Dados multi e hiperespectrais da cultura da soja (Glycine Max L.) e sua relação com doses de gesso e calcário no solo

Abstract

In Brazil the need for soil correction is a recurring problem and constitutes a limiting factor of productivity. Therefore, it is recommend the application of gypsum and lime to rise the soil pH and allow better agricultural use. Thus, geotechnologies have been increasingly use in the agriculture system production. The remote sensing provides faster and low cost results in comparison to conventional methods. This study aimed to evaluate the spectral behavior of soybean plants cultivated under different gypsum and lime doses in soil, by field spectroradiometry and digital images obtained by UAV. As to specific goals: identify which soybean crop development stage better express the spectral behavior due to gypsum and lime application; evaluate the applicability of vegetation index generated from red, green and near infrared bands, due to the evaluation of soybean plants under different gypsum and lime doses; establish the correlation between information obtained by field hyperspectral sensor and by UAV digital images, and set up the best development crop stage and the best vegetation index to these evaluations. The experiment was installed in field with experimental design of randomized blocks, with plots subdivided in 4x4 factorial scheme with 4 replications. The plots correspond to lime doses (0, 2.60, 5.35 e 8.10 ton/ha) and the split plots rates of gypsum (0, 4, 8 e 12 ton/ha). Two remote sensors collected the spectroradiometry data: spectroradiometer, with a 3 nm between 350 and 1400 nm and 30 nm between 1400 e 2500 nm spectral resolution at ground level; and two multispectral Vis/NIR digital cameras, with radiometric resolution of 8 bits, equipped with optical filters to select the incident radiation, attached to UAV at aerial level. The obtained data were processed by the Excel and SPRING software and six vegetation index were generated (NDVI, GNDVI, RNIR, GNIR, RVI e GRVI). The evaluations occurred in eight different periods: 27, 38, 48, 58, 63, 68, 83, 96 days after emergence (DAE) using the spectroradiometer. Besides that, in the 68, 83, 96 DAE also used multispectral cameras attached to UAV. The results show no significant spectral behavior of soybean crop to the gypsum application. On the other hand, lime application may be possible to be detect by remote sensing. Stands out that the lime influence was expressed at the R5.2 soybean development stage, with R² superior than 0.85 (NDVI, RNIR and RVI). Few treatments were significant to t test at Pearson correlation and many negative correlations were obtained when the ground and aerial sensors were compared