Inferences on the mantle viscosity structure and the post-overturn evolutionary state of Venus

Venus has similar size, density and bulk composition as Earth, but has tectonically evolved clearly differently and this divergence remains enigmatic. Surface observations such as gravity, topography and surface age constrain Venus' evolution, but interpreting these signals requires understanding of the surface-interior coupling and thus insight into the structure and evolution of the venusian mantle and lithosphere. Here, we investigate how such observables may be generated from interior dynamics using numerical forward models of global mantle convection that consistently link the thermochemical, magmatic and tectonic evolution of Venus. Venus' present surface gravity spectrum and its relation to topography is matched best by our models with a mantle viscosity profile featuring a sublithospheric minimum of  ∼ 2 × 1020 Pa s and a gradual increase by a factor of  ∼ 100 down to a depth of  ∼ 250 km above the core-mantle boundary. No pronounced viscosity jump around the mantle transition as inferred for Earth is favoured for Venus, which points to a relatively dry venusian upper mantle compared to Earth's as previously suggested. This holds true for both a pure stagnant-lid scenario and in the presence of episodic catastrophic overturns triggered by cumulative crustal growth due to on-going magmatism and volcanism. Overturns strongly perturb the surface gravity spectrum up to  ∼ 150 Myr after overturn cessation. Material deeply recycled by the resurfacing event annihilates the developed plume pattern, which needs much longer than those 150 Myr to recover to a state comparable to the pattern suggested by thermal emissivity anomalies observed on Venus. Moreover, overturns limit crustal thicknesses to reasonable values and are more capable than stagnant-lid evolutions in generating mean surface ages  > 500 Myr. These findings seem to confirm previous suggestions that the episodic regime is more applicable to Venus than a purely stagnant-lid regime. Yet, the relatively long time span required to recycle the entire surface (∼150−200 Myr) and the presently on-going volcanic resurfacing predicted by our models complicate the formation of a uniform surface age as indicated by Venus' crater population and may also suggest that the latest overturn has ceased longer ago than indicated by Venus' present mean surface age.