Use este identificador para citar ou linkar para este item:
Título: Detection of subpixel treefall gaps with Landsat imagery in Central Amazon forests
Autor(es): Chambers, Jeffrey Q.
Marra, Daniel M.
Ribeiro, Gabriel H.P.M.
Rifai, Sami W.
Niro Higuchi
Roberts, Dar
Assunto: Dinâmica Florestal
Emissão de gases
ISSN: 0034-4257
Revista: Remote Sensing of Environment
Volume: 115
Resumo: Treefall gaps play important roles in both forest dynamics and species diversity, but variability across the full range of gap sizes has not been reported at a regional scale due to the lack of a consistent methodology for their detection. Here we demonstrate the sensitivity of Landsat data for detecting gaps at the subpixel level in the Manaus region, Central Amazon. Spectral mixture analysis (SMA) on treefall gaps was used to map their occurrence across a 3.4 x 10(4) km(2) landscape using the annual change in non-photosynthetic vegetation (Delta NPV) as the change metric. Thirty randomly selected pixels with a spectral signature of a treefall event (i.e. high Delta NPV) were surveyed in the field. The most frequent single-pixel gap size detected using Landsat was similar to 360 m(2), and the severity of the disturbance (Delta NPV) exhibited a significant (r(2) = 0.32, p = 0.001) correlation with the number of dead trees (>10 cm diameter at breast height), enabling quantification of the number of downed trees in each gap. To place the importance of these single-pixel disturbances into a broader context, the cumulative disturbance of these gaps was equivalent to 40% of the calculated deforestation across the Manaus region in 2008. Most detected single-pixel gaps consisted of six to eight downed trees covering an estimated area of 250-900 m(2). These results highlight the quantitative importance of small blowdowns that have been overlooked in previous satellite remote sensing studies. (C) 2011 Elsevier Inc. All rights reserved.
ISSN: 0034-4257
Aparece nas coleções:Coordenação de Dinâmica Ambiental (CDAM)

Arquivos associados a este item:
Não existem arquivos associados a este item.

Este item está licenciada sob uma Licença Creative Commons Creative Commons