Blog

And although the global population grew by 18.6% from 2000 to 2015, the population in these regions outpaced this growth, increasing by 34.1% over the same period. This means that between 58 and 86 million people were affected by flooding in these areas over 15 years.

“It’s no surprise that floods will continue to increase,” says Beth Tellman, co-founder of flood mapping startup Cloud to Street and lead author of the study. “But what amazed me was that people moved to places where we have seen flooding in the past.”

Researchers looked at over 3,000 events in the Dartmouth Flood Observatory database, which records reported floods in the media. They matched the events, for which there was position data, with satellite images from MODIS, an instrument mounted on two NASA satellites, each of which has been capturing images of the Earth daily since 2000.

The researchers used an algorithm to map where the flood occurred, sorting which pixels were covered in water and which were not. They then added population data to see how trends in flooded areas changed over time.

Low- and middle-income countries have experienced the fastest population growth in flood-prone areas in the past two decades, with the highest growth rates in sub-Saharan Africa and South Asia.

Some of the movements can be attributed to socioeconomic factors, Thälmann said. Vulnerable groups may have no choice but to settle in floodplains where land can be cheaper and more accessible.

Using satellite imagery, the researchers were able to more accurately describe the effects of real floods than traditional models. Models can record some types of flooding, such as those that occur around rivers and along coastlines. But for others, triggered by heavy rainfall or random events such as dam breaks or storm surges that coincide with tide, satellite imagery provides a clearer picture.

The 913 mapped floods still represent only a fraction of the tens of thousands that occur worldwide each year. “This is just the tip of the iceberg,” says Thälmann.

MODIS captures images with a resolution of 250 meters, which is roughly the length of two football fields. This means that researchers cannot map small floods or floods in most cities. Clouds also interfered with the imaging algorithm, and since satellites passed over a specific location on Earth only once or twice a day, they also missed short-term floods.

“The new instruments are much higher resolution and can see through the clouds,” says Bessie Schwartz, co-founder and CEO of Cloud to Street. These tools, along with artificial intelligence, can now map floods even more accurately. But in order to systematically display floods over time, researchers had to stick to images from a single source using long-standing technology.

These efforts give scientists a clearer picture of the magnitude of recent floods and their anthropogenic impact than any other resource. “And the results will be especially useful for modelers trying to predict risk,” says Philip Ward, who studies flood risk assessment at Vrije Universiteit Amsterdam and was not involved in the study.