Correlação linear entre os índices de vegetação NDVI e NDRE com a produtividade do milho segunda safra

Abstract

Due to the agribusiness demand, the rural producers need to increase their level of expertise, mainly with the maize crop, that its extremely influenced by weather conditions. In this way, the producer seeks to adapt to the new remote sensing technologies, which can also be used for small and medium-sized agricultural properties. Therefore, the aim of this paper is to estimate the linear correlation between the yeld of interim-harvest and the NDVI (Normalized Difference Vegetation Index) and NDRE (Normalized difference red edge) indexes, using free resources as free-code software and sattelite imagery. Productivity data and Sentinel-2 imagery of a property in the Tabaporã municipality, at northwest of Mato Grosso, were used. It was used the Quantum Gis software to create the productivity map, the atmospheric correction of images (through the Dark Object Subtraction method) and the vegetation indexes (NDVI and NDRE). For comparison purpose, a pixel-to-pixel extraction in a amostral regular grid was done. All statistic calculation were perfomed in Microsoft Excel. It was verified that between the two indexes, the NDRE was more representative than the NDVI, when there is a high spatial variability of the productivity, there is a trend to higher correlation coefficients, being also strongly related to the low productivity, evidencing the relation between the vegetation indexes and corn sanitation conditions. Through these results, it was observed that with the use of free resources it is possible to obtain and process data for the generation of productivity maps and vegetation indexes.