Análise do comportamento dos coeficientes de retroespalhamento da mineração em imagens sentinel-1: um estudo de caso na floresta Saracá-Taquera (PA)

Abstract

Mining plays a crucial role in the national economy, contributing significantly to the well-being and improving the quality of life of present and future generations. However, it is essential that this activity is conducted with social responsibility and within the precepts of sustainable development, aiming to build an equitable society. The present study aims to analyze the values of the backscatter coefficient in mining areas in the Saracá-Taquera National Forest, in the state of Pará, using radar images. In the adopted methodology, three main classes were considered: two mining subclasses (reservoir and extraction), in addition to the soil and forest classes. In order to determine the number of samples per class, stratified random sampling was used, a method that ensures the representativeness of each stratum in the sample. The Google Earth Engine platform was used to process radar images and obtain backscatter values. Data analysis was performed using the Kruskal-Wallis non-parametric statistical test, complemented by the Dwass-Steel-Critchlow-Fligner multiple comparison. The results revealed that the mining subclasses presented values similar to those of the soil class within the range of -15 dB to approximately -23 dB, indicating that the identification of bauxite mining areas based on backscatter coefficient analysis may result in inaccuracies. . Despite the difficulties encountered, this study provides important information for understanding the dynamics of mining areas in the studied region. These findings highlight the need for more refined and integrated approaches to effectively identify and monitor these areas, aiming to balance economic development and environmental conservation.