Dados de sensor LiDAR na identificação e caracterização de clareiras e estradas na floresta Amazônica

• Ano de defesa: 2019
• Autor(a) principal: Favarin, José Augusto Spiazzi
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Santa Maria; Brasil; Recursos Florestais e Engenharia Florestal; UFSM; Programa de Pós-Graduação em Engenharia Florestal; Centro de Ciências Rurais
• Palavras-chave em Português: Abertura em dossel florestal; Escaneamento laser; Floresta tropical; Densidade de pontos; Forest canopy openning; Laser scanning; Tropical forest; Point density; CNPQ::CIENCIAS AGRARIAS::RECURSOS FLORESTAIS E ENGENHARIA FLORESTAL
• Link de acesso: http://repositorio.ufsm.br/handle/1/20243

Tropical forests have great importance in maintaining biodiversity, yet suffer from illegal logging and deforestation. On the other hand, sustainable forest management practices contribute to the rational use of forest resources. In this way, it is salutary that new technologies and methods are used to follow these activities. This study aimed to identify and delimit gaps in tropical forest from LiDAR data with the use of different density of returns. The study area is located in Fazenda Cauaxi, munipality of Paragominas-PA, in which the forest management activity is carried out. Forest data were obtained from forest inventory from 22 plots of 20 x 500m, and trees with DBH greater than or equal to 35cm were measured. LiDAR data were obtained on a flight that covered an area of 1.216ha composed of 20 scenes consisting of a cloud of points. A minimum gap area of 34m² was defined, based on the canopy rays measured in the forest inventory. The cloud of points was processed in the FUSION/LDV and a segmented raster was obtained in gap areas for each of the density of returns tested (37, 28, 18, 9, 4 and 1ppm², corresponding to the treatments of that study). The areas were grouped into three size classes, Class 1, Class 2 and Class 3 (34 – 149, 150 – 399 and greater than or equal 400m², respectively). The roads were identified by the spatial distribution pattern in the area with the aid of the DTM. The statistical analysis as oerformed in the R and the Kruskal-Wallis test was used to evaluated if there was a difference between the evaluated treatments, which did not show any significance at the 0.05 level, so the densities did not differ. Due to the identification of roads, the areas were reclassified into Small gaps, Large gaps and Roads. The number of areas in small gaps varied between treatments from 80.7 to 87.4% of total gaps, which is expected for areas smaller than 150m², in relation to the area, the variation was from 50.4 to 62,3%. By accounting for areas of gaps and roads, the Roads had the greatest coverage in the study area, varying between treatments from 68.3 to 55.5%. It was possible to infer that some gaps were opened by the activity of selective extraction of wood due to the spatial arrangement of the areas. Rare gaps were not identified, areas greater than 400m². Therefore, working with the reduction of the density of points did not affect the identification and delimitation of gaps in tropical forest. LiDAR technology has proven to be an effective tool for monitoring forest canopy disturbances. Thus, it can cover its application to the monitoring of the activity of forest management, deforestation and illegal logging in the Amazon.