Although the global population has increased by 18.6% from 2000 to 2015, the population growth rate in these regions has exceeded this growth by 34.1% over the same period. This means that between 58 and 86 million people in these places will face floods in 15 years.
“It’s not particularly surprising that flooding will increase,” said Beth Tellman, The co-founder of a flood surveying start-up company Cloud to street And the main author of the study. “But what shocks me is that people are moving into places where we have observed flooding in the past.”
Researchers observed more than 3,000 events Dartmouth Flood Observatory A database that records floods reported in media reports.They associate events with location data from Fashion, An instrument mounted on two NASA satellites, which have taken images of the Earth every day since 2000.
The researchers used an algorithm to map out where the flood occurred by sorting out which pixels are covered by water and which are not. Then they added population data to see how trends in flooded areas changed over time.
Over the past 20 years, flood-prone areas in low- and middle-income countries have experienced the fastest population growth, and sub-Saharan Africa and South Asia have the highest population growth rates.
Terman said that socioeconomic factors may explain some of the sports. Vulnerable groups may have no choice but to settle in flooded areas, where land may be cheaper and easier to obtain.
By using satellite imagery, researchers can describe the effects of real floods more accurately than traditional models. The model can capture certain types of floods, such as those occurring near rivers and coasts. But for other situations caused by heavy rainfall or random events (such as dam rupture or storm surges aligned with high tides), satellite imagery provides a clearer picture.
The 913 floods mapped are still only a small part of the tens of thousands of floods that occur every year around the world. “This is just the tip of the iceberg,” Terman said.
MODIS captures images at a resolution of 250 meters, about the length of two football fields. This means that researchers cannot map smaller floods or floods in most cities. Clouds also interfere with image processing algorithms, and because satellites only pass through a specific spot on the earth once or twice a day, they also miss short-term floods.
Newer instruments have higher resolution, which can be seen through the clouds, saying Bessie Black, Co-founder and CEO of Cloud to Street. Together with artificial intelligence, these tools can map today’s floods more accurately. But in order to systematically map the flood over time, researchers must insist on using images from one source, using existing technology.
This effort allows scientists to understand the scale and human impact of recent floods more clearly than any other source.The results are particularly useful for modelers trying to predict risk, saying Philip WardHe studied flood risk assessment at Vrije Universiteit Amsterdam, but did not participate in this research.