Análisis de cambios de cobertura del suelo mediante imágenes Sentinel-2 y Landsat-8 en la subcuenca del río Burgay

Abstract

The analysis of images captured by satellites has demonstrated their importance in the identification and monitoring of the peculiarities of the earth's surface. These studies make it possible to detect alterations over time. Our research focused on analyzing the evolution of land cover in the Burgay River sub-basin, using images from Landsat-8 OLI and Copernicus S2 Sentinel-2 MSI satellites, available at the United States Geological Survey (USGS) library. The methodology included coding in the Google Earth Engine platform and use of the Random Forest supervised classifier algorithm. Land cover maps for three time periods (2016, 2018, and 2020) were produced independently for each satellite. The accuracy of the classifications was high, reaching 95.33% (2016), 96.75% (2018), and 98.62% (2020) for Landsat-8, and 99.28% (2016), 93.65% (2018) and 98.63% (2020) for Sentinel-2. The results highlight the importance of spatial resolution in land cover identification, with a difference in change of 6.37% (2016-2020) for Landsat-8 (resolution: 30 m) and 8.97% (2016-2020) for Sentinel-2 (resolution: 10 m) .These results not only indicate land cover change and persistence, they also demonstrate the advantages and limitations of using Google Earth Engine and the public archive database in its platform to track and monitor these transitions over time.