Reconstruction of overlapped subaerial and subaqueous deposits in Coprates Catena, Mars

1.INTRODUCTION

Here we simulate the complete process of channel carving and the development of a valley-fan type system in the Coprates Catena region on Mars (Figure 1). Previous works [1,2,3] hypothetize that this stepped fan-like structure was formed in different stages, with an intermediate time-lapse. We reproduce the initial and final phases to analyze their different conditions and determine whether they represent alluvial fans (subaerial) or deltas (subaqueous). By iteratively combining topographic reconstructions and mechanical modeling, we can constrain the formation conditions of these surface processes that took place in the past on Mars’s surface, including if they are compatible with a permanent sheet of water covering the canyon (Figure 2).

Figure 1. Top left, location of Coprates Catena on the Mars globe (14º 59' S; 60º15' W), south of Valles Marineris. Below is a delimitation of the two episodes considered on a mosaic of CTX images, the subaqueous (oldest) in blue and the subaerial (youngest) in orange.

 

2.METHODOLOGY

We begin with a "backward reconstruction,” modifying the topography to approach the area shape before the triggering event [4,5]. We subtract the volume of fan-shaped material, backfilling the channel with the equivalent mass (Figure 2). Through interpretive geology, we identify which points of the observed topography were not modified after the event and reconstruct the altered areas using a thin plate spline interpolation algorithm (Figure 3). The mobilized material, an initial condition necessary to feed the numerical model, is then obtained with the difference between both reconstructed topographies, which is then interpolated in an unstructured computational mesh.

 

Figure 2. Topographic profiles (see Figure 1) for the thalweg (blue) and slope edge for both the older (red) and younger (orange) episodes. The vertical black line marks the boundary between deposition and transport. The dashed blue line shows the paleolake level.

 

Figure 3. Above, points used for topographic reconstruction of the channel (Orange: youngest episode; Black: oldest episode; Blue: thalweg; Yellow: reference points outside the channel; White: points belonging to altered zones to be calculated). Below is a reconstruction of the channel and canyon using thin plate spline interpolation in MATLAB.

 

For the "forward reconstruction” of the event, a systematic numerical study is performed using a two-layer model based on the depth-integrated Navier-Stokes equations [6] with different constitutive models combined with bottom friction laws [7,8,9]. The resulting model allows the dynamic simulation of fluid displacement under subaerial and subaqueous conditions, providing the flow time and runoff distance, as well as the final volume and shape of the deposits.

Considering that this fan structure formed in several phases, with a time span in between [1,2,3], we have divided the event into two distinct episodes: The older episode (Figure 2, in blue), which is assumed to be a subaqueous deposition due to the presence of a deep lake within the canyon [2]; and the more recent episode (Figure 2, in orange), which is considered to be an alluvial fan that was deposited on top of the previous subaqueous delta [3].

 

3.RESULTS

Tectonics related to the Coprates horst/graben system were reactivated after the channel’s creation in the oldest episode (Figures 4E,4F), producing several contractions and faults that have altered its shape. Consequently, applying the mechanical model directly to the observed topography today results in the material not flowing properly. Therefore, a "backward reconstruction" is required to reverse these modifications without excessively altering the original shape of the surface so we can provide a valid input for the mechanical modeling. Thanks to the method used (Figure 3), we obtained the channel topographies preceding (Figures 4A,4E) and following (Figures 4B,4F) the two episodes, removing the sedimentary deposits at its mouth.

 

Figure 4. "Forward reconstruction" of the youngest (A-D, orange in Figures 1 and 2) and oldest (E-H, blue and red respectively in Figures 1 and 2) episodes. Topography before (A and E), after (B and F), and the event and initial condition were obtained with the difference between both (C and G). The result of the numerical modeling and comparison of the observed fan shape (bottom) and its reconstruction (top) are shown for the youngest (D) and the oldest (H) episodes. 

 

For the "forward reconstruction,” we conducted multiple simulations of the formation of these deposits using different rheological laws for Newtonian, frictional, and Bingham fluids [7,8]. The mixture’s density, viscosity, and yield strength are related by the constitutive equation. The best match between simulations and actual observations is achieved for a Bingham-type mudflow consisting of a kaolinite-water mixture [4]. The interstitial water content is calculated from the empirical expression for kaolinite-type water-clay mixtures [10]. Table 1 shows the final parameters that best fit the observed morphologies.

	Youngest episode	Oldest episode
Density (kg/m3)	1722	1606
Viscosity (Pa·s)	0.114	0.958
Elastic limit (MPa)	1.53	0.07
Water content (%)	55	62

Table 1. The final parameters used in the mechanical model to obtain the results are shown in Figure 4D (youngest episode) and 4H (oldest episode).

Regarding the most recent episode, our results (Figure 4D, Table 1) are compatible with a kaolinite-type clay-water mixture mudflow with an interstitial water content of about 55%. For the older episode, using the same material and water content as in this episode, the "forward" reconstruction of the event by numerical simulation does not reproduce the observed final shape and height of the deposit. With these same parameters and considering different depaleolakes depths filling the canyon, as suggested by [3], it does not match the observations either. However, by increasing the interstitial water content of the sliding mass to 62% and defining a 1500m deep water sheet, the reconstruction of the event does match the observations (Figure 4H, Table 1). This finding supports the hypothesis of two different episodes for the formation of the fan, with different amounts of water, and the existence of a paleolake of considerable depth during the oldest episode studied.

 

4.ACKNOWLEDGMENTS

A.Molina thanks EDRIM project (VAPC 202250I104), D.Pascual JAE Intro 2022 program (JAEINT_22_00950, JAEINT22_EX_0770), T.Martínez-Pérez PTA2022-021846-1(MCIN/AEl 10.13039/501100011033yFSE), and A.Caramanico INTIME project (Grant agreement ID: 823934; DOI 10.3030/823934).

 

5.REFERENCES