出版社：Grupo de Pesquisa Metodologias em Ensino e Aprendizagem em Ciências
摘要：The objective was to evaluate the potential of Sentinel-2 satellite images to estimate the biomass of Tectona grandis L.f., in Rolim de Moura - RO. The methodology consists of 7 steps: 1) Forest inventory in 10 plots; 2) Selection and felling of 20 trees to obtain biomass from laboratory samples; 3) Estimation of biomass by plot using mathematical models; 4) Biomass estimation by remote sensing in dry and rainy periods, based on Sentinel-2 satellite images (NDVI, EVI-1, EVI-2 and SAVI); 5) Analysis of correlation, adjustment and selection of equations with biomass data (dependent variable) as a function of variables derived from the images (independent variables); Evaluation of the difference of independent variables acquired in the dry and rainy periods; and 7). Quantification and spatialization of biomass for the study area. The Schumacher and Hall - Log. model was the best for obtaining biomass. Image variables differ between dry and rainy periods. Correlations between biomass and image variables were higher in the rainy season. The variables that best estimate biomass were obtained in the rainy season, with equation 1 being selected (Y=35.625331–528.123868*B02)–2.653078*B03+273.491*B04+92.572857*EVI-1– 108.785571*EVI-2–22.807912*NDVI). Biomass estimates using Sentinel-2 images are accurate and useful for monitoring forest biomass, however, adjustments in mathematical models and studies with different sensors are recommended.
