The low-friction and viscous detachments drive deformation on Mars

Introduction

The Tharsis region on Mars represents one of the most prominent volcanic and tectonic provinces in the Solar System, shaped over the past four billion years through sustained magmatic and tectonic activity [1]. The prevailing hypothesis attributes the formation of Tharsis to a long-lived mantle plume beneath a relatively thin lithosphere [2,3]. Alternative models propose a superplume origin [4] or invoke a combination of isostatic uplift, lithospheric flexure, volcanic loading, and intrusive thickening [5,6]. Simulations have been employed to test these hypotheses and explore variations in crustal thickness and mantle flow [7,8], while geological and structural observations support the presence of plume-induced stress centers that have governed the development of step-like topographic decreases from the center outward and the formation of compressional landforms [9,10].

A key manifestation of this tectonic evolution is the widespread system of compressional landforms called “wrinkle ridges”, which encircle the Tharsis rise and extend into the Northern Plains. These landforms, typically several hundred meters in height and tens of kilometers in width, are best preserved around Syria, Thaumasia, and Lunae Planum [11]. Their morphology, spacing, and radial distribution suggest a genetic link to underlying detachment and a link with Tharsis uplift. Their formation remains debated due to limited subsurface data, with hypotheses ranging from fault-propagation folding to blind thrusting over detachments [12–13]. In several regions, wrinkle ridges display regular spacing that decreases with distance from the Tharsis center, particularly between Sinai Planum and Solis Dorsa [14].

This study investigates (i) the mechanical connection between circumferential compressional features and large-scale detachment and (ii) the influence of detachment rheology—frictional versus viscous—on stress transfer within wedge-shaped topography. These questions are addressed using two-dimensional thermomechanical models driven by gravitational forces arising from isostatic and volcanic loading, in the absence of external tectonic boundary conditions.

Methodology

Primary datasets were sourced from NASA’s Planetary Data System (PDS) and the Mars Global GIS repository. Topographic profiles were extracted from the MOLA-HRSC blended digital elevation model at 200 m/pixel resolution [15]. THEMIS daytime infrared mosaics (~100 m/pixel) [16] were used as the base for identifying circumferential wrinkle ridges [17]. We applied the Swath Profiler Toolbox [18] to derive average topographic cross-sections within 1,500 km-wide swaths focused on well-preserved wrinkle ridges. Features overprinted by Amazonian-era resurfacing were excluded from analysis.

Numerical simulations were performed using a finite-difference thermo-mechanical code [19,20] based on a marker-in-cell approach, which couples a fixed Eulerian grid with freely advected Lagrangian markers. Governing equations for conservation of mass, momentum, and energy were solved on a static grid, while the fourth-order Runge-Kutta method was used to update marker positions. Our domain consisted of an 8 km-thick sticky-air layer above a 7 km-thick upper crustal unit characterized by a quartzite rheology. Below this, a 2 km-thick mechanically weak detachment layer was introduced, implemented either as low-frictional (fluid-overpressured shale) or low-viscosity (salt). The base of the crust comprised an 8 km-thick megaregolithic layer modeled using dry olivine flow laws. Detachment depths were validated by the shallow seismic profiles of the InSight mission [20]. We conducted 60 exploratory runs to probe the parameter space and selected 10 representative models for detailed analysis. For the frictional case, mobile shale detachment zones were simulated under fluid overpressure ratios from 0.7 to 0.99. In viscous cases, the salt layer viscosity was varied from 10¹⁷ to 10¹⁹ Pa·s.

Results and Conclusion

Our numerical modeling reveals distinct behaviors for low-frictional versus viscous detachments. Specifically, for the wedge structure, notably low alpha and beta angles imply that the mechanically homogeneous wedge requires an exceptionally low friction coefficient to match the gravity-driven geometry. This supports the presence of a low-friction detachment, potentially shale, possibly activated by fluid overpressure. Simulations with an initial overpressure of 0.7 show no deformation, with localization beginning only beyond a threshold of 0.85. Between 0.85 and 0.99, proximal normal faulting is followed by deformation in a transmission zone and the formation of a distal thrust front. In contrast, viscous detachment models—representing salt with a viscosity of ~10⁷ Pa·s—produce diapiric upwellings, with normal faults shifting from radial to offset configurations, indicating unstable wedge deformation. When viscosity increases to 10⁸ Pa·s, the detachment transmits stress from the Tharsis uplift without localized diapirism, aligning best with observed topography.

Here, we propose that the presence of such low-friction detachment is likely linked to alteration of pyroclastic materials, and a viscous detachment, particularly around Tharsis, and in the Valles Marineris region, may be attributed to previously suggested [22] chlorite-based hydrothermal alteration processes. This mechanism could plausibly reduce both the frictional strength and viscosity of the detachment zone, thereby facilitating the observed tectonic features.

References

[1] Carr & Head (2010) EPSL, 294, 185–203. [2] Dohm et al. (2001) JGR, 106, 32943–32958.
[3] Breuer et al. (1996) JGR, 101, 7531–7542. [4] Baker et al. (2007) Superplumes, 507–522.
[5] Banerdt et al. (1982) JGR, 87, 9723. [6] Thurber & Toksöz (1978) GRL, 5, 977–980.
[7] Breuer et al. (1998) GRL, 25, 229–232. [8] Harder & Christensen (1996) Nature, 380, 507–509. [9] Anderson et al. (2001) JGR, 106, 20563–20585. [10] Mège & Masson (1996b) PSS, 44, 1471–1497. [11] Watters (1988) JGR, 93, 10236–10254. [12] Schultz, R. A. (2000) JGR: Planets, 105(E5), 12035–12052. [13] Karagoz et al. (2022) EPSL, 595, 117759. [14] Montési, L. G., & Zuber, M. T. (2003) JGR: Planets, 108(E6), 5048. [15] Fergason et al. (2018) USGS.
[16] Edwards et al. (2011) JGR, 116, E10008. [17] Knapmeyer et al. (2006) JGR, 111, E11008.
[18] Pérez-Peña, J. V. et al. (2017) Comput. Geosci., 104, 135–150. [19] Gerya (2019) Introduction to Numerical Geodynamic Modeling. [20] Ruh et al. (2020) Front. Earth Sci., 8, 605239. [21] Ehlmann, B. L. et al. (2011) Nature, 479(7371), 53–60. [22] Osterloo, M. M. et al. (2008) Science, 319(5870), 1651–1654.