Science Service System

Summary of Proposal LAN0103

TitleAssessment of land cover and canopy structure of a tropical ecosystem in Central Sulawesi, Indonesia
Investigator Erasmi, Stefan - Georg-August-Universität Göttingen, Inst. of Geography, Dept. of Cartography, GIS and Remote Sensing
Team Members
Dr. Ardiansyah, Muhammad - Institut Pertanian Bogor (IPB), Agricultural University of Bogor, Indonesia, Dept. of Soil Sciences
SummaryThe two central objectives of the proposed TerraSAR-X project are 1) to evaluate the suitability of spaceborne X-band SAR data for tropical land cover monitoring and 2) to assess the potential of TerraSAR-X data for the estimation of canopy structure variables in tropical forest and agroforestry systems. As sub-goals, the project intends to analyze the synergy of high resolution optical and X-Band SAR data for mapping land cover and crop phenology. Concerning the estimation of structural parameters of tropical forests, it is aspired to use both full-polarimetric X-Band data (TerraSAR-X) and L-Band data (ALOS PALSAR) synergistically.
The intended work is part of the multi-disciplinary, bi-national (Indonesian-German) research project "Stability of rainforest margins in Indonesia" (STORMA, SFB-552). The STORMA project examines the processes of destabilisation at the forest border and analyses factors and processes that may conserve forest systems.
The research area for the acquisition of TerraSAR-X data covers a region that is situated in the province of Central Sulawesi, Indonesia. The project area contains five districts of the province of Central Sulawesi (Sulawesi Tengah) and surrounds the Lore Lindu National Park (LLNP). More than 80 % of the project area are covered by natural rain forest (lower and upper montane rain forest). The focus within the currently ongoing research within the project area is on the investigation of the rainforest margin, hence the transition zone between natural forest and landscapes that are dominated by human acitivities. This means that the border of the LLNP roughly represents the data frame for the acquisition of TerraSAR-X data. The land use in the vicinity of Lore Lindu National Park is dominated by paddy rice cultivation in the valley bottom and coffee and cacao cultivation on sloping land.
Data acquisition will be focused on four sub-areas within the project area using the stripmap imaging mode (SM, 15x50 km). The projects aims to employ Multi Look Ground Range Detected (MGD), spatially enhanced (SE), quad-polarization products recorded at relatively steep (20°) incidence angles.
The land cover analysis will evaluate the hypothesis that full polarimetric TerraSAR-X band data will substantially improve the accuracy of crop type mapping at the rainforest margin. The combined use of multi-temporal backscatter and mono-temporal reflectance data for an enhanced mapping of annual cropping systems and rice crops will be evaluated. The methods used for the integration of the different data sources consist of state of the art image merging techniques as well as decision tree classifiers.
Further to the mapping of land cover, the computation of structure variables from the satellite data will be investigated using fully polarimetric X- and L-band SAR backscatter data that will be retrieved from TerraSAR-X and ALOS-PALSAR. Concerning the well known characteristics and limitations of X- and L-band backscatter data (e.g. penetration depth), the X-band data will be evaluated with regard to 2D-canopy surface structure (roughness) whereas the analysis of the L-band data focuses on the modelling of 3D-stand structure parameters. The combined use of X- and L-band satellite SAR data comprises an innovative approach in tropical ecosystem analysis and is expected to deliver information about the structure of tropical forests that are complementary to the estimates from optical satellite data.
The outcome of the analysis will contribute to the overall subject of the multi-disciplinary research project and will improve the possibilities to characterize the spatial and temporal variability in a tropical ecosystem at the rainforest margin in Central Sulawesi.

Back to list of proposals

© DLR 2004-2016