Application of hyperin EO-1 data in eucalyptus spp. plantations / Aplicação de dados hyperion EO-1 no estudo de plantação de euclalyptus spp

AUTOR(ES)
DATA DE PUBLICAÇÃO

2008

RESUMO

The access to orbital hyperspectral data has offered new perspectives to both academic and corporative sectors in terms of their potential to solve some demands in the forestry science. The main objective here was to evaluate the potential and constrains of the hyperspectral data from the Hyperion EO-1 sensor in the Eucalyptus spp. stands study. Two approaches were adopted. The first one was based on a species classification procedure and the second one included evaluations of the relationship between the spectral data and the stand logging. It was defined some multiple linear regression models as descriptors of that relationship. In both approaches the relief effect on the Eucalyptus spp. canopy reflectance was taken into account. The Hyperion EO-1 data were converted to BRF surface values, which were considered the basis of the entire work. The study area was located in Capão Bonito city (São Paulo State, Brazil) and included Eucalyptus spp. stands managed by the Votorantin Papel e Celulose company. Multi-spectral data of TM/Landsat and HRV/SPOT sensors were simulated from the Hyperion EO-1 ones. In the classification evaluation it was evident the canopy architecture effect in the stand age differentiation despite the species identification. This identification has been improved when the vegetation anisotropy was taken into account. In volume x radiometric data relationship evaluation the vegetation indices calculated from the hyperspectral data increase the R2 values that were interpreted as an advantage comparing to the multi-spectral data in this kind of evaluation.

ASSUNTO(S)

caracterização espectral anisotropia da vegetação volume de madeira biophysical parameters estimation vegetation mapping mapeamento da vegetação quantificação de parâmetros biofísicos.spectral characterization vegetation anisotropy logging

Documentos Relacionados