Estimativa da área e do tempo de permanência da vegetação secundária na amazônia legal por meio de imagens LANDSAT/TM

AUTOR(ES)
DATA DE PUBLICAÇÃO

2008

RESUMO

- Secondary vegetation, formed by natural regeneration of areas that are deforested and later abandoned, have many functions relevant to the ecology of the landscape in which it is embedded such as fixation of atmospheric carbon, maintenance of the biodiversity and the connection among forest remnants and restoration of soil fertility. The latter function is used in low technology agricultural systems for the reutilization of the land that presents reduction in the carrying capacity after a few years of agricultural use. There are evidences that the biomass, the development rate and the permanence of the secondary vegetation may be associated to the type of land use and also the stage of consolidation of the agricultural frontier. The objective of this dissertation is, through the analysis of these associations by a sampling approach, to estimate the area occupied by secondary vegetation in the Brazilian Legal Amazon (BLA) in 2006, to characterize its spatial distribution and to determine its permanence time. The sampling scheme is based on a stratified approach according to the level of deforestation observed in the 229 TM-Landsat scenes that cover the BLA and it was designed in order to keep the secondary vegetation area estimative error below 5%.In this manner 26 scenes were selected for four dates (1997, 2000, 2003 and 2006), distributed into seven strata according to the level of deforestation (0-1%, 1-5%, 5-10%, 10-20%, 20-30%, 30-55%, more than 55%) in which the areas of secondary vegetation were mapped. A regression model was constructed to estimate the area of secondary vegetation in the remaining images using as independent variables the deforestation area and area represented as hydrography (water bodies and floodplains) in the PRODES map, data of agrarian structure provided by IBGE, indicating the level and type of land use in the scenes and the area of the conservation unities provided by IBAMA, indicating the level of preserved areas within the images. The regression analysis found an adjusted R2 of 0.84 e positive coefficients for the proportion of hydrography in the image (2.055) and for the agrarian structure (0.197), and negative coefficients for the level of deforestation in the image (-0.232) and for the proportion of Conservation Unities in the image (-0.262). Using the multivariate regression model it was estimated an area of 131,873 km2of secondary vegetation for the year 2006. Applying a Monte Carlo simulation, it was estimated an uncertainty of approximately 12,445 km2for the area estimate, which varied between 121,722 and 145,608 km2. The permanence of the secondary vegetation was represented by the half-life of the exponential trend line adjusted to the decay curve of the proportion secondary forest observed in 1997 in the following dates for each of the stratum, weighted by the area of secondary vegetation estimated for the strata. A half-life of 4.9 years was estimated for the secondary vegetation.

ASSUNTO(S)

amazônia remote sensing amazon (region) uso da terra land use statistical models sensoriamento remoto vegetação secundária secondary vegetation modelos estatísticos

ACESSO AO ARTIGO

Documentos Relacionados