Interior Dynamics and Thermal Evolution of Mars – a Geodynamic Perspective

Fully dynamical 3D simulations that self-consistently model the evolution of mantle flow are routinely used to investigate the thermal history of the interior of rocky planets, and through their nature, they can address both global and local observations in regions of interest. Over the past years, models of the thermal evolution of Mars have been combined with crustal thickness variations derived from gravity and topography data, tidal deformation models, and calculations of the seismic velocities and seismogenic layer thickness to estimate the present-day seismicity.

Recent data from the InSight mission about the crustal thickness, upper mantle structure, and core size provide the most direct constraints on the interior structure of Mars. Here we use InSight’s estimates in addition to constraints from tidal deformation, elastic lithosphere thickness, and recent volcanic activity to model the thermal history and present-day state of the interior of Mars.

Similar to previous studies (Plesa et al., 2016, 2018), the geodynamical simulations include crustal thickness variations (Fig. 1a). We use the most recent global crustal thickness models of Mars (Wieczorek et al., 2022) that are derived from gravity and topography data and constrained by crustal thickness estimates at the InSight landing site from seismic observations (Knapmeyer-Endrun et al., 2021; Wieczorek et al., 2022). 

Fig. 1: a) crustal thickness model derived from gravity and topography data and anchored by the crustal thickness at InSight location; b) corresponding heat flow and c) elastic thickness variations at present day; d) average melt fraction and e) average melt zone depth at present day; f) seismogenic layer thickness computed using the 573 K isotherm.

The present-day thermal state in the 3D geodynamical models is obtained by running the entire thermal evolution of Mars. The models assume that 55 to 70% of the bulk amount of heat producing elements are stored in the crust as suggested by Knapmeyer-Endrun et al. (2021) in order to avoid wide-spread melting in the interior of Mars at present day. The strong crustal enrichment, and thus a mantle depleted in heat sources, leads to an average present-day lithospheric thickness of about 500 km in good agreement with the thermal lithosphere thickness estimated from the analysis of seismic events (Khan et al., 2021; Duran et al., 2022).

Our models show that crustal thickness variations have a first order effect on the temperature distribution in the shallow mantle. Thus, the patterns of the surface heat flow and elastic lithosphere thickness at present day correlate with the crustal thickness variations (Fig. 1b,c). The present-day elastic lithosphere thickness at the north pole of Mars is one of the strongest constraints for thermal evolution models and indicates that the mantle contains less than 45% of the total heat production and/or that the polar cap has not yet reached elastic equilibrium (Broquet et al., 2020).

The radius of the core in our models has been set to 1850 km, a value that lies well within the current core radius estimates (Stähler et al., 2021). Such a large core prevents the stability of an endothermic phase transition from wadsleyite/ringwoodite to bridgmanite that was previously proposed to explain the formation of the martian crustal thickness dichotomy and Tharsis through a low degree convection pattern (Harder & Christensen, 1996, Breuer et al., 1998). Current models show a mantle characterized by several mantle plumes distributed throughout the mantle. During the evolution, strong mantle plumes that produce melt up to recent times become focused in regions covered by a thick crust, in particular in the Tharsis province (Fig. 1d,e). 

We compute the tidal deformation using the interior structure and present-day thermal state from the geodynamic models. For the tidal deformation calculations, we use a semi-analytical model based on the normal mode theory for radially stratified viscoelastic bodies (Sabadini & Vermeersen, 2004) and choose an Andrade rheology. Results show that thermal evolution models with a large core and a dry mantle viscosity can match the latest tidal deformation estimates determined from radio tracking measurements from Mars Odyssey, Mars Reconnaissance Orbiter, and Mars Global Surveyor (Konopliv et al., 2020). 

Similar to the study of Plesa et al. (2021), we compute the seismic velocities in the entire 3D thermal evolution model. We use the mantle compositions of Taylor (2013), Yoshizaki & McDonough (2020), and include the most recent composition of Khan et al. (2022). Independent of the assumed compositional model, seismic velocity variations show a pattern similar to that of the crustal thickness and these variations can extend to depths of 400 km or deeper for models with a thick lithosphere. 

Our models can be used to estimate the seismogenic layer thickness and the present-day seismicity of Mars and to compare the results with the recent estimates from InSight (Banerdt et al. 2020). Using isotherms of 573 K and 1073 K that have been suggested to mark the base of the seismogenic layer (Knapmeyer et al., 2006 and references therein), we calculate the thickness of the seismogenic zone at each location on Mars (Fig. 1f) and estimate the present-day seismicity. The seismogenic layer thickness follows crustal thickness variations and could be used to discriminate between geodynamic models, if the source depth and location of seismic events is known. The present-day seismicity calculated from the geodynamic models that use the crustal thickness and core size estimates from InSight is compatible with the values derived from InSight's seismic data. 

Future models need to analyze in greater detail the consequences of the large present-day elastic lithosphere thickness at the north pole of Mars from recent estimates on the present-day thermal state. Geodynamic models combined with seismic waves propagation calculations need to investigate the effects of spatial variations of seismic velocities in the mantle and lithosphere, and future studies are necessary to test the spatial distribution of seismicity by comparing model predictions to observations.