“People die and lose their homes and livelihood,” said Beth Tellman, a human-environmental geographer at the University of Arizona and lead author of the paper. The study estimates that up to 86 million people have moved into places where there have been floods, which Tellman says is only “the tip of the iceberg.”
Tellman and her colleagues collected satellite imagery of 913 large flood events around the world, from 2000 and 2018. The database included floods from rivers, tropical storms, melting ice and snow, as well as dam or levee breaks. Researchers then compared the population of the flooded areas between 2000 to 2015.
The change in population in flood zones varied by location. In Russia and Sri Lanka, for instance, the number of people living in those areas shrank. Jamaica stayed about the same. But many places, such as Bangladesh and India, saw large increases — of up to 14.3 million and 44.8 million people, respectively.
Overall, the paper found that the population of flooded areas grew at nearly twice the average global rate. Proportionally, as much as 24 percent more people lived in those areas by 2015. Previous estimates, Tellman said, put the number closer to about 2 percent.
“If people are moving into places that flood, it’s going to flood again,” said Tellman, noting the areas they examined flooded an average of about three times since 2000. That, she said, can lead to not only loss of life and property, but longer-term setbacks in economic development. “It’s only a matter of time.”
The study did not look at why people might move to flood-prone areas — though Tellman said the reasons could range from migration to urban centers near water to the affordability of land in flood-prone areas. It did find that the issue was particularly pronounced in areas of the global south, such as sub-Saharan Africa and Southeast Asia, which Tellman says are seeing large population growth or increased flooding risk due to climate change, or both.
“This is a back-burner issue that people don’t really talk about,” said Jason Thistlethwaite, a professor studying climate risks and extreme weather at the University of Waterloo in Canada. “Studies like this shed light on just how really urgent the situation is.”
With climate change, migration and population growth, Tellman said, the number of people living in flood zones is on track to increase. Using climate models from the World Resources Institute, the study estimated that between 2010 and 2030, 57 of 106 countries examined will see such a jump. That would potentially expose as many as 179 million additional people to flooding.
“We can’t be complacent about flood risk,” said Hannah Cloke, a professor of hydrology at the University of Reading in the United Kingdom. “We’re not acting fast enough.”
If anything, Cloke and other experts say, the paper is underestimating the number of people with potential flood exposure.
The satellite instrument that the scientists used — called a Moderate Resolution Imaging Spectroradiometer, or MODIS — takes images twice a day. But it does not work through clouds, Tellman said, which often accompany flooding, and it does not necessarily pick up floods that happen on a shorter time scale, such as the flash flooding that recently hit Europe. The 250-meter resolution of the imagery also makes it harder to see urban flooding.
In total, Tellman said, there were more than 2,000 recorded flood events that the researchers didn’t have images for. That means many more people may be exposed to floods than this study captured.
Some see these satellite limitations as problematic for the paper’s findings.
“You compound all those problems, and you struggle to see if you can trust the conclusions they are putting out,” said Oliver Wing, chief research officer at the flood modeling company Fathom. He added that advances in satellite and other technologies have made better modeling data available than what is found in the study. “It’s probably not something I would use as a flood modeler myself.”
But others said they would use the data set, which is publicly available online. “This gives us some real data to check our models against,” said Phillip Ward, a professor of global water risk dynamics at Vrije Universiteit in Amsterdam.
Roxy Mathew Koll is a climatologist at the Indian Institute of Tropical Meteorology. He said he found the study useful because climate change models don’t always take into account human impacts. Calling the work a “wake-up call,” he said global-level analysis like this can help point to where more precise, localized modeling is most needed.
The data could also be useful in the United States. “Our current regulations are based on historical data that’s no longer accurate,” said Sandra Knight, an engineer working on water policy at the University of Maryland. A recent U.S. Government Accountability Office report found that most U.S. flood insurance maps “do not reflect how climate change may affect flood risk.”
Tellman hopes scientists will use her work to improve their flood models and has been in touch with insurance companies that have also expressed interest in the data. She said she wants to continue to improve on the study by adding more detailed satellite, or radar, data.
But, ultimately, she’d like to see improvements to public policy that take into account the impact of flooding on populations.
“The countries we expect that have struggles with flooding [should] have the financial support to continue to adapt,” she said, noting that isn’t always the case. The Central African Republic, for instance, had the highest expected increase in proportion of population exposed to floods in the study. But it’s relatively low on the list of recipients of international climate adaptation funds.
Tellman said many other tools — including zoning restrictions and managed retreat planning — could also be used to address the growing proportion of people living in flood zones.
“We can reverse these trends. It’s not inevitable,” she said. “Those are human decisions.”