Avaliação de técnicas geoestatísticas no inventário de povoamentos de Tectona grandis L.f.

• Autor(a) principal: Santana, Rogério Alves
• Data de Publicação: 2011
• Tipo de documento: Dissertação
• Idioma: por
• Título da fonte: LOCUS Repositório Institucional da UFV
• Texto Completo: http://locus.ufv.br/handle/123456789/4044
• topic: Inventário florestal; Dependência espacial; Predição; Forest inventory; Spatial dependence; Prediction; CNPQ::CIENCIAS AGRARIAS

The general objective of this work was to evaluate if estimates of wood volume obtained from geostatistics techniques are more accurate than those obtained from classical statistical techniques.The study was conducted in stands of Tectona grandis L.f. with the dendometric volume trait measured in 101 plots laid along a sampling spacing with UTM coordinates, distributed in 17 plots with a total area of 391.87 ha. By adjusting the experimental semivariogram model, index of spatial dependence and cross-validation statistics it was obtained two estimates of the forest inventory: one by using the classical statistics and the other by using geostatistics. Methodology of systematic sampling with intensity of sampling of one plot for every 4 ha was used in the conduction of the classical estimation. A semivariogram model, used to assess the structure of spatial dependence of the variable volume per ha in the study area was adjusted in the prediction by geostatistcs. Then, block kriging was used to obtain predictions of the average volume of each plot. Comparison of the two methods of estimation was made based on the sampling error, in which a minor error with geostatistics was found. Therefore, it is concluded geostatistics should be used for forest inventory when there is spatial dependence structure in the region under study inasmuch as it provides more accurate predictions.