Open-source tools for flood risk assessment for multiple spatio-temporal scales under scenarios of change

• Ano de defesa: 2024
• Autor(a) principal: Rápalo, Luis Miguel Castillo
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: eng
• Instituição de defesa: Biblioteca Digitais de Teses e Dissertações da USP
• Palavras-chave em Português: correção de modelos digitais de elevação; digital elevation model correction; digital twin; early warning system; gêmeo digital; human instaiblity risk; modelagem espacialmente distribuída de inundações; risco de inestabilidade humana; sistemas de alerta antecipada; spatially-distributed flood modeling
• Link de acesso: https://www.teses.usp.br/teses/disponiveis/18/18138/tde-07102024-094249/

Flooding poses a significant threat to communities worldwide, with climate change expected to exacerbate the frequency and intensity of such events, highlighting the need to enhance resilience to cope with the future risks. To this end, the main objective of this work is to develop multiple novel open-source tools to advance flood risk assessment and forecasting capabilities across multiple spatiotemporal scales under climate change scenarios within poorly gauged areas. The developed tools are related to a coupled scheme of hydrologic and hydrodynamic model (HydroPol2D) based on cellular automata approach, capable of parallel computing, allowing for overcoming spatial scale limitations in order to simulate from a watershed up to a continental scale. This thesis is divided into an introductory chapter, four research papers (Chapters 2 to 5), and a general conclusions chapter. Chapter 2 presents a novel open-source algorithm, Forward Moles, for enhancing hydrodynamic modeling through digital elevation model (DEM) correction by performing stream burning to better represent drainage networks, overcoming hydraulic discontinuities for improved terrain depiction and spatially distributed flood modeling accuracy. The application of the fully-distributed 2D hydrodynamic model HydroPol2D, Chapter 3 shown methods to assess human instability flood risk (HIFR) in an urban watershed, comparing theoretical and empirical approaches, where the empirical method critically identifies HIFR zones overlooked by conventional inundation assessments. Furthermore, climate change tends to increase such risk areas and potential injuries/fatalities. Following this line, in Chapter 4 includes multiple vulnerability degrees for HIFR analysis, as well as a downscaling method for climate change data, which reveals children as most vulnerable based on age, gender, weight, and height, with their instability highly sensitive even to localized intense rainfall pulses, aiding urban planning and risk communication efforts. In Chapter 5, we developed an open-source coupled system integrating hydrodynamic modeling, satellite rainfall data, and numerical weather predictions as a Digital Twin and an early warning system (EWS), successfully capturing extreme events like hurricanes for near real-time forecasting in data-scarce regions and monitoring hydrological conditions to enhance disaster preparedness. Finally, Chapter 6 presents the overall conclusions and outlines future steps or challenges that could be addressed to enhance the tools developed in this thesis.