Estimativa da altura do dossel forrageiro por meio do mapeamento com VANT

Abstract

The improving pasture management techniques in integrated systems, or even intensive production, aims to ensure the sustainability of cattle ranching in biomes such as the Cerrado. Vegetation cover sensing helps to distinguish different temporal changes in ecosystems by quantifying reflected electromagnetic energy (REM). In these relationships, vegetation indices (VIs) are used to determine production factors in various crops, including pasture. This research applied the mapping of vegetation indices - VIs (Leaf Area Index - LAI, Green Leaf Index - GLI and Visible Atmospheric Resistance Index - VARI) to pasture management systems using an Unmanned Aerial Vehicle (UAV) in order to estimate the height of the forage canopy. Flights were made over the areas of interest to capture images with RGB and multispectral optical sensors. After processing the images in Pix4Dmapper software, VIs mosaics of forage height were generated. The maps were adjusted using samples of the average height of the forage, using an indirect method that measures the actual height of the canopy with a graduated stick specifically for height management.Comparative statistical tests were applied to validate the stratification into four classes, in which the optimum predictability of the VIs tested was verified. The results confirmed that UAV mapping, calibrated with samples of the actual height of the forage canopy, can estimate its height per stratified area. The mosaics measured and recorded gaps in the pasture that are generally not quantified in simple average sampling, where exposed soil and/or sparse forage appears. They also make it possible to visually verify, with the choropleth’s maps, the general state of pasture occupation on the terrain in terms of the weighted average height of the forage. The results show that the UAV mapping of the IAF and GLI studied, was efficient in estimating the height of the forage canopy and can be applied to the management of productive systems, with a view to automation and verification of the general state of the pasture.