GEOMORPHOLOGICAL AND MINEROLOGICAL CHARACTERIZATION OF SAND DUNES USING MULTISPECTRAL LANDSAT 7 ETM+, AND SPOT, EASTERN PART OF ABU DHABI, UNITED ARAB EMIRATES

Various remote sensing data sets are useful to extract information of various spatial and spectral scales. In this study, Landsat 7 ETM+, and SPOT datasets were used to study the geomorphology, and mineralogy of different types of sand dune types in the eastern part of Abu Dhabi in United Arab Emirates. The extraction and mapping of detected components were carried out using principal component analyses, spectral analyses and matching methods. The results of the present study showed that the sand dunes are classified into three main classes based on composition. The first class is low dunes composed of a mixture of quartz and carbonate, the second class is dunes covered with desert grass and shrubs and is stable, and the third class is dunes without vegetation, and believed to be mobile. Morphologically, the dunes in the investigated area belong to three types. The first is linear and is trending NE-SW. This type shows transitional change in mineralogy from carbonate rich components to iron oxide rich components. This change was observed on Landsat images as spectral variations and was confirmed from ground truth data. This type of dunes mostly reflects eastward wind direction. The second type is barachonid dunes consisting mainly of chains of crescent and indicate eastward wind direction. The third type is star dunes and appears in images as radially symmetrical to mound shape that characterize multidirectional wind systems. The existences of star dunes indicate a low sand supply environment. Eastward of the study area, the intensity of dunes at the rise, and appeared as clusters of low level white dunes. In some locations, as depicted from ground truth, linear dunes exhibited grain size and sorting patterns in which sand becomes coarser, better sorted and finally skewed from interdune to crest. This can be due to differential movement of sand in creep and saltation populations, where the fines will tend to be concentrated in the interdunes because they can penetrate between coarser grains, where they are prevented from migration. The results of this study show that multispectral data can be used to differentiate between different dune types and associate mineralogy and to determine dynamic processes shaping the formation of the dunes such as prevailing wind directions.