Abstract
Given the expansion of soybean cultivation, remote sensing (RS) has become strategic for validating estimates of planted area. This study compared three methodologies in the 2022/2023 harvest in 32 municipalities in Santa Catarina: the Epagri/Cepa method, MapBiomas, and the Crop Enhancement Index (CEI) applied to Sentinel-2/MSI images. The analysis included Friedman tests, linear regression, and interpretation of scatter plots and dot plots. The estimates showed a high correlation (R² > 0.92), especially between CEI and MapBiomas (R² = 0.97). The Friedman test did not indicate an overall difference (p = 0.26); however, regression analysis revealed a positive bias when comparing Epagri/Cepa with RS methods, indicating overestimations. The results reinforce the potential of RS as a robust and replicable method, indicating the need to harmonize approaches to improve official agricultural statistics.