Delineamento de zonas potenciais para manejo diferenciado em nível de talhão apartir de dados de colheita e de imagens de satélite

Abstract

Precision agriculture has been used to assist farmers to choose the best management strategies, by identifying variability present in the area and from that, to determine the limiting factors and propose different management alternatives, according to the needs of each zone. It is important to create instruments that respond to one of the main issues, common to all of those who have ambition to the adoption and implementation of production modernization techniques: the technical and economic viability of modern production agriculture. The aim of this study was to relate stability maps for yield productivity and stability maps by Normalized Difference Vegetation Index (NDVI) for determination of management zones. The study was based on an analysis of five years of crop yield data and five satellite images Landsat 5 and Landsat 8 for Vegetation Index (NDVI) data for 2007 (wheat), 2008 (soybean), 2009 (soybean), 2013 (wheat) and 2014 (soybean). With processed data, stability maps for yield productivity and stability maps by Normalized Difference Vegetation Index (NDVI) were generated, following data parameters and coefficient of variation. Both maps were compared and validated by the Kappa index. As a result, we obtained a weak Kappa index relationship (0.2623), but 57.48% of the area could be classified correctly, when comparing the correlation maps. We concluded that crossing data of areas classified by productivity yield and NDVI is valid and that the zoning method for NDVI method is a way to determines areas, but with greater emphasis in areas classified as "Stable High."