Stratigraphy, morphology, microtexture of the ILDs in Capri Chasma: insights into their formation.

Capri Chasma is a canyon located at the exit of Valles Marineris. Its flanks are piled with lava flows. It also contains in places, bright Interior Layered Deposit (ILD) revealing a complex geological history. Some slopes are also the site of a phenomenon known as RSL (recurrent slope lineae). In the northern part of the Chasma, within the thickest portion of the Interior Layered Deposit (ILD), rock units are exposed along some walls, that are part of a ∼30-km circular depression. A well-preserved stratigraphy in the northwest-facing slope reveals several sulfate-rich units, including kieserite and poly-hydrated sulfates, some of which are associated with red and gray hematite [1,2]. A presumed impact has disrupted the southern stratigraphy, redistributing these mineral layers and contributing to a jumbled geological structure. The presence of sulfates and hematite suggests initial ILD deposition by evaporation of magnesium and sulfur-rich brines, followed by groundwater upwelling events that penetrated the ILD, allowed gray hematite precipitation, and induced diagenetic alteration. Alternative explanations involve atmospheric deposition of sulfur-rich aerosols and dust particles within the ice deposits followed by weathering and formation of the sulfates during Martian climatic cycles. As a result, the origin of the Capri Chasma ILDs is still rather vague, and more information needs to be gathered on the stratigraphy, morphology, and color of the deposits, as well as on the microtexture of the materials that make them up. In order to better understand the nature of the deposits and the conditions of their formation, our investigation is based on three types of data: (i) CTX and HiRISE digital terrain models refined by photoclinometry (ii) a multi-angle sequence of hyperspectral images acquired by the CRISM spectro-imager on which we invert Hapke's photometric model after atmospheric correction (iii) a series of images from the CaSSIS multispectral sensor on which photometric and atmospheric correction is performed to extract the intrinsic surface reflectance in four spectral bands, resulting in true color images.

First, we analyze the spectrophotometric data generated from the fusion of the sequence FRTB385 of 11 hyperspectral CRISM images, estimating both atmospheric optical thickness and surface reflectance at a 200 m resolution using the MARS-Reco tool [3]. An unsupervised k-means classification clusters the reflectance data into five photometric classes(fig.1), and Hapke photometric model parameters are computed using the GLLiM algorithm [4] to interpret surface microtexture properties. These parameter maps are overlaid on geological images, and mean volume phase function parameters (b, c) are compared with laboratory data to infer material textures (fig.2). The analysis shows that bright sulfate deposits (cyan class) have translucent grains throughout the wavelength range with high surface roughness, consistent with crystals formed by evaporation of brines. Hematite-bearing materials (blue class) are strongly backscattering at visible wavelengths, where internal absorption is high, and become significantly forward scattering in the shortwave infrared, suggesting round, clear grains, similar to textures observed by the Opportunity rover at Meridiani. This likely indicates very similar conditions of formation.

Fig.1 False color CRISM image (left) and classification map (right) superimposed on a context CTX image of a semi-circular depression in the Eos Chasma, Mars. The classification of the terrains is based on a kmeans clustering of their CRISM photometric curves at a wavelength of 755m.

Fig.2 Qualitative interpretation of the microtexture of the different terrain classes based on a comparison of the spectral behavior of the phase function with laboratory measurements in the Hapke (b, c) parameter space.

Second we use a CTX DTM and its associated CTX orthoimage D07_029766_1668_XN_13S047W, both products generated by a photogrammetric processing of an image pair. The goal is to remove artifacts from the initial CTX DTM at 18m.pixel-1 and to densify it to 6m.pixel-1 using the ortho-image, which exploits intensity variations associated with slopes. To do this, we use the HDEM photoclinometric method [5] based on the CRISM reflectance model previously established for the scene. Prior to this, the initial DTM is filtered to an effective resolution of 288 m.pixel-1 after being decomposed on a spatial wavelet basis using the 2D Isotropic Undecimated Wavelet Transform algorithm, followed by the elimination of high frequencies. The result is highly satisfactory, allowing a much finer, less error-prone topographic characterization of the terrain (fig.3). A similar process was applied to a HiRISE DTM and its associated PSP_008958_1665 orthoimage, resulting in a resolution of 25 cm.pixel-1. In this case, morphometric terrain analysis is possible with a relative vertical resolution estimated at 15-20 cm, giving access to the details of aeolian figures, for example.

Fig 3. Topographic profiles extracted for a mound in the Capri Chasma circular depression.

Thanks to the use of novel CTX, HiRISE and CaSSIS image processing methods, we are carrying out a detailed morphological, stratigraphic and textural characterization of the sulfate- and hematite-rich units over an interesting part of the Capri Chasma. Valuable information has already been obtained that complements the spectroscopic studies to understand the origin and evolution of these formations, as well as the surface phenomena that are currently shaping these landscapes.

[1] Roach, L.H., Mustard, J.F., Lane, M.D., Bishop, J.L., Murchie, S.L., 2010. Icarus 207, 659–674.

[2] Weitz, C.M., Noe Dobrea, E.Z., Lane, M.D., Knudson, A.T., 2012. Journal of Geophysical Research (Planets) 117, E00J09.

[3] Ceamanos, X., Douté, S., Fernando, J., Schmidt, F., Pinet, P., Lyapustin, A., 2013. Journal of Geophysical Research (Planets) 118, 514–533.

[4] Kugler, B., Forbes, F., Douté, S., 2022. Statistics and Computing 32, 31.

[5] Douté, S., Jiang, C., 2019. IEEE Transactions on Geoscience and Remote Sensing 1–14.

Acknowledgment : S.D. is grateful to the Centre National d’Etudes Spatiales (CNES) for supporting his CaSSIS and HiRISE related work through the Program “Exobiologie, protection planétaire et exoplanètes”.