Desenvolvimento de um modelo de data science para prevenção de enchentes

• Ano de defesa: 2023
• Autor(a) principal: Barcelos, Daniel Visentini de
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Santa Maria; Brasil; Administração Pública; UFSM; Programa de Pós-Graduação em Gestão de Organizações Públicas; Centro de Ciências Sociais e Humanas
• Palavras-chave em Português: Gestão pública; Inundações; Python; Public management; Floods; CNPQ::CIENCIAS SOCIAIS APLICADAS::ADMINISTRACAO::ADMINISTRACAO PUBLICA
• Link de acesso: http://repositorio.ufsm.br/handle/1/29689

Urban flooding is common in Brazil. This is due to the acceleration of urbanization, which leads to changes in land use. These floods cause economic losses, environmental effects, disturbances in the normal lives of the affected people and in addition to human losses. In recent decades, Brazil has experienced a rapid process of urbanization that has occurred without obstacles without adequate urban planning, especially without inspection of infrastructure systems. These changes in cultivation and occupation have a direct impact on drainage systems due to this incompatibility. In general, floods are the most frequent natural disasters and have increased significantly in recent years, causing human casualties and great economic losses worldwide. Despite reaching a wider rural area, the greatest losses and management difficulties are concentrated in urban areas. Therefore, this dissertation aims to develop a framework that helps in mapping areas susceptible to flooding, through automated monitoring and prediction of flood occurrence. For this, a code was developed in the python language using googlecolab. Design Science was chosen as the methodology of the study. The results indicate that the framework is capable of predicting floods and with a greater volume of data the level of accuracy of the system will be improved, as possible future work intends to test the framework with more robust databases.