Source: Dissertation Abstracts International, Volume: 70-06, Section: B, page: 3388.

Adviser: David Bercovici.

A major question in Earth and planetary science is why does Earth exhibit plate tectonics and the other terrestrial planets do not? The difference in tectonic behavior between Earth and Venus is even more striking given their similarity in size, composition, and average density. While our general knowledge of material behavior suggests that all the terrestrial planets, Earth included, should not exhibit plate tectonics, the requisite weakening allowing this tectonic style is in operation on our planet but not Venus. To address this set of questions I have taken a two-part approach: development and implementation of a two-phase damage rheology in models of lithospheric deformation and planetary convection to determine the plate generating capabilities of such a shear localizing model, and consideration of the effect of a planet's climate on a lithospheric damage rheology and how this can explain the different modes of convection displayed by Earth and Venus.

I have implemented a two-phase damage rheology into a two-dimensional simple-shear calculation as a means for understanding the depth-dependent localization occurring in the lithosphere. The results from this work imply that while void-generating damage is likely only important in the upper 15 km of the lithosphere, grain size reducing damage is able to produce strong localization. The combined interaction of void and grain size reducing damage can also lead to the formation of a central shear zone dominated by small grains with high porosity bands bordering this region. Implementation of two-phase damage rheology into a convectively driven lithospheric-mantle model suggests similar behaviors, namely that grain size reducing damage is far more efficacious at producing localization necessary for plate generation.

The hypothesis regarding the effect of climate on planetary convection suggests that temperate surface temperatures (due to the role of free surface water controlling the carbon cycle) moderate healing due to grain-growth in the lithosphere, thereby leading to enhanced damage and localization necessary for plate boundary formation. Venus on the other hand has much higher surface temperatures 400 K higher), which leads to increased healing and a suppression of localization and plate tectonics. I tested this hypothesis in a convectively driven lithospheric-mantle model along with a scaling analysis, and I found that variations in the lithospheric healing rate (which serves as a proxy for surface temperature) can lead to a transition in planetary convective modes. A more rigorous test of this hypothesis was attempted by developing a variable viscosity convection model for planetary convection. Here I tested the effect of varying surface temperature on the development of localization with a grain size and temperature dependent rheology. The results from this work suggest that a surface temperature increase of approximately 250 K relative to Earth values could lead to a transition from plate tectonic convection to an episodic mode of convection potentially similar to what is exhibited by Venus.