Date of Award
Open Access Dissertation
Floods are among the most devastating hazards on Earth, posing great threats to a large amount of population in the world. As the severity and frequency of flood events have noticeably increased, there is a growing need to improve the flood awareness and exposure analysis to assist flood mitigation. Fortunately, the Era of Big Data has fostered many innovative spatial data sources as well as spatial data analytics. This dissertation advances the existing flood monitoring studies by obtaining enhanced flood awareness via the development of a data fusion enable and deep learning supported flood monitoring framework that systematically integrates remotely sensed observation with in situ documentation from crowdsourcing platforms. In addition, this dissertation advances flood exposure studies via the application of long-term nighttime remote sensing series for the estimation of hurricane exposure in U.S Atlantic/Gulf coasts and the development of a spatially explicit population disaggregation method for comparative assessment of the exposed population within 100-year floodplains in the entire Conterminous United States (CONUS). In the Big Data Era, the important theoretical, methodological, and contextual knowledge gained in this study could greatly benefit local authorities and federal agencies for better preparedness of flood as well as other types of natural disasters in a geospatial framework.
Huang, X.(2020). Remote Sensing and Social Sensing for Improved Flood Awareness and Exposure Analysis in the Big Data Era. (Doctoral dissertation). Retrieved from https://scholarcommons.sc.edu/etd/5851