Estimativa de biomassa acima do solo para uma área queimada e uma área de corte seletivo no município de Feliz Natal MT por meio de dados LiDAR

Favorite this paper

Remote sensing techniques have aided measurement and estimation of forest area and the identification of deforestation and forest degradation. Light Detection and Ranging (LiDAR) allows mapping the vertical structure of forests and helps obtaining information in areas of difficult access. This study was conducted in two areas in the municipality of Feliz Natal, Mato Grosso, Brazil. The first area (area 1) was burned in 2006, 2008 and 2011, while the second area (area 2) was subjected to selective logging in 2006 and 2007. Both areas were inventoried in the field: area 1 in 2013 and area 2 in 2015, totalizing 27 samples. In addition to the field data, airborne LiDAR data were acquired for the two areas in August 2013. The objective of this study was to use LiDAR data to estimate aboveground biomass (AGB) in these areas and understand the differences in their carbon stocks as a result of fire and selective logging. Structure metrics extracted from the point cloud data were linearly and highly correlated with AGB. The multiple regression model created with the stepwise procedure presented an R2 of 0.96 and a root mean square error of 8.7 Mg/ha (25.3%). Using LiDAR data, it was possible to model the relationship between AGB and LiDAR metrics for areas that have been degraded by fire and selective logging. The results showed a difference in carbon stocks of 15.8% for these areas, indicating that the degradation by fire was considerably more intense in this site.