Answers are grounded in this paper and cite their sources.
The characterization of land uses and covers is crucial for the management and monitoring of agricultural territories and natural resources to facilitate decision-making by various entities. This research aimed to provide updated information on the main agroproductive systems using remote sensing techniques to aid decision-making for both governmental and non-governmental entities involved in territory management. The methodology involved adapting computer code in the Google Earth Engine (GEE) platform, utilizing Sentinel-2 images from the year 2021 and field survey data. Three classification algorithms were applied to the image mosaic, and their accuracy was validated, with "Random Forest" showing the highest overall accuracy at 82.95%, followed by "Classification and Regression Tree (CART)" at 81.50%, and finally "Minimum Distance" at 57.51%. According to the "Random Forest" results, the main land uses identified were forest cover (136,162.78 ha), followed by pasture classes (25,497.24 ha), sugarcane cultivation (11,081.02 ha), and coffee cultivation (5,512.17 ha). Additionally, two agricultural suitability maps were generated for sugarcane and coffee crops to identify potential and non-potential areas for their development. The research encourages the application of these tools for national-level territory management, benefiting the agricultural sector.
No takes yet. Share an insight, caveat, or question.
Céspedes et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: