Avaliação da eficiência do monitor de secas para definição de secas em Sergipe

• Autor(a) principal: Santos, Kelly Marina Silva
• Data de Publicação: 2020
• Tipo de documento: Dissertação
• Idioma: por
• Título da fonte: Repositório Institucional da UFS
• Texto Completo: http://ri.ufs.br/jspui/handle/riufs/13618
• topic: Engenharia civil; Secas; Monitoramento; Gestão de secas; Políticas públicas; Monitoring; Drought management; Public policy; ENGENHARIAS::ENGENHARIA CIVIL

Drought is a recurring natural phenomenon in several regions around the world and causes serious effects on the environment and the population. Historically, governments of affected locations usually adopt emergency measures to fight the impacts of the drought. In the last decades, drought preparedness programs were developed in many countries, like United States, Portugal and Mexico. In Brazil, the implementation of the Brazilian Drought Monitor (MSB) shows the beginning of a transition from crisis management to risk management. This tool consists of a drought monitoring based on multiple indices and observed data, through a participatory process between federal and state institutions. Due to the importance of consolidating the Monitor as a means to support the decision for public policies, this research aimed to evaluate the performance of MSB regarding the identification of droughts in the State of Sergipe. State climate data and emergency decrees were used for association analysis with the Monitor classification of drought, through the application of statistical tools: Chi-Square independence test and Cramer's V coefficient. In addition, the Minimum Standard Form (FMP), a questionnaire answered by members of the process to verify the local perception of impacts of drought, was analyzed to check the correspondence with the MSB impacts table. Through the statistical tests applied, an association between monthly, quarterly and half-yearly rainfall and the MSB categories was observed, since the null hypothesis of independence was rejected, with a p-value of 2,5x10-5, 2,3x10-10 e 8,7x10-15, respectively. However, the results of Cramer’s V coefficient were close to zero (0,21, 0,28 and 0,32, in this order), indicating a weak association between the studied classifications. Similar results were found for monthly, quarterly and half-yearly precipitation anomalies, with p-values of 6,7x10-5, 1,3x10-5 e 8,3x10-15, respectively. In this case, Cramer’s V values were 0,20, 0,21 and 0,32. NDVI was the climatic variable with the best results, rejecting the null hypothesis with p-value of 7,8×10-133 and Cramer's V coefficient of 0,52. Therefore, it can be stated that, for the state of Sergipe, NDVI has a higher association with the Monitor categories than the other variables. Since the promotion of a common understanding of the drought situation between the federal and state institutions is one of the MSB's objectives, this work also analyzed the existence of a correlation between Civil Defense emergency decrees with short and long-term impacts of the Monitor. In this case, the Chi-Square test resulted in p-value of 0,09 and Cramer’s V was equal to 0,12, suggesting that there is no association between the MSB maps and emergency decrees. Finally, the main problem identified through the analysis of the forms was the subjectivity in its questions and answers. Thereby, a reformulation of the FMP was proposed, in order to make it more objective. In general, it can be concluded that the Monitor has represented the state of Sergipe with low efficiency, requiring adjustments and improvements, due to its growing importance as a decision support tool.