Wetland landscape based on Sentinel-2 images and geo-tagged photographs in Centla, Tabasco

Abstract

This is an analysis of the geographic landscape in the Centla wetlands of Tabasco, Mexico. A map shows the use of remote sensing data combined with easily understood and conveyed visual descriptive data which show the ecological conditions of the landscape. The central map of this article presents a land use and land cover study, obtained from Sentinel-2 MSI data for the Centla wetland zones. The support vector machine algorithm is used to classify Sentinel-2 images. The results show a high general precision of 90.0%, as well as high precision in separating types of wetlands. Information obtained during fieldwork at the ground level is inserted in the map, comprised of photos taken with a mobile application along the Grijalva and the San Pedro y San Pablo rivers. These photos provide visual verification of the map.