Andrew Molthan (ST11) was selected as a Co-I to work with PI Beth Tellman of commercial entity Cloud2Street, evaluating the potential use of commercial satellite imagery from Planet in the detection of water and flood water extent, using machine learning approaches. The team will be complemented by Co-I Iksha Gurung of ST11/IMPACT and two additional students who will help lead the organization of image labeling of water features and other activities necessary for building relevant training data and optimizing the use of machine learning tools. Over the 18 month period, the intent of the proposal is to develop an optimized method for detecting flood water from the imagery and to provide feedback on the utility of the commercial data to NASA. The selection of this work is complementary to ongoing MSFC Earth Science Disasters Team efforts to improve capabilities for floodwater detection from multiple platforms, and to build new partnerships with the external community -- here, a new commercial partner operating in the disasters-thematic space.
