Search

The Brazilian Amazon forest and cerrado savanna encompasses a region of enormous ecological, climatic, and land-use variation. Satellite remote sensing is the only tractable means to measure the biophysical attributes of vegetation throughout this region, but coarse-resolution sensors cannot resolve the details of forest structure and land-cover change deemed critical to many land-use, ecological, and conservation-oriented studies. The Carnegie Landsat Analysis System (CLAS) was developed for studies of forest and savanna structural attributes using widely available Landsat Enhanced Thematic Mapper Plus (ETM+) satellite data and advanced methods in automated spectral mixture analysis. The methodology of the CLAS approach is presented along with a study of its sensitivity to atmospheric correction errors. CLAS is then applied to a mosaic of Landsat images spanning the years 1999–2001 as a proof of concept and capability for large-scale, very high resolution mapping of the Amazon and bordering cerrado savanna. A total of 197 images were analyzed for fractional photosynthetic vegetation (PV), nonphotosynthetic vegetation (NPV), and bare substrate covers using a probabilistic spectral mixture model. Results from areas without significant land use, clouds, cloud shadows, and water bodies were compiled by the Brazilian state and vegetation class to understand the baseline structural typology of forests and savannas using this new system. Conversion of the satellite-derived PV data to woody canopy gap fraction was made to highlight major differences by vegetation and ecosystem classes. The results indicate important differences in fractional photosynthetic cover and canopy gap fraction that can now be accounted for in future studies of land-cover change, ecological variability, and biogeochemical processes across the Amazon and bordering cerrado regions of Brazil. [ABSTRACT FROM AUTHOR]

The Amazon rain forest constitutes one of the major global stocks of carbon. Recent studies, including the last Intergovernmental Panel on Climate Change report and the Coupled Climate Carbon Cycle Model Intercomparison Project, have suggested that it may reduce in size and lose biomass during the twenty-first century through a savannization process. A better understanding of how this ecosystem structure, dynamics, and carbon balance may respond to future climate changes is needed. This article investigates how well a fully coupled atmosphere–biosphere model can reproduce vegetation structure and dynamics in Amazonia to the extent permitted by available data. The accurate representation of the coupled climate–biosphere dynamics requires the accurate representation of climate, net primary production (NPP), and its partition among the several carbon pool components. The simulated climate is validated against precipitation (within 5% of four datasets) and incident solar radiation (within 7% of observations). The authors also validate (i) simulated land cover, which reproduces well the observed patterns; (ii) NPP, within 5% of observations; and (iii) respiration rates, within 15% of observations. The performance of simulated variables that depend on carbon allocation, like NPP partitioning, leaf area index, and aboveground live biomass, although good on a regional mean, is significantly low when spatial patterns are considered. These errors may be attributed to fixed carbon allocation and residence time parameters assumed by the model. Carbon allocation apparently varies spatially, and to simulate this spatial variability is quite a challenge. [ABSTRACT FROM AUTHOR]

The Brazilian government annually assesses the extent of deforestation in the Legal Amazon for a variety of scientific and policy applications. Currently, the assessment requires the processing and storing of large volumes of Landsat satellite data. The potential for efficient, accurate, and less data-intensive assessment of annual deforestation using data from NASA’s Moderate Resolution Imaging Spectroradiometer (MODIS) at 250-m resolution is evaluated. Landsat-derived deforestation estimates are compared to MODIS-derived estimates for six Landsat scenes with five change-detection algorithms and a variety of input data—Surface Reflectance (MOD09), Vegetation Indices (MOD13), fraction images derived from a linear mixing model, Vegetation Cover Conversion (MOD44A), and percent tree cover from the Vegetation Continuous Fields (MOD44B) product. Several algorithms generated consistently low commission errors (positive predictive value near 90%) and identified more than 80% of deforestation polygons larger than 3 ha. All methods accurately identified polygons larger than 20 ha. However, no method consistently detected a high percent of Landsat-derived deforestation area across all six scenes. Field validation in central Mato Grosso confirmed that all MODIS-derived deforestation clusters larger than three 250-m pixels were true deforestation. Application of this field-validated method to the state of Mato Grosso for 2001–04 highlighted a change in deforestation dynamics; the number of large clusters (>10 MODIS pixels) that were detected doubled, from 750 between August 2001 and August 2002 to over 1500 between August 2003 and August 2004. These analyses demonstrate that MODIS data are appropriate for rapid identification of the location of deforestation areas and trends in deforestation dynamics with greatly reduced storage and processing requirements compared to Landsat-derived assessments. However, the MODIS-based analyses evaluated in this study are not a replacement for high-resolution analyses that estimate the total area of deforestation and identify small clearings. [ABSTRACT FROM AUTHOR]