Geoprocessamento aplicado ao mapeamento da suscetibilidade a escorregamentos no município de Porto Alegre, RS

BRITO, M. M. Geoprocessamento aplicado ao mapeamento da suscetibilidade a escorregamentos no município de Porto Alegre, RS. 2014. 167 f. Dissertação (Mestrado em Engenharia) – Programa de Pós Graduação em Engenharia Civil, UFRGS, 2014. Mass movements are natural phenomena that are part of the surface dynamic processes that shape the Earth’s surface. When they occur in urban areas, can cause economic losses, social impacts and, in extreme cases, losses of human lives. The implementation of preventive measures to mitigate those damages requires the identification of the areas prone to these events. Therefore, the aim of this dissertation was to develop a methodological approach to map landslide susceptibility at the municipality of Porto Alegre-RS, with the use of geoprocessing tools. The variables used were: slope, geology, distance from lineaments, flow accumulation and land use, which were processed in the Idrisi software. After the standardization of the variables through the application of fuzzy functions and the direct assignment of notes, the importance of each one on the predisposition of landslides was defined, based on the opinion of experts, with the support of the technique entitled analytic hierarchy process (AHP). The integration of the variables was based on weighted linear combination, resulting in two scenarios: the first with the natural susceptibility and the second with the induced susceptibility. The results indicate that the most susceptible areas to landslide are located in the hills of the Crista de Porto Alegre and in the isolated hills located at the south of the municipality. The remaining area has, in general, low susceptibility. The validation of the results obtained an overall accuracy of 95% and 94%, respectively for Scenarios 1 and 2. The areas under the ROC (Receiver Operating Characteristic) curve were respectively 0.960 and 0.966, indicating that the variables and weights were adequate and both scenarios are robust. Thus, it is noted that geoprocessing tools are useful to map the landslide susceptibility, since they make the mapping process and its updates agile, and give more clarity and transparency to the susceptibility assessment process. In addition, the proposed method reduces efforts in the early stages of risk management, enabling the manager to prioritize the most susceptible areas for detailed mappings. This demonstrates that decision support routines, such as those used in this dissertation, can be incorporated into susceptibility studies, providing important information for the urban planning. Key-words: susceptibility mapping; multicriteria evaluation; weighted linear combination; geographic information system.