Dune Morphometry in the Age of Digital Elevation Models

Abstract:

Dune patterns can be characterized in many different ways. Relationships between dune height, width and spacing, and the spatial variation in these parameters have been widely employed to provide quantitative information that can be used to describe dune patterns and make comparisons between dunes in widely separated areas, as well as to identify different generations of dunes. Digital elevation models (e.g. ASTER GDEM) provide a rich resource of data for analyses of dune patterns at landscape scales in several ways, including: (1) more extensive analyses using traditional measures, such as relationships between dune height and spacing, and the spatial variation in these parameters; and (2) estimation of sediment thickness on a regional scale. Analyses of data for Arabian and Namibian sand seas and dune fields show that dune height and spacing relationships are much more variable than previously reported and call into question existing models. Regional patterns of sediment thickness reveal areas of erosion, bypass, and accumulation that can be related to transport pathways and wind regimes. The widespread occurrence of complex dune patterns as well as the magnitude of the newly available data sets however requires more sophisticated analyses than simple extraction of dune morphometric parameters using GIS approaches. Geostatistical analyses using spatial autocorrelation, Fourier, and Wavelet methods have been employed in analyses of sub-aqueous bedforms and show promise for dune systems. Automated or semi-automated identification of dune length, width, spacing, and trends using advanced image analysis techniques such as linear segment detection is a potentially transformative approach. The strengths and weaknesses of these methods to provide pertinent geomorphic information are currently being evaluated, but they have the potential to provide new insights into the nature of dune patterns.