A Positive feedback mechanism on crustal growth at the origin of the Martian dichotomy

The dichotomy in topography on Mars likely results from a dichotomy in crustal thickness. Insight seismological data provided the first estimate of the Martian crustal thickness at the landing site that is ~20±5 km or 39±8 km depending on the number of crustal layers. This estimate, together with the inversion of topography and gravity data, provides a global crustal thickness map of Mars and constrains the amplitude of the hemispheric difference in crustal thickness. Assuming the same crustal density for both hemispheres, the southern crust appears on average thicker than the northern one by at least 12 km and up to 33 km. Several explanations have been proposed for the origin of this dichotomy, involving external processes, such as a large impact, or internal ones, such as a degree-one convection. Here we propose that the development of the crustal dichotomy may have resulted from an initial crustal or lithospheric thickness anomaly and a positive feedback between crustal thickness growth and partial melting in the mantle. Indeed, because the crust is enriched in heat-producing elements, the lithosphere of a one-plate planet is thinner where the crust is thicker, inducing a lower pressure at the base of the lithosphere. Because of the pressure-dependence of the mantle solidus, partial melting is more important at low pressure causing a larger rate of crustal extraction and growth where the crust is thicker. To model this, we used a parametric bi-hemispheric thermal evolution model which assumes a well-mixed mantle. A small initial perturbation in lithosphere or crust thickness is imposed between the North and South and we follow crustal extraction and growth in both hemispheres. The thermal history of Mars is then calculated over 4.5 Gyr for a large number of parameters. Our results show that this positive feedback can indeed create a significant crustal dichotomy with a North-South crustal thickness difference that depends on model parameters and easily explains the difference revealed by InSight, reaching even more than 100 km for some sets of parameters. In addition, our modeling shows that, for large enough crust thicknesses, temperatures allowing for partial melting and differentiation can be reached at the base of the crust during its growth, in particular in the hemisphere showing the thickest crust.