Improving Flood Monitoring Methods
ECE Assistant Professor Pau Closas, Professor Deniz Erdogmus, CEE Professor Edward Beighley, and ECE Postdoc Tales Imbiriba were awarded a $98K grant from the National Geographic Society’s Committee for Research and Exploration for “Spaceborne Multispectral Image Fusion for Water Mapping and Flood Management”.
Natural disasters such as floods lead to enormous economical and social impacts as property is damaged, crops are lost and people are relocated or lose their lives. The occurrence of such natural hazards is increasing due to climate change making prediction models less accurate which is particularly relevant for flood risk assessment, underscoring the need for observation-driven flood monitoring methods. This demands improved remote sensing tools for monitoring, detecting, and estimating the intensity of key natural phenomena. In this project, we focus on the flood estimation and detection problems leveraging multispectral satellite imaging using instruments with different spatial, spectral and temporal (revisiting times) resolutions. The main drawback in flood assessment with satellite images is related to the resolution/revisiting time tradeoff where lower resolution images (in the order of kilometers) are available at smaller revisiting time (1 to 2 days) while higher resolution images (in the order of tens of meters)are available at larger revisiting times (5 to 16 days). Our goal is to provide a methodology and software capable of providing a spatially detailed assessment of flooding incidents by exploiting multiple spaceborne imaging sources. This will ultimately lead to improvements in water mapping, flood detection, and flood mapping.