Uso do NDVI para análise multitemporal das áreas impactadas pela lavra de Bauxita na região norte do Brasil entre os anos de 2018 a 2021

Abstract

Bauxite is a rock primarily found in tropical and subtropical regions of the planet, used as raw material to obtain aluminum. Brazil is one of the world's largest producers of this ore, with the state of Pará being the federal unit that contributes the most to this result. The method used in bauxite ore mining typically requires the removal of parts of the native vegetation, making environmental restoration necessary, which can be carried out in tandem with mining operations. In light of this, the present study aims to analyze the development of areas impacted by bauxite mining in the Northern region of Brazil through a multi-temporal analysis of satellite images using the Normalized Difference Vegetation Index (NDVI). To achieve this, the research was divided into four stages. The first stage involved selecting the main bauxite-producing companies in the country. The next stage involved the acquisition and processing of georeferenced data. The third stage was based on the interpretation of images using NDVI with bands 4 (RED) and 8 (NIR). Finally, the fourth stage involved the multi-temporal assessment of bauxite-producing areas through the creation of a time series of thematic maps and vectorization of classification images of areas impacted by bauxite mining. Analysis of mineral production reports allowed for the selection of the three main bauxite-producing companies in Brazil, all located in the state of Pará: Mineração Rio do Norte, Mineração Paragominas, and Alcoa World Alumina, respectively. The acquisition and processing of georeferenced data enabled the identification and visual analysis of these companies' areas. Determining the NDVI of the areas made it possible to identify non-vegetated surfaces, exposed soil, low vegetation, moderate vegetation, and high vegetation for all years through coverage classes, represented by colors ranging from red to green. However, the multitemporal assessment of bauxite-producing areas had limitations, as there was no atmospheric window that allowed the acquisition of representative images of all areas during the analyzed period. Nevertheless, despite these limitations, the technique proved effective in identifying certain points for the different companies analyzed, such as possible vertical advances in mining over the time series, changes in the qualitative landscape layout of the NDVI indicating environmental restoration efforts, and an increase in exposed soil areas suggesting the expansion of bauxite mining.