The Western Amazonian region is considered a biodiversity hotspot because of its high biological diversity and the fast growing anthropogenic pressure on the landscape, including forest conversion to agriculture, exploitation of mineral resources (e.g. oil) and more frequent incidences of forest fires. These characteristics of the region demand for instrumental tools supporting decision making for prioritizing biological inventories, conservation and sustainable utilization. The recent availability of high resolution satellite imagery provides us with more data and information over extensive areas on the diversity of biological, biophysical, land-use conditions at the landscape scale. The mapping of tree diversity, applying satellite imagery, at the species or groups of species level, gives us valuable insights in ecological processes and its spatial characteristics (e.g. heterogeneity, spatial distribution). Especially in the tropical forest environment, this is a challenge because of its high levels of tree diversity and its complex forest structures. This study presents an approach of tree diversity mapping, using a combination of the high resolution satellite sensor data of Quickbird (0.6 ~ 2.8 m) and aerial photos of different resolutions (0.03 ~ 0.3m) taken from airborne platforms (Helium balloon and airplane) in the tropical lowland forests. Relationships between image objects identified in the sequence of multi resolution imagery and ground collected taxonomic plus tree crown structure data are being studied. Results of this research should facilitate the identification of areas with high(er) priority for inventories, conservation and sustainable use planning. This research is also part of the HERB (Hydrology, Ecology and Regional Biodiversity) project.
Submetido por Edwin Keizer em 24-MAR-2004
Tema Cient�fico do LBA: LC (Mudan�as dos Usos da Terra e da Vegeta��o)