Northren forests boundary mapping of Iran Using Landsat7 ETM+ data

• Main Author: Khosro Mirakhorlo
• Format: Article
• Language: fas
• Published: Research Institute of Forests and Rangelands of Iran 2003-09-01
• Series: تحقیقات جنگل و صنوبر ایران
• Subjects: Classification; satellite data; boundary mapping; Accuracy assessment; Caspian forests
• Online Access: http://ijfpr.areeo.ac.ir/article_109356_29f61c8f0c0b4831d07f1772d0a09c78.pdf

For sustainable management of the Caspian Forests of Iran, a most appropriate and fastest method of data collection and combination is essential. Combination of Satellite digital data with field data, using Geographic Information Systems, (GIS) could be applied to obtain accurate information for multiple decision making which is influenced by different factors. In this study eight senses of Landsat 7ETM+ data accosted from Caspian Forests of Iran, June and July 2000, were used. Based on OIF index and statistical analysis of the ETM data, color composite 3,4 and 5 were selected for unsupervised classification. The ground observation information were collected from 3860 plots(30*30), using unsupervised map as a primary map. Overall, combining the ETM and the ground data throw the field information, using the supervised classification method, boundary of the Caspian Forests of Iran were achieved in six classes (Dense forest, Sparse forest, Rangeland, Farm land, Water and Bare soil). The classification accuracy assessment showed that the accuracy rates of the dense forest, sparse forast, rangaland, farmland, water and bara soil were 83, 78, 54, 76 and 98 percent, respectively.