Investigação das semelhanças entre mapas de ndvi de plataformas de sensoriamento

Abstract

The use of remote sensing for Earth observation with the use of satellites began in the 1970s with the launch of Landsat. Since then, new constellations and platforms have been launched, allowing for a diversity of images of the Earth's surface. Thus, the objective of this is to evaluate the Normalized Difference Vegetation Index (NDVI) data of soybeans from the 2023/2024 summer harvest, obtained from the Sentinel 2, Landsat 8-9 and Planet platforms. The work was conducted with images of an area of FAECA that contained the development of the soybean crop in the 2023/2024 harvest. The images were downloaded from the artificial satellites Landsat, Sentinel-2 and Planet, made available on free platforms, the Sentinel Hub for the first two, and access to the Planet images was done through the Education and Research Program. The number of images was one per month (from October 2023 to March 2024). The images were downloaded from the artificial satellites Landsat, Sentinel-2 and Planet, made available on free platforms, the Sentinel Hub for the first two, and access to the Planet images was done through the Education and Research Program. The number of images was one per month (from October 2023 to March 2024). The images were reprojected for Universal Transverse Mercator Projection – UTM, and then the NDVI was calculated, later reprojected to a 3 x 3 m matrix, thus standardizing the NDVI pixel for all imagers. Subsequently, the following analyses were performed: descriptive, correlation, area calculation and Kappa index. From the images obtained, it can be seen that there were small differences between the NDVI values between the platforms, the most relevant being that of January/2024 between the Planet and Landsat data. Therefore, the soybean NDVI index obtained by the images from the Planet and Sentinel 2 platforms is more similar than when compared to the data from the Landsat 8-9 platform.