Salahudin Zahedi, Kaka Shahedi
Watershed Science of Sanru and Natural Resource Researcher of Agriculture and Natural Resources Research Center of Kurdistan, Kurdistan, Iran
Watershed Sience Department, Faculty of Natural Resources, SANRU, Sari, Iran
Key words: production estimation, semi-arid shrub lands, multivariate regression, TM sensor.
Rangelands vegetation cover is the main resource production of protein in Iran. Inappropriate usage and misknowing of species combination, is the agent of decreasing of valuable species in rangelands. Annual production definition based on growth form or possibly could base on species is the major factor of accurate rangelands management specially grazing programing and natural or intentional fire prevention. This study applied double date TM imagery to estimate production of shrubberies forms of rangelands of Qeshlaq dam watershed. The images were processed by ERDAS IMAGINE software. The rangeland yield clipping and weighing system applied to measure green herbage biomass from ground truth sites by means of 300 medium plots (5m2). Ground truth sites were selected to represent five rangeland types and four sites were sampled by systematic random method in each type to calibrate the relationship between satellite-derived Wavebands, vegetation indices and green yields. These yield data were compared with yields estimated by 6 main wavebands also 4 synthetic bands of 2 scenes of TM data in corresponding time and go through linear multivariate regression processing to make the model. Remote sensing yield estimation model were also analyzed for their precision and checked by actual 10 percent measured yields on four ground truth sites. Results showed that relationship between shrubberies growths is meaningful with, ND53 and TM5/TM3 bands. Resulted model have higher accuracy in estimating shrub forms growth production in order to permanent management of rangelands in comparison with traditional models.
Get the original articles in Source: Volume 5, Number 1, July 2014 – JBES
Published By: International Network for Natural Sciences