Description
In the last few years, the need for monitoring and mapping vegetation increased, partly because of deforestation, and in part due to the need of technological advancements in agriculture and silviculture. The use of vegetation indexes (VI), generated based on remote detection data, constitutes an important strategy for the detection of natural or anthropic changes. IVs are obtained through calculations based on data acquired through remote detection, due to its simple and efficient manner of highlighting the green signal while minimizing variables due to solar irradiation and the effects of the canopy substrate. This article’s main objective is to present a comparative study of NDVI and SAVI maps generated from scenes obtained by the CBERS-4 satellite, MUX sensor, and LANDSAT-8 satellite, OLI sensor, the newest of its families, with the aim of investigating if there are significative differences between them and which one is most suitable to study the intended area. This study was performed in Sao Francisco de Paula/ RS where agricultural economic activities are extensively explored and possesses significant ecological value in its forests. The digital processing was performed through the SPRING 5.4.3 software. The NDVI presented smaller values than SAVI regarding every class employed in both images. The images from CBERS-4 produced more satisfactory results, since it better evidenced the different types of vegetation. The images of vegetation indexes NDVI and SAVI confirmed its applicability as a tool for detection and characterization of the vegetation cover. The reflectance of the analyzed targets was statistically different. The effect of the soil in the studied area was visible. Therefore, considering the results obtained in winter, it is recommended the use of SAVI instead of NDVI due to the strong influence of the soil.
