Elsevier

Planetary and Space Science

Volume 218, 1 September 2022, 105505
Planetary and Space Science

Pre-landslide topographic reconstruction in Baetis Chaos, mars using a CaSSIS Digital Elevation Model

https://doi.org/10.1016/j.pss.2022.105505Get rights and content

Highlights

  • Digital Elevation Model are needed to estimates landslide's material volumes and distribution in order to fully understand their mobility.

  • CaSSIS (Colour and Stereo Surface Imaging System) DEM have not yet been used to conduct such landslide analysis.

  • Three reconstruction on a CaSSIS DEM have been used to estimate the volume and the material distribution of a landslide in Baetis Chaos on Mars.

  • CaSSIS data can be used to conduct analysis and provide additional coverage to study medium-scale landslides or other landforms (scale 5–15 km).

Abstract

Planview detailed morphological analysis of martian landslides is usually performed using orbital imagery such as from the ConTeXt camera (CTX) at 6 ​m/pix, the Colour and Stereo Surface Imaging System (CaSSIS) at 4.5 ​m/pix or the High-Resolution Imaging Science Experiment (HiRISE) at 0.25–0.5 ​m/pix. However, topographic information is key to fully understand a landslide's formation mechanism and its mobility, by estimating the material volumes mobilised and the spatial distribution of erosion and deposition. Digital Elevation Models (DEM) are required to carry out these analyses; nevertheless, there is currently a gap in landslide-volume studies between those using Mars Orbiter Laser Altimeter (MOLA) dataset at ∼450 ​m/pix or HRSC at 50–200 m/pix and those using HiRISE data at 1–2 m/pix, which is only partially filled by CTX elevation data at ∼20 ​m/pix. The CaSSIS camera on board the ESA/Trace Gas Obiter (TGO) can be used to produce DEMs, but so far, such data have not yet been used to conduct a landslide volume analysis. Here, we use three reconstruction methods (semi-automatic, morphology-based and tilted) on a CaSSIS DEM to estimate the initial topography and hence the volume and the distribution of erosion and deposition of a 6 ​km long landslide in Baetis Chaos. Despite the complex topography of the surrounding area due to the presence of an ejecta deposit beneath the landslide, we were able to estimate the landslide's volume and mass distribution. Using a tilted plane as part of estimating the initial topography produced the best results. We evaluated the success by considering the quantifiable balance between erosion and deposition (given the uncertainties) and more subjectively by considering whether the volume distribution matched with what was expected based on the morphology in images alone. Therefore, we recommend the use of this method for individual landslide studies in complex topography where detailed knowledge of the deposit-thickness distribution is required. The semi-automatic reconstruction method produces satisfactory volume estimates and would be better suited to studies where hundreds of landslides are present. We found that CaSSIS data can be used to successfully conduct such analyses, providing additional DEM coverage to study martian medium-scale landslides or other landforms of similar scale (5–15 ​km) with the notable benefit that it provides single-pass stereo image acquisition.

Introduction

Martian landslides are common features that can have morphologies that resemble Earth debris slides (e.g., Crosta et al., 2018a), mudflows (Guimpier et al., 2021), or giant rock avalanches (McEwen, 1989; Quantin et al., 2004; Magnarini et al., 2019). They can mobilise large quantities of material up to 1012 ​m3 and spread over areas of up to 109 ​m2 (Quantin et al., 2004; Crosta et al., 2018b). The typical morphology of a landslide is composed of three distinct zones (Highland and Bobrowsky, 2008): the erosion zone at the top of the landslide, the transport zone through which the material transits and the deposition zone where the material accumulates.

In order to better understand the dynamics of landslide formation, it is important to quantify the volume of mobilised material and understand both the distribution and thickness of the deposit. The dynamics of landslides can be influenced by parameters such as the presence of water or ice within the sliding material (Cruden and Varnes, 1996), so a better understanding of their dynamics permits a better understanding of the conditions in which these landslides are formed. When compared with the vertical and horizontal distances travelled, the volume of the sliding mass can be used as a measure of the event mobility and used to compare it to other landslide events (Lucas and Mangeney, 2007). Moreover, the mass distribution of the deposit can be compared with the outputs of 3D flow simulations of landslides (Crosta et al., 2018b; Magnarini et al., 2019; Guimpier et al., 2021; Pajola et al., 2022), which can be used to better understand the physical mechanisms involved in the mass movement. In the case of recent terrestrial landslides, these volume calculations can be obtained by differencing Digital Elevation Models (DEMs) from pre- and post-landslide (Tsutsui et al., 2007).

For places where no pre-landslide topographic data are available, a topographic reconstruction is needed to obtain a pre-landslide DEM. This kind of reconstruction is carried out on Earth when no pre-event topography exists, such as on volcanic events using a DEM with a spatial sampling of 5 ​m/pix (Rodriguez-Gonzalez et al., 2010) and on landslides using DEM with a 0.1–0.2 ​m vertical resolution and 2 ​m spatial sampling (Conoscenti et al., 2015).

On Mars, topographic reconstruction has been carried out for landslides using the Mars Orbiter Laser Altimeter (MOLA) data, which has a 1.5–2 ​m vertical resolution (Quantin et al., 2004; Lucas and Mangeney, 2007; Lucas et al., 2011). However, because of the spatial sampling of 463 ​m/pix (Smith et al., 2001a) only landslides bigger than 10 ​km have been studied. Also, data from the High Resolution Stereo Camera (HRSC), which has a maximum of 10 ​m vertical resolution (Neukum and Jaumann,), has been used for landslide reconstruction (Crosta et al., 2018b). These authors used HRSC DEMs with spatial sampling of 15 ​m/pix allowing the study of landslide <10 ​km, yet the vertical resolution leads to significant uncertainty in the volume calculation for smaller landslides. Finally, elevation data from the High Resolution Imaging Science Experiment (HiRISE) with 1–2 m/pix spatial sampling and <1 ​m vertical resolution has been used to perform topographic reconstruction for kilometre-scale gullies (Conway and Balme, 2014; de Haas et al., 2015) and landslides (Guimpier et al., 2021; Pajola et al., 2022). However, features larger than 6 ​km cannot generally be accommodated in a single DEM due to relatively narrow HiRISE imaging swath (McEwen et al., 2007).

Out of the 3000 landslides identified on Mars (Crosta et al., 2013, 2018b), around 1/3 fall between these two spatial scales. The Colour and Stereo Surface Imaging System (CaSSIS) instrument on board the Trace Gas Orbiter (TGO) can be used to produce DEMs with a 4–5 ​m vertical resolution and 15 ​m/pix spatial sampling and have a spatial coverage of 9.4 by 47 ​km (Thomas et al., 2017). Such DEMs have never been used to perform topographic reconstruction of landslides, and together with CTX could be used to fill the scale-gap between HiRISE and HRSC-MOLA. The benefit of the CaSSIS camera over CTX is that it uses a 180° rotation mechanism to capture stereo images of a given site in a single pass (Thomas et al., 2017).

Our first aim is to test whether topographic reconstruction of a terrain is possible given the spatial sampling and quality of CaSSIS DEM data and to determine whether it is possible to derive useful information for the analysis of landslide dynamics (estimation of the volume and thickness of the deposit). For this purpose, we use a 6 ​km long landslide located in Baetis Chaos region where no stereo HiRISE nor exploitable stereo ConTeXt (CTX at 6 ​m/pix) imagery datasets are available. We then test three reconstruction methods to estimate the mass distribution and volume of the landslide with the aim to understand which method produces the best result.

In this manuscript, we first introduce previous methodologies used to reconstruct topography in landslide studies on Mars. Then we present the geographic context and the morphology of the landslide used in this study. Afterwards, we describe the methodology used to create the CaSSIS DEM followed by a description of the three reconstruction methods and associated uncertainties. Finally, we present the results of each reconstruction method and provide an assessment on the use of CaSSIS stereo data for pre-landslide reconstruction.

Section snippets

Brief summary of previously used topographic reconstruction methods

In order to better understand the dynamics of martian landslides, several studies have already been carried out that estimate landslide volumes or deposit distributions. In the following section, we will briefly review the different methods used to carry out these estimates. We first describe topographic reconstruction methods applied to large numbers of landslides in population-scale studies (Section 2.1) and then reconstruction methods applied at local scales for selected individual case

Overview of the study area

The study area is located in Baetis Chaos (0°14′ S, 60°34’ W) (Fig. 1a and b). Martian chaos terrains were first described by Sharp (1973) and they are commonly characterised by a rough floor topography where irregular jumble of blocks with different sizes occur (ranging from hundreds of metres to several kilometres in size). These blocks generally have a fairly flat top-surface, flanked by steep slopes and can reach several kilometres in diameter (e.g., Pajola et al., 2016). The exact

Methodology

In order to perform the pre-landslide topographic reconstruction, we need a post-landslide DEM of the area within which the landslide is clearly recognisable (which is not the case for MOLA or HRSC). CaSSIS provides a more detailed representation of this landslide and allows, with its stereo capability, a more complete analysis than could be performed with the other existing data (Fig. 3b). In section 4.1., we detail the DEM construction from the CaSSIS stereo images using 3DPD (three

Results

The three reconstruction methods provided different results, highlighting that the reconstruction is sensitive to the assumptions made during the reconstruction process and to the complex topography in the deposit zone. The ejecta deposits on the Chaos floor mean that the topographic reconstruction is more uncertain if compared to the landslides that superpose the flat floor of Valles Marineris, for example (e.g., Quantin et al., 2004).

The hillshaded relief view of the original and

Assessment of three reconstruction methods

The semi-automatic reconstruction method produces a similar estimate of the overall volume of the landslide to the other two methods despite producing an unrealistic volume distribution. Therefore, this quick method would be appropriate if only an estimate of the volume is required. However, if we remove the second unrealistic erosion zone, which is located in the transport and deposition zones, the expansion coefficient reaches about 344%, revealing that the amount of erosion in the erosion

Conclusions

Our successful reconstruction of pre-landslide topography with a CaSSIS DEM means that these data can be added to the already available coverage of high-resolution stereo-topographic data with CTX and/or HiRISE. By using three different reconstruction methods from the simplest to the more complex we have shown that CaSSIS data can be used to perform successful and detailed topographic reconstructions. We evaluated this by considering the quantifiable balance between erosion and deposition and

Author statement

A. Guimpier: Conceptualization, Methodology, Writing - Original Draft; S. J. Conway: Conceptualization, Methodology, Writing - Review & Editing, Supervision; M. Pajola: Resources, Writing - Review & Editing; A. Lucchetti: Resources, Review & Editing; E. Simioni: Resources, C. Re: Resources, A. Noblet: Resources; N. Mangold: Review & Editing, Supervision; N. Thomas: Resources; G. Cremonese: Resources.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgments

SJC, NM and AG are grateful for the support of the Programme National de Planétologie and the French Space Agency (CNES). The authors thank the spacecraft and instrument engineering teams for the successful completion and operation of CaSSIS. CaSSIS is a project of the University of Bern funded through the Swiss Space Office via ESA's PRODEX programme. The instrument hardware development was also supported by the Italian Space Agency (ASI) (ASI-INAF agreement no. I/018/12/0), INAF/Astronomical

References (42)

  • N. Thomas et al.

    Absolute calibration of the colour and stereo surface imaging system (CaSSIS)

    Planet. Space Sci.

    (2022)
  • D. Watson

    The natural neighbor series manuals and source codes

    Comput. Geosci.

    (1999)
  • L.P. Chew

    Constrained delaunay triangulations

    Algorithmica

    (1989)
  • N.M. Coleman et al.

    Surface morphology and origin of outflow channels in the Valles Marineris region

  • S.J. Conway et al.

    Decameter thick remnant glacial ice deposits on Mars

    Geophys. Res. Lett.

    (2014)
  • G.B. Crosta et al.

    Global scale Analysis of martian landslide mobility and paleoenvironmental clues

    J. Geophys. Res.: Planets

    (2018)
  • G.B. Crosta et al.

    A new inventory of martian landslides

  • G.B. Crosta et al.

    Introducing a new inventory of large martian landslides

    Earth Space Sci.

    (2018)
  • D. Cruden et al.

    Landslide types and processes. Special report - national Research council

    Transport. Res. Board.

    (1996)
  • T. de Haas et al.

    Recent (Late Amazonian) enhanced backweathering rates on Mars: paracratering evidence from gully alcoves

    J. Geophys. Res.: Planets

    (2015)
  • C. Gross et al.

    Further evidence for multiple flooding events at Juventae Chasma and Maja Valles

    MARS

    (2009)
  • View full text