Uso do LIDAR (Light Detection and Ranging) para estimação da altura de árvores em povoamentos de eucalipto

• Autor(a) principal: Pires, Jonas Machado
• Data de Publicação: 2005
• Tipo de documento: Dissertação
• Idioma: por
• Título da fonte: LOCUS Repositório Institucional da UFV
• Texto Completo: http://locus.ufv.br/handle/123456789/3007
• topic: Árvore; Medição florestal; Sensoriamento remoto; Eucalipto; Economia florestal; Trees; Forest mensuration; Remote sensing; Eucalypt; Forest economy; CNPQ::CIENCIAS AGRARIAS::RECURSOS FLORESTAIS E ENGENHARIA FLORESTAL::MANEJO FLORESTAL

The total height (tH) of a tree has high correlation with its volume, being essential for its calculation in individual level or stand of trees reserved for cutting the youthful phase, it is the only variable to be measured for the accompaniment of the growth of slices and experimental parcels. The ordering productive capacity allows operational planning better, through the knowledge of the productivity placed in each slice. The recurrent equations of the relations between the dominant height (dH) and age (A) are used to get the productive capacity. However, the measurement of the height in the field demands involvement of resources as, staff, vehicle, equipment and time. The technology Lidar (Light Detection and Ranging) allows to measure with accuracy the total height (tH) of individual trees in standing of eucalyptus reserved for cutting. This study was carried through in an area of approximately 306 ha, in commercial planted area of eucalyptus, of the Aracruz company Cellulose S/A, located in the region of São Mateus, in the north of the Espírito Santo, South latitude 18º 05', longitude 39º 30' West of Greenwich. The system Lidar was installed in an airplane that flew over the area to an average height of 1.000 m, with the width of the band of sweepings of 650 m, allowing accuracy horizontal line of 50 cm and vertical line of 15 cm. The general aim of the study was to elaborate a methodology for automatic extraction of total height (Ht) of individual trees in standing of eucalyptus reserved for cutting, with raised data for the Lidar sensor. Specifically, one searched esteem the total height (Ht) of individual trees in planted area of eucalyptus, with Lidar data, using the average total height Lidar (Ht lidar) for ordering productive capacity. In the estimate of the total height of individual trees (tH Lidar) with Lidar data, had been measured in the field 46 trees in 24 different parcels. One applied test t for two dependent samples, and as resulted, the tH Lidar, the tH Obs and the tH Suunto, with averages 25,8m, 25,1m and 25,4m, respectively, not differing among themselves 1% from probability. Evaluating prompt values, it is observed that only four points had had superior difference the 3,0 m, and these superestimates for Lidar had always occurred that the showed tree was a suppressed individual. This difference can be attributed to an imperfection in the identification of suppressed trees next the dominant trees, or same to an error of measurement in the field. In the ordering of productive capacity using the average total height Lidar (Ht lidar) had been launched thoroughly fifty and seven circular parcels: 52 with 360 m2 and 5 with 507m2, these last installed in roughed-hew slices. In each parcel, average dominant height (Hd obs). By means of an automatic ordering, measured it calculated it average total height (Ht lidar) for parcel. The coefficient of correlation between (Ht lidar) and (Hd obs), was 0,9533 and the hypothesis of identity between the two vectors heights was rejected 1% of probability (F(H0) = 18,2572*). Despite this significance of test F, when analyzing the trends (Figures 20 and 21), consistency between the methods is verified. Of the point of view of the forest measuring, this fact indicates the possibility of use of gotten information of height with Lidar in studies of ordering of productive capacity. Given the limitation of the data, this possibility could not be proven in this study.