The proposed research will advance flood inundation mapping and enhance situational awareness in disaster response situations through a combination of machine learning techniques and Synthetic Aperture Radar (SAR) data. One of the most difficult challenges during the early stages of a flood event is acquiring timely, unobstructed Earth observation data that can provide lifesaving insight into the situation on the ground and safely direct first responders to where they are needed most. Unfortunately, satellite images of the affected areas are often obscured by cloud cover. SAR is an especially promising technology for addressing these challenges, as it can continually gather ground-level data, regardless of cloud cover or even time of day. The complementary field of machine learning, and especially the subdiscipline of deep learning, offers significant potential for effectively monitoring and interpreting SAR imagery in near-real-time. By combining these two technologies, this project will support the rapid delivery of accurate flood inundation maps that will enable first responders, humanitarian relief organizations, and other decision-makers on the ground to effectively route resources and identify highly impacted areas, both during and following extreme weather events.
