The landscape can be considered a spatial unit, whose heterogeneity is modified mainly by the established relation between society and nature, producing a configuration marked by the fragmentation or connection between its elements. The monitoring of the landscape structure is fundamental to designate its configuration, acting as a subsidy to the understanding of the anthropic impacts in the natural landscapes. The aim of this project is to monitor the areas of the Cerrado (Brazilian Savanna) and Atlantic Forest in the Tietê-Jacaré Hydrographic Basin - São Paulo - Brazil through the Sentinel 1 and 2 satellite data. In order to develop and test methods of detecting forest changes we will use the machine learning (ML) approach. The information will be analyzed in Geographic Information Systems using the ArcGIS software and Python programming language with the geographic projection latitude/longitude - Datum SIRGAS2000 for all information plan. From the overlap of data from environmental characterization and the evaluation of the proposed models, it will be possible to observe environmental landscape structure, generate essential information about the current state of such regions and allow the evaluation of environmental responses to anthropic impacts.
News published in Agência FAPESP Newsletter about the scholarship: