Proposição de metodologia para construção de gráfico de risco de inundações em bacias urbanas: estudo de caso Bacia do Córrego Cachoeirinha

Abstract

Flood events are recognized as one of the leading causes of economic damages and loss of life worldwide. Unfortunately, it has been observed that its occurrence has become increasingly frequent and of greater magnitude. An analysis developed by the Center for Research in Disaster Epidemiology (CRED) showed that in 2016, among the types of natural disasters, floods corresponded to 50% of occurrences in the world. Among 164 countries, in which the largest number of people were affected, Brazil ranked 11th in 2015, according to the World Resources Institute. In response to flood events, Brazilian cities, like other cities in the world, have sought to extinguish or mitigate the risks to which people are exposed. To minimize the risks to which the population is exposed, it is necessary to invest in the incorporation of tools that aid in the decision-making process related to urban drainage management. With this aim, it is necessary to foster the incorporation of tools which may contribute in the decision-making process in urban drainage management. In this context, the present work proposes a methodology for the construction of a flood risk graph in urban basins, as a tool to help in the elaboration of action plans and guidelines to deal with flood events, whether in the prevention or mitigation phase of impacts. For that, the SWMM model was applied in the Cachoeirinha watershed (Belo Horizonte, Minas Gerais). The model was calibrated and validated with precipitation data and water levels recorded in monitoring stations recorded located in the study area.Nine events were selected for calibration and nine for validation. The choice of events was based on the type of risk alert that was generated in the Cachoeirinha stream. The risk alert levels are defined as: yellow, orange and red, corresponding respectively to the water level occupying 50%, 80% of the cross-section of the channel and overflow condition. The performance of the model reached a satisfactory result according to the objective function: the Nash-Sutclife coefficient was 0.72 for the calibration and 0.70 for the validation. Design precipitations with a return period varying from 2 to 100 years and several durations were simulated and the maximum water levels obtained were classified according to the defined risk alert levels. From this classification design precipitation intervals were established, in which the limit value between the risk alert ranges was contained. Precipitation values contained in these intervals were simulated until the change in the alert range of the graph was detected. In order to evaluate the proposed graph, the selected events were plotted and a good adjustment of the risk bands was verified