Robert Zomer
Center for Spatial Technologies and Remote Sensing (CSTARS)
Dept. of Land, Air, and Water Resources
University of California, Davis
email: rjzomer@ucdavis.edu
Return to
Robert Zomer Homepage
Establishing Baseline Data for Land-Use and Cover-Change Analysis: A cost-effective methodology for remote and rugged terrain areas based upon satellite remote sensing data.
Highly complex mountainous landscapes in the Himalaya are characterized by spatial, biological and cultural diversity. Lack of accurate or timely data is a major constraint for managers, policy makers, and researchers in rugged or remote mountainous regions. Establishing baseline data for land use and cover change (LUCC) analysis, as well as for Geographic Information Systems (GIS) modeling capability, is essential to identification and understanding of change processes. Change detection and modeling of these processes within a GIS environment based upon satellite remote sensing may be a cost-effective strategy in remote mountainous terrain.
Due to the complex physiography of mountainous landscapes, accurate terrain modeling is an essential prerequisite to a wide variety of landscape level analyses, e.g. extent and distribution of terrain and landscape features, watershed and stream networks, and vegetation, landuse, and habitat types. The use of satellite remote sensing data for LUCC analysis in steep and highly heterogeneous terrain is evaluated in a case study of the Makalu-Barun National Park and Conservation Area (MBNPCA) of eastern Nepal.
The MBNPCA encompasses more than 8000 meters of vertical relief within 2300 sq. km. of protected area, with bio-climatic zones ranging from tropical to alpine. Although a great diversity of vegetation types are found within the area, including significant stands of various closed canopy late-successional forest types, substantial pressures on the natural resources of the area are evident. Agricultural practices of the 32,000 inhabitants of the area, primarily subsistence agriculturalists, range from terraced hill farming to swidden practices and high altitude trans-humance pastoralism. Although tourist visitation to the Park is currently relatively low, substantial increases in tourism are expected as access and infrastructure improve.
As part of an overall methodology for establishing baseline datasets for terrain and landscape analysis, a set of precision geocorrected and orthorectified digital base maps have been produced of the study area from satellite imagery. The maps are intended to facilitate and georeference both further research and classification of the satellite imagery. An overview of the larger research project includes:
1.) Extensive field survey and ground truthing conducted from 1991 to 1995. Forest and vegetation communities were sampled and observations were georeferenced using Global Positioning System (GPS) receivers. Multivariate statistical analysis identified significant relationships between community composition, distribution, and several site and topographic characteristics, including disturbance levels.
2.) A precision geocorrected DEM extracted stereoscopic SPOT imagery is compared with a DEM generated from digitized 1:250k contour vectors. Both Landsat TM and SPOT imagery were orthorectified and precision geocorrected using a corrected DEM. The higher spectral resolution Landsat TM data were merged with higher spatial resolution SPOT data to produce both a digital and hard copy satellite photomap. This base map is useful both in the field, and as a visual aid for image classification.
3.) Vegetation and land use types will be delineated from the satellite imagery based upon spectral characteristics and ancillary data, and utilizing topographic normalization techniques to compensate for high relief and topographic shadows. Results allow GIS modeling capability for LUCC and wildlife habitat analysis. By georeferencing these efforts to regional databases such as the ICIMOD's Nepal 1:250 k GIS Database, results can be integrated into larger regional or global change models.
For more information or comments, send email to: rjzomer@ucdavis.edu