Out of Africa: Radarclinometry of the Sand Seas of Namibia and Titan

Introduction: Recent observations by the Cassini spacecraft show widespread regions of longitudinal dunes on Saturn’s moon Titan [1]. The exact composition of the dunes is not known, but its radar and infrared return suggest an organic component [2]. Given the broad extent of the dunes (they cover 40% of Titan’s equatorial areas), they constitute an important portion of Titan’s overall organic inventory. To determine the exact volume of dune sands on Titan, it is necessary to estimate their heights over large areas. The height and spacing of dunes also provides information about the depositional environment in which the dunes were formed [3]. Given the paucity of altimetry data available on Titan, the best constraint on the height of the dunes comes from backscatter variations visible in high resolution SAR (synthetic radar aperture) images taken by the Cassini spacecraft. Radar images of planetary surfaces provide a wealth of information about the surface being imaged, since radar backscatter depends on three surface properties – topography, roughness, and composition. Because of the dependence of radar backscatter on local incidence angle, an individual SAR image can reveal topography by radarclinometry, or “shape-fromshading” [4]. This technique is limited in the sense that radar backscatter does not depend exclusively on topography, but also on roughness and composition. Sand seas (with sand-covered interdune areas) make excellent targets for radarclinometry, since they are essentially uniform in composition, and somewhat uniform in roughness. This makes topography the only changing variable, so differences in radar backscatter can be interpreted as differences in local incidence angle with some confidence. Radarclinometry has been performed on the longitudinal dunes observed on Titan, yielding slopes of 6 to 10 degrees, and heights of 100-150 m [1]. Nonetheless, given the assumptions and approximations inherent in radarclinometry, it is instructive to validate the technique using independent sources of topography, where such sources are available. This gives us an understanding of how robust the technique is when applying it to planets where independent sources of topography are limited or unavailable, such as Titan. The spacing and height of the dunes on Titan appear to be similar to the dunes observed in the Namib desert of western Africa. This region has been imaged by both Spaceborne Imaging Radar (SIR-C) and the Shuttle Radar Topography Mission (SRTM), providing independent SAR and topography data. In this work, we perform radarclinometric analysis of the SIR-C images of the Namib sand sea, and compare the resultant height profiles to topography data from the same region. We then use this technique to estimate the height of dunes on Titan. In this way, we judge the reliability of radarclinometry in determining the topography of sand dunes. Method: To carry out any form of radarclinometry, a functional relationship between surface slopes and image signals is required. These scattering models describe the radar backscatter function, σ0, as a function of local incidence angle, i. For the Namib desert, we used the scattering model given for smooth, gently sloping, dry desert sand in Ruck et al. [5], scaled to the observed changes in backscatter over the SIR-C radar image. For Titan, we used a scattering model developed by Wye et al. [6], fit to scatterometry data from dune areas on Titan (Equation 1).