Monitoramento de indicador de qualidade de ar por processamento de imagens de satélite

Abstract

The scarcity in the number of meteorological centers for sampling in Brazil makes data on air quality scarce. The lack of frequent monitoring of air quality indicators is only noticed when it reaches levels that are harmful to health. There are several studies on air quality indicators, mainly focused on the PM2.5 indicator. Contamination by PM2.5, which is 2.5μm particulate matter, can cause heart problems, asthma in children and many respiratory problems. These risk factors demonstrate the importance of studying this indicator, since the type of particle it represents can affect the entire population of an area, impairing the quality of life and health of the inhabitants. Remote sensing is a tool that uses satellite images to calculate the conditions of various indicators, such as temperature, soil quality and air quality. In order to have a good idea of air quality, this study proposes to deepen the knowledge about remote sensing and satellite image processing, applied to the problem of estimating the amount of PM2.5 in the air. A systematic review of the approaches and methods being used to address this problem was carried out. After analyzing this review, the objective of this work was to study the functioning of systems and concepts, focusing on measurements, for evaluation and estimation. Finally, a system to predict PM2.5 was developed, and used to predict PM2.5 in São Paulo, obtaining '2 = 0.112 and RMSE = 17.15. Then, this work raises the possibility of applying the methods studied in Brazil, and a conclusion of the most current situation in Brazil, arriving at the need for the most recent meteorological studies for Brazil.