A spatio-temporal assessment of urban flooding within the United States
Schroeder, Amanda Jean
MetadataShow full item record
Urbanization leads to drastic changes in land cover, which has major effects on local hydrology. Recently documented urban flooding has highlighted the need for more research on the unique interplay between hydrometeorology and the built environment, so this dissertation explored three major components of the urban water cycle that affect the development and magnitude of flash floods: atmospheric contributors, hydrologic response, and land use/land cover characteristics. To explore the atmospheric component, quantitative analyses were performed on thermodynamic variables from forty urban flood events. Spatio-temporal hydroclimatological trends and characteristics associated with urban flooding events for several cities were examined to address the hydrologic response and land use/land cover aspect. Finally, the three components (atmospheric, hydrologic, and land use/land cover) were combined via an observational case study of a historic flood in Oklahoma City to identify what urban-atmospheric-hydrologic interactions played a role in the evolution of the event. Additionally, various observational networks, including the unique, intra-urban Oklahoma City Micronet, were examined to identify optimal observing methodology. Results showed that detailed assessments of the atmospheric and surface/subsurface conditions (from both a hydrologic and land use perspective) are crucial to obtain accurate assessments of the potential for, and resultant magnitude of, urban flash floods. When assessing the atmospheric conditions prior to a heavy rain event, the precipitable water (PW) anomaly with respect to the local PW climatology, as well as the warm cloud depth (WCD), MUCAPE, and wind shear are key. From a hydrologic and land use perspective, results illustrated that urban basins experienced a higher percentage of flood events during abnormally dry conditions and were more “flashy”. Finally, from an observational perspective, gage-biased radar mosaic QPE provided the best continuous representation of the rainfall distribution over the area but were not adequately able to capture all details of the observed distribution from the Oklahoma City Micronet, illustrating the need to include both in situ and remotely sensed observing networks for an improved characterization of urban rainfall.