Previsão de produção de biomassa de cana-de-açúcar por índice de vegetação por diferença normalizada (NDVI)

Abstract

The technological development of sensor and orbital systems in recent decades has provided significant advances in studies by remote sensing. One example is the crop forecasts by the so-called vegetation indices indexes, this technology has been applied to sugarcane, allowing to improve its management, and to reduce costs and time with field samples, and to improve estimates by covering large areas extensions rather than spot sampling. Therefore, the objective in this work was to evaluate the use of images of the Normalized Difference Vegetation Index (NDVI) in the estimate of the biomass production of sugarcane cultivars, by the methodology of sum of pixels per plot (NDVIs), in the municipalities of Iturama and União de Minas, Triângulo Mineiro.. The evaluation was performed by modeling, in regression analysis, with the independent variable being the sum of the NDVI values of the pixels (NDVIs) and the dependent variable the biomass production of the respective plot. Altogether, different modeling was performed for the 5 cane cultivars (CTC4, CTC15, RB835486, RB92579, RB867515) and a general one, independent of cultivar. With NDVIs images of the cane at 7, 8, 9 and 10 months after planting or harvesting the previous crop. It was concluded that the modeling using NDVIs as an independent variable, instead of the average NDVI values, was promising in estimating the production of sugarcane biomass. Welladjusted models of biomass production were obtained based on NDVIs in cultivars CTC4, RB835486, RB92579 and for the global (in distinction of cultivars) for all periods after planting / harvesting the previous crop studied. From the results, it can be said that the use of vegetation indices, in this case the NDVIs, is a viable technique for forecasting sugarcane crops, assisting logistics of harvesting and marketing sugar and ethanol.