Remote Sensing for Identification and Characterization of Zibar Sand Dunes in Sandy Alluvial Plains within Thar Desert, India

Abstract

Zibars, a kind of small dunes with coarser particles in the desert, are often difficult to identify in field. This study was carried out in the vicinity of Jodhpur to test the use of remote sensing techniques to identify the zibars and study their major characteristics. Following satellite data were used in analysis: multi-spectral data in the form of FCCs of AWiFS and LISS-IV, hyperspectral data from Hyperion, and microwave data from RADARSAT-2. CartoDEM data were used to find out the elevation differences in the study area, while a field spectro-radiometer was used to find out the reflectance properties of different land surfaces. It was found that neither all kinds of image combinations, nor all conventional digital analysis techniques provide useful results. Spectral angle mapper classification of Hyperion data were found better than a grain size index mapping and some conventional interpretation techniques. Since there is very little control of soil moisture variation in the zibar pattern, microwave data were also of little use. The study concluded that zibar pattern gets manifested on the satellite images due to grain size variation. Keywords: Zibar, spectral reflectance, AWiFS, LISS-IV, Hyperion, RADARSAT-2, Spectral Angle Mapper, Grain Size Index.