group faced was that vast amounts of U.S. tornado data from prior to the creation of the National Weather Service in 1870 weren’t available in a digital format. The researchers found records in a 1993 book by meteorologist Thomas P. Grazulis, Significant Tornadoes, 1680–1991. “There was a big effort involving a lot of people to go through the book and put it into a format that’s accessible by geographic information systems,” Maas said. Some of the tornado locations were recorded as descriptions such as “five miles north of this town,” he said. “But the town didn’t exist anymore, so we had to pull out old maps to find it.” More than a dozen people worked on the project at any one time, describing about 7,000 tornadoes in total.

For records outside the United States, the team downloaded existing databases from the Internet. Sometimes tracking down a database instead involved sending someone an email to get their spreadsheet or accessing a Ph.D. thesis, Maas said.

“The tricky part is that every single dataset keeps track of this information in a slightly different way,” Maas said. “This is the most challenging part, because you have to massage everything to get it to come together.”

Nascent Research

Outside the United States, tornado research is only recently picking up speed. Records in these regions are harder to come by, and populations generally have less knowledge of the phenomenon.

In May 2019, seven tornadoes hit near cities in southern Chile, spurring several research groups to begin studying what atmospheric ingredients give rise to tornadoes there, explained meteorological researcher Julio César Marín Aguado of the Universidad de Valparaíso in Chile. “We now know more about the synoptic conditions and some mesoscale characteristics associated with these events, but there are still many aspects to be investigated,” he said. A global, unified database could help fill in vital gaps.

A similar genesis of tornado research has happened in East Africa. Sosten Chiotha, regional director for the nongovernmental organization LEAD International and an environmental scientist based in Zomba, Malawi, published the country’s first paper about local tornadoes after an automated weather station installed at his office picked up strange readings in 2017. South Africa has a long track record of publishing tornado research, but Malawi has limited knowledge and capacity to systematically track and document tornadoes, he said.

“The first picture of the tornado that we published was captured by someone who thought it was smoke from burning tires and took the picture out of curiosity,” Chiotha said. The hope is that observations from more regions, included in a global database, would help provide a more complete picture of tornado occurrence across the world.

Fatal Tornadoes

The United States accounts for 21,548 of the recorded fatalities in the database published by Maas and his collaborators. But tornadoes in other countries wreak havoc as well: Bangladesh accounts for 8,325 fatalities in the database, India has seen 1,473, and the rest of the world combined accounts for 3,824. “The frequency, intensity, and lifespan of tornadoes in South Asia are generally less severe than those in the United States; however, due to the dense population in these countries, even low-intensity, short-lived tornadoes can result in significant deaths and damages,” said meteorologist Nasreen Akter at Bangladesh University of Engineering and Technology.

Diagram showing density of tornado fatalities across the world
The global distribution of total recorded tornado fatalities per 10,000 square kilometers is shown here. Data are not available for South African and most Chinese tornadoes. Credit: Maas et al., 2024, https://doi.org/10.1175/BAMS-D-23-0123.1

“Other countries may turn into hot spots due to increasing global temperature, so credible data is critical to reduce vulnerability of people to tornadoes.”

And that impact could shift in the future. Ashraf Dewan, an environmental geographer from Bangladesh now at Curtin University in Australia, said that although the United States is currently a hot spot, the climate is changing. “Other countries may turn into hot spots due to increasing global temperature, so credible data is critical to reduce vulnerability of people to tornadoes,” he said. The researchers hope the new database can be used to collate more data from the Global South, especially where government agencies don’t have the capacity to collect such data.

Subhash Chander Bhan, a recently retired agrometeorologist and forecaster with the India Meteorological Department, agreed with the new study’s methodology but said the next big challenge is to incorporate information not available in official records or existing research.

“I know about quite a few reports of tornadoes in northwest India in various electronic platforms which have not been systematically documented,” he said. “Collating all available information and its digitization, in terms of exact time, coordinates, path, and damage, would certainly help researchers work further on climatological, synoptic, dynamic, and thermodynamic aspects; and for incorporating the knowledge in early warning systems.”

Maas acknowledged that in a relatively small dataset, any trends might turn out to be statistical artifacts; he added that his dream would be a central international organization that keeps track of tornadoes.

—Andrew J. Wight, Science Writer

Citation: Wight, A. J. (2025), A new tornado database helps researchers worldwide, Eos, 106, https://doi.org/10.1029/2025EO250005. Published on 3 January 2025.
Text © 2025. The authors. CC BY-NC-ND 3.0
Except where otherwise noted, images are subject to copyright. Any reuse without express permission from the copyright owner is prohibited.