Author(s): SOARES, V. B.; SILVA, G. B. S. da; BOLFE, E. L.; PARREIRAS, T. C.; CASTRO, V. H. M. e de; SANO, E. E.

Summary: This study focused on assessing the dynamics of smallholder agricultural areas between 2019 and 2023 in Breves, Pará State, Brazil, using geographical information systems and remote sensing techniques. The classification of Sentinel-2 images resulted in an overall accuracy of 99% and a Kappa coefficient of 0.98 for distinguishing between forest and non-forest, and an accuracy of 89.7% and a Kappa coefficient of 0.86 for distinguishing between different types of land cover. The analysis revealed a significant increase in the areas of secondary vegetation and bare soil, while the areas of forest, pasture, and water remained relatively stable throughout the period of analysis. The transition between land cover classes indicated a conversion of areas from forest to secondary vegetation and then to bare soil, possibly reflecting logging and small-scale farming activities. These results highlight the importance of remote sensing in characterizing and monitoring small-scale agricultural production, providing valuable insights for sustainable rural development.

Publication year: 2024

Types of publication: Paper in annals and proceedings

Unit: Embrapa Environment

Keywords: Agricultural transition, Amazonia, Cobertura da terra, Cobertura do Solo, Floresta aleatória, Land cover, Land use, Produção Agrícola, Random Forest, Transição agrícola, Uso da Terra, Uso do solo