North Carolina State University has developed a new mapping tool that utilizes machine learning and satellite imagery to predict flooding in urban areas. The tool aims to provide valuable data for urban planners to identify flood-prone areas and allocate necessary resources for flood prevention and resilience.
Urban environments pose unique challenges when it comes to collecting satellite data and tracking water flow. The presence of tall buildings casting shadows and complex drainage systems makes it difficult to predict where flooding may occur. Additionally, flooding in urban areas tends to start and end quickly, making it challenging to gather sufficient data.
To overcome these challenges, graduate student Rebecca Composto used satellite data from Hurricane Ida to train a machine-learning model to recognize and map flooding. By hand-drawing polygons onto satellite images, she provided the model with training data to identify characteristics associated with flooding and create accurate flood extent maps.
Comparing the resulting flood extent map to Federal Emergency Management Agency flood zones revealed that more flooding was occurring in minimal hazard areas than in designated flood zones. This discovery highlights the importance of identifying and addressing flood risk in areas that may not be classified as high-risk based on traditional criteria.
Moving forward, Composto plans to simplify the model for easier use and share the code with emergency-response leaders. By integrating a new map displaying flood depth, the tool can provide enhanced insights into urban flooding and support informed decision-making in flood-prone areas.
The research paper detailing this innovative mapping tool, titled Quantifying urban flood extent using satellite imagery and machine learning, underscores the potential of combining technology and spatial data analysis to improve flood risk assessment in urban environments. With further refinements and applications, this tool could become a valuable resource for supporting flood mitigation efforts and enhancing urban resilience in the face of changing climate patterns.
