Ruedas, Thomas, Paul J. Tackley, and Sean C. Solomon. “Thermal and compositional evolution of the martian mantle: Effects of phase transitions and melting.” Physics of the Earth and Planetary Interiors 216 (2013): 32-58.
URL1 URL2
We present new numerical models for the thermochemical evolution of the mantle of Mars over the past 4 Gy. Specifically, we have developed a parameterized model of composition and thermoelastic properties of mantle material and combined it with a two-dimensional, anelastic, compressible convection and melting algorithm in a spherical annulus geometry; these models include a detailed treatment of the effects of solid–solid phase transitions and of compositional changes that accompany generation and removal of mantle partial melt during magmatism. Among other questions, we examine if a perovskite + ferropericlase layer (pv + fp) exists at the base of the martian mantle, if long-lived plumes can explain the volcanic provinces, under which circumstances Mars may still be volcanically active, and how iron- and radionuclide-rich Mars is. Results of the models are compared with geophysical and chemical observations from spacecraft and information from martian meteorites. Most models yield crustal thicknesses between ∼75 and 90 km, ancient depths for the Curie temperatures of possible magnetic carriers that include the entire crust, and mechanical lithosphere thicknesses that increased from less than 100 km in the Noachian to 200–250 km now. Generally, models with a large core, Mg# = 0.75, and radionuclide contents derived from those suggested by Wänke and Dreibus (Chemistry and accretion history of Mars. Philos. Trans. R. Soc. Lond. A 349, 285–293, 1994) tend to explain observations best, but none of the models reproduces the full range of concentrations found in martian meteorites, thus implicitly supporting the idea that ancient chemical heterogeneities survive in the mantle. Only a subset of the models develops a pattern of mantle convection that evolves towards two or three large, long-lived plumes, and it takes at least ∼2 Gy before this stage is reached, whereas models with a larger number of plumes or with plumes that are less stationary are more common. As Tharsis and probably Elysium are older than 2.5–3 Gy, either model assumptions more complex than those made in this study or a dynamical mode different from low-degree convection with long-lived large plumes is required to explain these major volcanic provinces. The effect of the mid-mantle phase transitions on mantle dynamics is rather weak, so that whole-mantle convection prevails; however, a basal pv + fp layer would form a dynamically separate unit and decouple the mantle and core to some extent.