UC Santa Cruz

This dissertation examines the influence which a geophysical process (giant impacts) has on a geochemical marker (composition) during terrestrial planet formation. Simultaneously studying all planets maximizes the available constraints and permits examination of controls on the overall composition of the Earth. I also examine the Galilean satellite system to determine the universality of the terrestrial conclusions.

The late stages of planetary accretion involve stochastic, large collisions. Impact-related erosion and fragmentation can have profound consequences for the rate and style of accretion and the bulk chemistries of terrestrial planets. However, the previous predominate assumption in computer models of accretion was that all collisions resulted in perfect merging despite the likelihood of these collisions producing a range of outcomes (e.g., hit-and-run, removal of material from target, or production of several post-collision bodies). In this work, I investigate the effects of late-stage accretion with multiple collision types and the consequences on the bulk (mantle/core) and isotopic (Hf--W) composition.

My model is composed of two parts: (1) N-body accretion code tracks orbital and collisional evolution of the bodies and (2) geochemical post-processing evolves composition in light of impact-related mixing, partial equilibration and radioactive decay. For terrestrial planets, Part (1) is Chambers (2013, Icarus) and incorporates multiple collisional outcomes. For Galilean satellites, Part (1) is Ogihara & Ida (2012, Icarus) and assumes perfect merging for all collisions thus the model is not self-consistent (it likely overestimates compositional changes).

For the terrestrial planets, the results are consistent with observed mantle/core ratios and tungsten isotopic anomalies. A moderate (approx. 0.4) core equilibration factor is preferred due to protracted accretion time. It is important to include multi-modal collisions when modeling planet formation if composition, timescales, or spatial distribution of mass are being investigated.

I could not reproduce the observed ice fraction gradient of the Galilean satellites, even with an initial compositional gradient and vaporization of water ice. Some other physical process(es) are needed, perhaps tidally-driven volatile loss at Io and Europa. Extensive inward radial migration smooths out initial compositional gradients.