ALBUQUERQUE, E. M.; http://lattes.cnpq.br/5467509964012082; ALBUQUERQUE, Erickson Melo de.
Abstract:
Agriculture is one of the main economic activities of Brazil, especially the cultivation of
sugarcane, raw material for the production of sugar and biofuel for cars, and other alcohol -chemicals. Grown in a wide range of latitude of the globe, development of sugarcane depends on environmental factors such as climate, the incident radiative energy on the region and soil water availability, which varies according to the region's location in the planet and proposes adjustments to the crop management. The information about productivity of sugarcane is fundamental to the planning of sugar and alcohol, and through remote sensing activities you can get it. Thus, the aim of this study was to estimate the productivity of sugarcane through an agrometeorological - spectral model, using as input data products derived from satellite SPOT - VGT and MSG, which focuses on the modeling of crop coefficient (Kc) by remote sensing, over Barretos/SP and Morro Agudo/SP, for the crop year 2010/2011. Products LSA SAF ET, SPOT DMP and SPOT NDVI-S10 were integrated into the GIS ILWIS, and meteorological data INMET and vector data provided by Conab containing planted areas with sugarcane over the study area. Seven in all processing steps were performed, starting by import of satellite products using the ILWIS processing routines batch files, including filtering of NDVI pixels contaminated using SM data, application of conversion factors and cropping images to the
municipal limits; then was computed the Fraction of Coverage Vegetation (FCV); then Kc,
requirement to calculate the crop evapotranspiration (ETc); later obtained the Yield Potential (Yp) of sugarcane; and, finally, the Estimated Yield (Ye) spatially distributed. In total, 365 ET images, 36 NDVI images and 36 DMP images were imported. After application of the SM data, Barretos presented average around 70% and Morro Agudo 80% success pixels, respectively. Through the NDVI product was identified during the phenological period of sugarcane. From this, the period for the computation of Ye was defined. The most intense precipitation events occurring during the rainy season in the region increased the number of contaminated pixels and contributed to decrease NDVI and DMP values. Factors dependent on these parameters were also affected, such as FCV and Yp. However, the observed relationships between the profiles of NDVI, FCV and DMP proved consistent and between the Kc and ET too. In turn, the Kc modeled by remote sensing showed similar behavior to that found in the literature. The Yp and Ye showed follow the profile of NDVI. Finally, the Ye maps pictured their spatial distribution, with different patterns for the two municipalities. Compared to the results of other studies, the mean value of Ye was underestimated in Barretos, which can be attributed to lack of data in images, and overestimated in Morro Agudo, which may be related to contamination in shades of planted area.