Zircon U/Th model ages in the presence of melt heterogeneity

In situ U-series zircon dating has yielded unique insights into magmatic processes and the complexity of zircon crystallization. However, the approach requires some knowledge of the state of isotopic disequilibrium of the melt from which zircon crystallizes. Current practices for correcting initial 230Th include use of an isochron array (defined by several coexisting zircons) or two-point isochrons based on a tie between each zircon and a common glass (or whole rock) measurement. However, magmas are complex and measured U/Th in zircons from a single extrusive can vary by up to a factor of seven, casting doubt on the assumption that a single glass composition is representative of the magma from which each zircon crystallized. We propose a correction scheme using the measured zircon 238U/232Th, a U/Th partition coefficient ratio between zircon and melt, and the observation that most magmas are within 15% (1σ) of the equiline. Using this correction scheme, we show that uncertainties can be underestimated by up to a factor of three and that published dates are potentially biased towards older ages.