Mapping floods in urban areas from space at local risk level
Dr. Guy Schumann
Open EO data has long held promise for wide-area flood mapping and many algorithms exist to serve flood maps across large spatial scales. A lot of those maps are being used to support situational awareness assessments. However, typically, open EO imagery works well over open water rural areas but in areas where most people and assets at risk are located, i.e. urban areas, traditional flood mapping algorithms applied to free satellite data have serious limitations. However, recently, advances in using SAR signal coherence change for mapping floods coupled with an increase in powerful cloud computing, make urban flood mapping a reality. In this talk, we present examples of use cases using an urban flood map algorithm on an online cloud-based EO processing platform to rapidly process Sentinel-1 SAR images into urban building geometries that are then used to derive an accurate urban flood map using SAR signal coherence change.