Drones como ferramenta de gestão florestal: desempenho dos índices IRAV (Índice Resistente à Atmosfera na Região Visível) e ITV (Índice Triangular Verde) na estimativa de volume de Eucalyptus.

Abstract

This work aimed to associate reflectance data of RGB aerial photographs captured by drone and to comcant them with wood volume values using two vegetation indices (IRAV and ITV). The study was conducted in Macaíba-RN in the experimental eucalyptus plantation that is part of the TECHS-UFRN research program. The captured photographs were processed using the QGIS software, generating a map for each index evaluated. To compare with the volume values of the trees of each plot, a value of the index of the average of the pixels in the radius of 1.5 m of each tree was generated, which, after the flight, were felled for cubation. The correlation analysis was performed by multivariable linear regression, with the index value being the dependent variable, and the volume value being the independent variable. The IRAV index showed the highest correlation with the volume values of the trees, accounting better statistically than the ITV index, especially the C3 plot (E. grandis x E. camaldulensis), was the one that presented the best correlation with volume (m³), with a value of r = 0.60. The ITV index showed lower correlation with the volume of the trees, obtaining a significant difference between the two indexes evaluated. Nonlinear regression analysis was performed where regression models that best represented the indices in each genotype were chosen by comparison by comparison via AIC. The chosen regressions were applied to the index map for comparison with the volumetric production values (m³/ago) through the segmentation process. The two indices showed a positive relationship with the values estimated by the direct method, showing the highest volumetric production in the most densespacings. The results obtained with this study show the potential of the use of drones monitoring tool to estimate forest productivity, through regression analysis of vegetation indices obtained by aerophotos.