Towards the automated geomorphometric extraction of talus slopes in Martian landscapes

Terrestrial talus slopes are a common feature of mountainous environments. Their geomorphic form is determined by their being constituted of scree, or similar loose and often poorly sorted material. Martian talus slopes are governed by the different nature of the Martian environment, namely: weaker gravity, the wide availability of loose material, the lack of fluvial erosion and the typicality of large escarpments; all these factors make talus slopes a more striking areomorphic feature on Mars than on Earth. This paper concerns the development of a numerical geomorphometric analysis, parameterization and detection of the talus slopes method. We design inventive variables, a multidirectional visibility index (MVI) and a relief above (RA) and propose two techniques of talus slope extraction: ISOcluster and progressive Boolean overlay. Our Martian digital terrain model (DTM) was derived from the ESA Mars Express HRSC imagery, with a resolution of 50 m. The method was tested in the study areas of Nanedi Valles and West Candor Chasma. The major challenge concerned the quality of the DTM. The selection of robust variables was therefore crucial. Our final model is to a certain degree DTM-error tolerant. The results show that the method is selective concerning those slopes that can be considered to constitute a talus slopes area, according to the visual interpretation of HRSC images. Based on an analysis of the DTM, it is possible to infer various geological properties and geophysical processes of the Martian and terrestrial environments; this has a range of applications, such as natural hazard risk management.