International Research Journal of Earth Sciences______________________________________ ISSN 2321–2527 Vol. 3(8), 14-31, August (2015) Int. Res. J. Earth Sci. International Science Congress Association 14 Grain Sorting on a Coastal dune of Orissa, Bay of Bengal and a comparison with the Desert Dune of Rajasthan, India Barendra Purkait 1* , Dipanjan Das Majumdar 2 and Ashis Kumar Paul 3 1 Department of Geology, University of Calcutta, 35 B. C. Road, Kolkata- 700 019 INDIA 2 Department of Remote Sensing and GIS, Vidyasagar University, Midnapore, West-Bengal – 721102, INDIA 3 Department of Geography and Environment Management, Vidyasagar University, Midnapore, West-Bengal – 721102, INDIA Available online at: www.isca.in, www.isca.me Received 13 th July 2015, revised 15 th August 2015, accepted 23 rd August 2015 Abstract Grain sorting patterns and textural parameters of a coastal dune of Konark, Orissa has been compared with the desert dune of Thar, Rajasthan, India. Sand samples were collected from the stoss, crest and the lee sides of the dunes. Grain-size distribution patterns were critically studied and compared with three hypothesized model distributions: log-normal, log- hyperbolic and the log-skew-Laplace to discriminate the coastal sands from the desert sands. In general, it is observed that the coastal dune sands are coarser and poorly sorted than the desert dune sands. Almost all the coastal and the desert dune sands, are positively skewed. Coastal dune sands are platy to leptokurtic distributed and desert sands are leptokurtic distributed. In terms of symmetry index, coastal dune sands exhibit a narrow range of symmetry index as compared to the desert dune sands. The log-skew-Laplace is the best-fit grain-size distribution model for coastal dune sands for stoss and crest samples whereas log-hyperbolic is the best-fit statistical model for lee samples; for desert sands, log-hyperbolic is the best-fit grain-size distribution model of the stoss and crest sands whereas log-normal is the best-fit statistical model for lee sands. The results of ANOVA and Tukey test signify the variance between coastal and desert dune sands. Finally, the canonical discriminant function yielded two distinctly separate spaces for coastal and desert dunes. Keywords: Coastal and desert dune sands, grain-size distribution, log-normal, log-hyperbolic, log-skew-Laplace, discriminant analysis. Introduction A considerable effort has been undertaken to determine the grain-size characteristics of aeolian sediments in order to elucidate the aeolian processes on coastal and desert dunes 1-3 and to distinguish different depositional environments based on grain-size characteristics 4 . Kar 4 in his earlier research made an attempt to discriminate the sands of desert dune from the coastal dune based on the granulometric characteristics. Kasper- Zubillaga and Dickinson 5 made an attempt to differentiate beach, dune and river environments in terms of their mineralogical components by using simple bar charts. Kasper-Zubillaga and Carranza Edwards 1 conducted grain-size discrimination from the slip face and crest of linear and sand sheet from the desert and coastal zone of the Altar Desert, Northwestern Mexico. They observed that the textural attributes of the dune fields are affected by their proximity to the sediment sources, tidal transport, long shore driftand wind direction. Coastal dune sands are coarser than the desert dune sands. They suggested that the linear discriminant analysis and ANOVA test are useful tools to discriminate between desert and coastal dune sands. Kasper- Zubillaga et al., 2 carried out sedimentological, compositional and geochemical determinations on desert and coastal dune sands of the Altar Desert. According to Livingstone et al. 7 , grain-size variations between dunes are related to the height of the dune. Although there is progressive fining of sand from dune base to dune crest 8-9 , this relationship cannot be adopted as a general rule, since grain-size distributions in the dune may be controlled by different modes of transport and wind pattern. The experimental study of Lancaster et al. 10 on a small reversing dune shows that particle size and sorting characteristics of surface sand samples at any point in time and space are highly dependent on wind competency and resultant sand transport rates. It is observed from the aeolian research that the coarser fractions of a grain- size population usually move near the bed because of their least susceptibility to lifting 10-15 . Therefore, it is assumed that grain- size should decrease with elevation from the bed surface. Contradictory evidence from field experiments 16-21 and wind tunnel experiments 21-22 suggest that this simple size decay concept may not be typical. Farrell et al., 23 studied the vertical distribution of grain-size for wind-blown sand and observed that a reverse in grain-size trends occurs at an inflection zone above the bed. Below this inflection zone, mean grain-size decreases steeply with elevation in the near bed region. There is a coarsening of grain-size with elevation occurs just above the inflection zone. Zhu et al. 24 studied the grain-size characteristics of aeolian dune and Gobi sands at different sections and observed that sorting improves with decreasing grain-size. A clear relationship occurs between skewness and sorting being