Amazon-wide Forest Gap Fraction and Selective Logging from Satellite Analyses
Gregory
Paul
Asner, Carnegie Institution, gpa@stanford.edu
(Presenting)
José
Natalino Macedo
Silva, EMBRAPA, natalino@cpatu.embrapa.br
Mercedes
Maria Cunha
Bustamante, Universidade de Brasilia, mercedes@unb.br
Michael
M.
Keller, USDA Forest Service, michael.keller@unh.edu
Amanda
Naslund
Cooper, Carnegie Institution, acoop@stanford.edu
Lydia
Olander, Carnegie Institution, lolander@globalecology.stanford.edu
David
E.
Knapp, Carnegie Institution, dknapp@globalecology.stanford.edu
Selective logging and forest disturbance can now be quantified throughout the Brazilian Amazon using multiple high spatial resolution satellite sensors (Landsat ETM+, EO-1 ALI, ASTER) and non-linear, Monte Carlo spectral mixture analysis. We have extensively tested the method in different forest environments in central and eastern Para, Northern Mato Grosso, and southern Acre in logged forest of 0.5 to 4.5 years post-harvest. We have also quantified the forest gap variation of mature canopies throughout nearly 5 million square kilometers of the Amazon basin. Results indicate that selective logging can be detected and quantified with high precision and accuracy. They also indicate that forest types and disturbance conditions can be measured on an operational basis. Forest gap fraction varies with canopy architecture, soil conditions, and topographic position. These findings will allow improved quantification of forest carbon stocks, productivity, fire susceptibility and biogeochemical processes throughout the Amazon basin.
Submetido por Gregory Paul Asner em 27-FEV-2004
Tema Científico do LBA: LC (Mudanças dos Usos da Terra e da Vegetação)