Quantificação de volume total em plantios de Pinus taeda L. por inventário florestal tradicional e por Tecnologia LiDAR
Visualizar/ Abrir
Data
2021Autor
Siqueira, Maria Lucia Marques, 1998-
Metadata
Mostrar registro completoResumo
Resumo : O inventário florestal pode ser conduzido por meio do método tradicional, adotado predominantemente, ou por sensoriamento remoto, utilizando a tecnologia LiDAR. As empresas de base florestal vêm buscando alternativas, como o uso do LiDAR, para alcançar bons resultados, com redução de custo e tempo no inventário. O objetivo deste trabalho é comparar as estimativas de volume por inventário tradicional e LiDAR para a espécie Pinus taeda. Foram realizadas 2 coletas de dados em plantios localizados em Sengés, PR. A primeira foi realizada por inventário florestal tradicional, no período de julho a outubro de 2018, onde 57 parcelas circulares foram instaladas em plantios de 6 a 14 anos. Em outubro de 2018, a outra coleta foi realizada por meio de voo com LiDAR, consistindo no sobrevoo de uma área de aproximadamente 1200 ha, e abrangendo todas as 57 parcelas instaladas manualmente, as quais foram utilizadas para o processamento. As variáveis mensuradas no inventário tradicional foram circunferência e altura total, e o volume total de cada parcela foi estimado por funções de afilamento. As variáveis registradas pelo LiDAR foram as coordenadas dos pontos X, Y e Z, intensidade do pulso laser, número do retorno e quantidade de retornos. Para a estimativa de volume total, a partir das variáveis registradas pelo LiDAR, foi utilizado o método de análise de componentes principais, e em seguida o ajuste de regressão linear, sendo a variável resposta o volume (m3 por área de parcela) e as variáveis independentes os componentes principais. Na comparação dos inventários tradicional e LiDAR, o teste de Qui-quadrado não foi significante, o valor do coeficiente de correlação de Pearson foi alto (0,96), e as estatísticas RMSE e Bias foram baixas. A análise de componentes principais se mostrou apropriada para a redução no número das variáveis obtidas pelo LiDAR, e para o ajuste de modelos de regressão, obtendo assim as estimativas de volume total. Abstract : The forest inventory can be conducted through the traditional method, or by remote sensing, using LiDAR technology. Forest-based companies have been looking for alternatives, such as using LiDAR, to achieve good results, with reduced time and costs of the inventory. The objective of this study is to compare the volume estimates between traditional inventory and LiDAR for the species Pinus taeda. Two sets of data were collected at the plantations located in Sengés, PR. The first was collected through a traditional forest inventory, from July to October 2018, where 57 circular plots were installed in plantations for 6 to 14 years. In October 2018, the other collection was carried out by flying with LiDAR over an area of approximately 1200 ha, and covering all 57 plots installed manually, which were used for processing. The variables measured in the traditional inventory were diameter and total height, and the total volume of each plot was estimated utilizing tapering functions. The variables recorded from LiDAR were the coordinates of points X, Y and Z, intensity, number of returns and return number. For the estimate of total volume, considering the variables recorded from LiDAR, the principal component analysis method was used, and then the linear regression was adjusted: the dependent variable was volume (m3 per plot) the principal components as independent variables. When comparing traditional and LiDAR inventories, the Chi-square test was not significant, Pearson's correlation coefficient was high (0.96), and the RMSE and Bias statistics were low. Principal component analysis proved to be appropriate for reducing the dimension of the variables obtained from LiDAR, and for adjusting regression models, thus obtaining the estimates of total volume.
Collections
- Engenharia Florestal [253]