MAPPING BANANA PLANTATIONS WITH SENTINEL-2 AND GOOGLE EARTH ENGINE IN THE MICROREGION OF ARARANGUÁ, SANTA CATARINA, BRAZIL

This paper analyzes the potential for using Sentinel-2 sensor data to identify banana plantations in the Araranguá microregion, southern Santa Catarina. Spectral bands and indexes were used to classify the 2023 land use into two classes: banana plantations and other uses. Data composition, training and land use classification process was carried out on the Google Earth Engine (GEE) platform. The classifier used was Random Forest. The results were validated using a reference sample set of 400 points. These points were compared to the classification results using confusion matrices. The accuracy assessment results indicate a strong potential for identifying banana plantations using Sentinel-2 data, with 0.934 as the highest user accuracy achieved.