FDL Waters of the United States

Stream networks across the planet expand and contract with floods and droughts. What if we could map these networks in near real-time and develop forecasts to classify streamflow throughout the United States? The objective would be to use machine learning to predict the level of waters in the United States using high-resolution, high frequency satellite imagery. This in turn would be a valuable tool for managing drought and ensuring the availability of water when and where it’s needed.