A Tale of Two Earths: Reconciling the Lunar and Terrestrial Hadean Records
- Author(s): Boehnke, Patrick
- Advisor(s): Harrison, T. Mark
- et al.
Studying early Earth history is complicated by the fact that the rock record doesn’t extend past 4 Ga and our only record for the Hadean (>4 Ga) comes to us from detrital zircons from the Jack Hills in Western Australia. The Hadean zircon record extends back to almost 4.4 Ga and has revealed that the early Earth may have had liquid water, a felsic crust, plate boundary interactions, and possibly a biosphere. On the other hand, analyses of lunar and meteoritic samples are used to argue for a hellish Hadean Earth where frequent, large impactors repeatedly destroyed the crust. Indeed, these two models stand in direct contradiction. The focus of this thesis is to examine the evidence for these two models and ultimately propose a reconciliation based on a new interpretation of the chronology of the lunar samples used to constrain the impact history into the early Earth-Moon system.
In order to improve the understanding of zircon crystallization in igneous settings, we undertook experimental studies of zircon saturation which were analyzed using a novel ion imaging approach by a secondary ion mass spectrometer. This study confirmed the original model for zircon saturation, that it is a function of only temperature, melt composition, and Zr content. Indeed, the primary implication for the early Earth from this work is that zircons are much more likely to crystallize in a felsic rather than mafic magma and therefore simply the existence of Hadean zircons suggests a high likelihood for felsic Hadean magmatism.
The majority of the thesis focuses on the interpretation of 40Ar/39Ar ages of lunar and meteorite samples, specifically with regards to impact histories derived from compilations of such ages. The primary complication with lunar and meteorite 40Ar/39Ar ages is that the vast majority show evidence for later disturbances due to diffusive loss of 40Ar. To try and extract meaningful thermal histories from these samples, we undertook investigations of samples from Apollo 16 and the Jilin chondrite. We then used an extension of the multi-domain diffusion model that can model samples containing multiple activation energies (i.e., whole rock samples with multiple K bearing minerals) to propose that the 40Ar/39Ar system can be used to recover shock heating temperatures and durations.
Having shown the effects of diffusive 40Ar loss on the accuracy of 40Ar/39Ar dating, we then explored the question as to whether or not compilations of disturbed 40Ar/39Ar ages simply misestimate the timing of bombardment episodes or are fundamentally inaccurate. For this we created a simple numerical model that simulates a chosen impact history on a surface and then creates a histogram of 40Ar/39Ar plateau ages. Our results show that rather than simply misestimate timing, compilations of 40Ar/39Ar ages can lead to inferences of illusory bombardment episodes.
Finally, we examine the 40Ar/39Ar ages of suite of geochemically related Apollo 16 rocks to examine the effects of mixing and brecciation on the accuracy of inferred ages. By analyzing multiple rocks from each soil sample, we show that three out of six samples are not compatible with a single thermal history. That is to say, despite their close proximity during sampling and geochemical similarities, analyzed rocks in the soil sample have unique chronologies. Based on these findings, we developed a simple numerical model which shows that internal isochrons of mixed samples can yield erroneous ages while retaining a statistically acceptable mean squared weighted deviation (MSWD).