A research team in China may have just developed a better way to track and predict flooding.
The team of researchers at Hunan University, just a few hours south of Dongting Lake, which experienced historic flooding in 2020, believes they've found a new approach to flood monitoring using synthetic aperture radar. SAR is basically creating a model image of a single area using satellite imagery from multiple locations.
Historically, satellite data has faced challenges from something as simple as cloud cover. However, the team from Hunan University looked at SAR data from a yearlong time period to help identify specific flood patterns, including when flood waters normally rise and when they recede.
The team used this method of gathering data over a matter of space and time, along with using "fuzzy logic," to wind up at accurate conclusions. Fuzzy logic is a way of variable processing that considers all available information, including information that may only be partially true or relative in nature.
Xinxin Liu, co-author of the paper and associate professor of electrical and information engineering at Hunan University, said, "The proposed method has good consistency and stability in mapping the development of flood events. It can also identify flood regions that are prone to misidentification or omission thanks to time-series modeling."
Liu believes this method can be used to quickly and accurately map a flood as it's happening.
Liu and the team tested this method by applying it to data from the Dongting floods in 2020. The team found that their model was highly accurate and outperformed other models.
As the planet continues to warm, flooding has become more frequent. A warmer planet means more water vapor in the air, which means more rainfall that can potentially lead to flooding.
This method, developed by Liu and her team, can not only be used to map floods as they occur but also to help understand changing flood patterns over time.
"Because our method can reuse the historical data, we are also thinking about whether we can create some benchmark pattern of flood events based on historical data and apply them to the rapid warning of flood risk," Liu said. "This should be more interesting and meaningful."
Join our free newsletter for weekly updates on the coolest innovations improving our lives and saving our planet.