|
|
|
Abstract
Land Use Changes Detection and Spatial Distribution with Using Comparison of the Potential of the ETM+ and ASTER imagery for Forest Area Mapping, Case Study Farim Drainage Basin
Ataollah Kelarestaghi
Faculty of Natural Resources, University of Tehran,
Iran
Email: ata_kelarestaghi@hotmail.com
Hassan Ahmadi
Professor
Faculty of Natural Resources
Univ. of Tehran
Mohammad Jafari
Professor
Natural Resources Faculty
University of Tehran
Jamal Ghodosi
Academic Staff
Research Institute of soil conservation and watershed management
Using satellite images can extract various, update, cheaper and repetitious information. Now, with the concern about many satellite images, evaluation of these images for different resources inventory is a requisite. The purpose of this study is evaluation and comparison of the potential of the Landsat 7 ETM+ and ASTER images for forest land use area mapping in Farim Drainage Basin in Mazandaran Province. Qualitative and quantitative analysis implied that no radiometric and striping error was found. These images were geo referenced with using 32 ground control point and the RMS error was 0.71 pixels. Then, different suitable spectral transformation such as rationing, PCA, Tasseled Cap transformation and Fusion were performed on the images. Image classification were done using supervised classification maximum likelihood and minimum distance classifier utilizing original and synthetic bands resulted from diverse spectral transformation and the forest area was separated from non forest area. The ground truth map was prepared through fieldwork using systematic random sampling. Results from accuracy assessment of the forest maps overlaid with ground truth map were shown; ETM+ and ASTER images concluded an overall accuracy of 93.02% and 95.4% respectively. Key Words: satellite images, forest land use map, ground truth map, accuracy assessment
|
|
|