Multi-resolution digital terrain models and their potential for Mars landing site assessments

Abstract

One of the key issues for planetary, especially Mars surface exploration, is how to make an objective assessment of criteria for landing sites selection for future rovers, sample return missions and landers considering topographic variables such as slope, altitude and roughness as well as their physical scattering properties. For these purposes, stereo vision analysis is here proposed as the best possible solution to provide reliable topographic data. Recently, a number of successful orbital missions to Mars have taken place including the Mars Express mission with the High Resolution Stereo Camera (HRSC) as well as the Mars Reconnaissance Orbiter spacecraft equipped with two major cameras—Context Camera (CTX) and High Resolution Imaging Science Experiment (HiRISE). A stereo processing chain has been developed to generate high quality Digital Terrain Models (DTMs)—up to a maximum grid spacing of 0.7m with HiRISE, 10m with CTX and 25m with HRSC along with terrain-corrected ortho rectified images. Applying this system, topographic datasets were produced over potential landing sites, which had been previously proposed based mainly on their purported geological significance. High (<10m) and ultra-high (<4m) resolution DTMs from stereo imagery were employed to confirm mainly topographical hazard free landing sites from an engineering standpoint as well as to assess the geology of the target areas. In particular, the Minimum Noise Fraction (MNF) approach has been applied to assess landing risks quantitatively based on the surface roughness of the resultant topographic products. In future, it is expected that topographic products can be integrated with other data sources such as hyperspectral imagery, radar backscattering and laser beam broadening in order to provide the physical properties for the landing site selection as well.