USE OF REMOTE SENSING IMAGES TO ESTIMATE DENDROMETRIC CHARACTERISTICS OF EUCALYPTUS FORESTS / USO DE IMAGENS DE SENSORES REMOTOS NA ESTIMATIVA DE CARACTERÍSTICAS DENDROMÉTRICAS DE POVOAMENTOS DE EUCALIPTO

AUTOR(ES)
DATA DE PUBLICAÇÃO

2008

RESUMO

The Forest Management Information System is an integrated system which can be used to support the planning, implementation and monitoring of forest management activities. Beyond collected field information, geoprocessing and remote sensing systems are essential for that management type. One of the goals of this dissertation is to develop an analysis methodology for data analysis (from a forestry database and extracted from remotely sensed digital images) that enhance the information generation capability to the forestry planning and operational control. During this work, we had access to forestry databases, inventories and ex works wood volumes provided from Aracruz about forests located in RS, as well as digital images provided by CBERS-2 satellite. We measured 72 characteristics acquired from forestry images. After correlation analysis, only 28 were considered for later analysis. The first part of this work deals with data organization in such a way as to correlate them with the images. A method to identify harvested areas and another for time correlation are needed to allow usage of data collected over two years which is related to only one image. The next part involves simple and multiple regression analysis. We were unable to find a single parameter to estimate volume or age by itself. Conversely, multiple regression models achieved correlation coefficients up to 99% and the root mean squared error was down to 20m2/ha of wood volume.

ASSUNTO(S)

estimativa de volume de madeira eucalyptus sp. índice de vegetação remote sensing vegetation index wood volume estimate engenharia de producao sensoriamento remoto eucalyptus sp.

ACESSO AO ARTIGO

Documentos Relacionados