Using vegetation indices of ASTER-L1B Imagery to estimate single trees crown cover in arid rangelands, Case study; Tag - Ahmadshahi, Southern of Khorassan

“Crown cover area” is one of the multiple-use ecological characteristics in natural resources studies. Current research has been done to evaluate capability of different vegetation indices, derived from ASTER-L1B image processing, to estimate crown cover area of single trees in very sparse forest stands in arid rangelands. The main hypothesis of the research is based on using spatial, spectral and radiometric resolution of ASTER imageries to estimate crown cover area of single trees, more accurately and precisely. Study area, a natural site of wild pistachio (Pistachio atlantica L.), is located in Tag-Ahmadshahi, southern Khorassan province. The geometric correction of the images has been accomplished, using non-parametric method of TPS and Digital Elevation Model (DEM); derived from topographic map of 1:50000, with a RMSE of less than 1 pixel. The filed data belong to 50single trees, sampled from a complete inventory of trees in the study area. Vegetation indices and soil line equation have been calculated and the values have been extracted from the indices in the place of the corresponding sampled trees. Then, regression equations have been established between vegetation indices and the crown cover area. The results indicate significant correlations between them. Among the indices, NDVI and MSAVI presented stronger statistical correlation than the other indices with crown cover area, having the correlation coefficient of 0.61 and the standard error of 5.58.