C. Brenhin Keller , T. Mark Harrison

Accurately quantifying the composition of continental crust on Hadean and Archean Earth is critical to our understanding of the physiography, tectonics, and climate of our planet at the dawn of life. One longstanding paradigm involves the growth of a relatively mafic planetary crust over the first 1-2 billion years of Earth history, implying a lack of modern plate tectonics, a paucity of subaerial crust, and consequently lacking an efficient mechanism to regulate climate. Others have proposed a more uniformitarian view in which Archean and Hadean continents were only slightly more mafic than at present. Apart from well-known complications in assessing early crustal composition introduced by crustal preservation and sampling biases, emerging effects such as the secular cooling of Earths mantle and the dramatic, biologically-driven oxidation of Earths atmosphere have not been fully investigated. We examine the limits imposed by these constraints on published models and find that several existing studies are not robust against these complications. We find that the data are most parsimoniously explained by a model with nearly constant crustal SiO$_2$, in which Earths early crust may be considered more mafic in relative to modern compatible element abundances, but not silica.