Document Type : Research Paper
Authors
1 Professor, Faculty of Natural Resources, University of Tehran
2 Assistant Professor, Faculty of Natural Resources, University of Yazd
3 Research Instructor, Research Institute of Forests and Rangelands
Abstract
This study was aimed to assess the applicability of LANDSAT ETM+ satellite images for estimating vegetation production and cover. The images were digitized using topographic maps and geometrized in 1:25000 scales. Required processes such as spectral ratio measurement and vegetation indices were applied on the images. Collection was carried out for vegetation cover and production in various vegetation types in homogeneous units. Sampling points' locations were recorded with GPS. Sampling method was random-systematic in such a way that in each unit, a circle with 20 meters radius was considered. One and 9 sampling plots were placed on the centre and on the perimeter, respectively. The plot size was 1m * 1m. In each plot, vegetation percentage was estimated and the production was calculated using double sampling method. Then, DN values for each sampling unit (9 pixels for one unit) were elicited in respect to primary bands' images, vegetation indices and spectral ratios. Correlation and regression analyses between geo-information and satellite information (Digital numbers) were carried out. Results revealed that 7th and 5th Bands and IR1, MIRV2 and VNIR2 indices had a significant correlation with production and given parameter could be estimated through regression models. Likewise, RA, IR1 and TVI indices had a significant correlation with vegetation percentage and this parameter could be estimated through regression models.
Keywords
)