HURRICANESTracking Flooding in Coastal Communities During Hurricanes Helene and Milton
A web-based application that gathers crowdsourced data to identify flooding and inform policy in coastal communities provided scientists with essential data from hurricanes Helene and Milton.
A web-based application that gathers crowdsourced data to identify flooding and inform policy in coastal communities provided scientists with essential data from hurricanes Helene and Milton.
The team of researchers led by USF St. Petersburg GIS and Remote Sensing Professor Barnali Dixon used the CRIS-HAZARD app to analyze real-time flooding in Pinellas County, which is home to 588 miles of coastline in Florida. The app also received data from users in neighboring Hillsborough County during Hurricane Milton, which allowed the team to expand its research.
The app, developed with researchers at Georgia Tech, combines volunteered geographic information and community crowdsourced data such as photos and videos with near real-time data on flooding. Using dynamic modeling and mapping tools, including artificial intelligence (AI), the app can identify and extract estimated water elevation from imagery submitted that will be used to inform emergency managers and policymakers.
The CRIS-HAZARD app had a soft launch on September 18. On September 26, Hurricane Helene hit Florida’s west coast north of Tampa Bay as a Category 4 hurricane. On October 9, Hurricane Milton hit Florida’s west coast south of Tampa Bay as a Category 3 hurricane.
Images gathered from the back-to-back storms gave the team important information about how much flooding occurred and where in the Tampa Bay area.
“People used it to report what was happening in their backyard,” said Dixon, who is also executive director of USF’s Initiative on Coastal Adaptation and Resilience (iCAR). “This information is their lived experience, and we want to take that experience and make data out of it using AI tools that will process images and figure out the depth of the water. We can use that data to calibrate and validate models.”
In addition to the data provided for researchers, the pictures and videos are accessible to other users so they can see where flooding is happening, which helps inform them of hazards in the community.
Dixon compares it to the real-time traffic app Waze, where drivers can see traffic back-ups and where police activity is happening along roads and highways.
“We make it a real-time display for about seven days, then the data is archived,” Dixon said. “For example, if there is a big rain event today, you want to know what is happening today, not a month or two months back. But the data stays there so we can access and analyze it.”
