Global Flood Observation with Multiple Satellites

A global historical geodatabase of flood-event extents is key to providing accurate exposure information in vulnerable areas that cannot afford traditional flooding models. Decades of Earth observing satellites and cloud computing make it possible to run water detection algorithms back in time to capture the spatial extent of floods globally. Cloud to Street uses the Google Earth Engine to map large numbers of flood events and trends with the Moderate Resolution Imaging Spectroradiometer (MODIS) and Landsat satellites. These methods are used to map flood histories and create decision-making tools for watershed managers in Rio Salado, Argentina, and the Nile Basin in Ethiopia, Sudan, and South Sudan. In Rio Salado the gentle slopes make flood modeling difficult, but the empirical observations obtained can be used to better identify wetland areas, support tax-relief programs for farmers, and improve land-use planning. Watershed managers in the Nile Basin are using flood histories to identify flood-model errors and improve modeling efforts in order to support early warning systems. Global flood mapping methods applied to local contexts can improve watershed resilience by providing new data for better decision-making.