Enhancing the temporal resolution of satellite-based flood extent generation using crowdsourced data for disaster monitoring

Remote-sensing satellite data are routinely used during disasters for damage assessment and to coordinate relief operations. Although there is a plethora of satellite sensors able to provide actionable data about an event, their temporal resolution is limited by their revisit time, presence of clouds, and errors in the reception of data. These limitations do not allow for an uninterrupted monitoring, which is crucial during disasters and emergencies. This research presents an approach that leverages the increased temporal resolution of crowdsourced data to partially overcome the limitations of satellite data. The proposed approach focuses on the geostatistical analysis of a combined satellite and Twitter data to help delineate the flood extent on a daily basis. The crowdsourced data are used to augment satellite imagery from Advanced Land Imager instrument on Earth Observating One (EO-1) satellite, Landsat 8, WorldView-2, and WorldView-3. The proposed methodology was applied to estimate the daily flood extents in Charleston, South Carolina, caused by the October 2015 North American storm complex. The results of the proposed methodology indicate that the user-generated data can be utilized adequately to both bridge the temporal gaps in the satellite-based observations and also increase the spatial resolution of the flood extents.