مقايسه روش هاي پيكسلي و تحت پيكسلي در پهنه بندي پوشش گياهي مراتع﴿مطالعه موردي: شهرستان سميرم، استان اصفهان﴾

چكيده انگليسي :

Comparison of Pixel and Subpixel Methods for Mapping Rangeland Vegetation Cover Case study Semirom Region Isfahan Province Mahdiyh Farazmand Farazmand 202@yahoo com Data of Submission February 15 2012 Department of Natural Resources Isfahan University of Technology Isfahan 84156 83111 IranDegree M Sc Language Farsi Supervisor Dr Reza Jafari Email address Reza Jafari@cc iut ac irAbstractAmong the various desertification processes and vegetation degradation are the main processes in rangelandsand it often starts with the reduction of vegetation cover This can result from single or combined effects ofovergrazing rainfall deficits The effects of these are the appearance of barren soils and an increasedsusceptibility to wind and water erosion Vegetation cover is one of the most important components of theearth s surface It strongly influences evapotranspiration infiltration runoff and soil erosion Vegetationcover is also the principal factor limiting stocking rates in managed grazing lands It has been widelyrecognized as one of the best indicators for determining land condition Therefore land condition can beassessed and monitored according to vegetation cover and its variations in time and space This component isoften used as the key indicator in the remote sensing of land condition Assessment and monitoring is animportant and essential step for rangeland management Nowadays satellite images are used in differentscales to study vegetation cover Usually data vegetation changes are evaluated with studying vegetationspectral characteristic This study aimed to compare and evaluate pixel vegetation indices VIs and sub pixel spectral mixture analysis SMA methods in mapping vegetation cover in Semirom region Isfahanprovince using Landsat TM data For this purpose Landsat TM image of the region in 2009 were preparedand processed First the image was geometrically with a RMSe 0 5 pixel and topographically corrected Thepercentage of canopy cover was determined using step point technique in radial direction 6000 point persite The collected field data was correlated with different groups of vegetation indices including slop based NDVI distance based PD54 orthogonal transformations GVI and SBI and plant water sensitive STVI 1 and also SMA To evaluate the performance of VIs and SMA techniques vegetation cover data wasclassified in three percentage groups including 0 25 25 50 and 50 In 0 25 group the SMA and PD54 indexhad the highest relationships with field vegetation cover data 0 52 0 45 and p 0 05 The GVI and STVIcorelated the best in group 25 50 vegetation cover percentage The NDVI index of the distance basedvegetation indices showed the strongest relationship with cover data 0 60 and P 0 01 where thepercentage of vegetation cover was higher than 50 In general results showed that relationships betweenSMA and field cover data decreased with increasing vegetation cover and among the pixel and sub pixeltechniques the PD54 and SMA appear to be the most suitable methods in arid environments with sparsevegetation cover The results also suggested that NDVI is a useful index for general cover assessment andmonitoring regardless of soil and vegetation variation especially in dense vegetation areas Because of thebroad extent of rangelands land condition monitoring and assessment using ground based methods is limitedin relation to the information they can provide Results of this research showed remote sensing techniques including VIs and SMA could be used as complementary approaches Keywords SMA NDVI PD54 GVI SBI STVI 1 Isfahan

فرازمند، مهديه

مقايسه روش هاي پيكسلي و تحت پيكسلي در پهنه بندي پوشش گياهي مراتع﴿مطالعه موردي: شهرستان سميرم، استان اصفهان﴾

SMA , NDVI , PD54 , GVI , STVI-1