BETTER DESCRIPTIONS OF U-Pb ZIRCON DATA AND APPLICATION OF BIAS ANALYSIS ON U-Pb AGES FROM SPOT PICKING WITH CATHODOLUMINESCENCE (CL) IMAGES OF ZIRCON

Laser Ablation-Inductively Coupled Plasma-Mass Spectrometry (LA-ICP-MS) is a cost-effective technique for in-situ U-Pb zircon geochronology due to high sample throughput and availability. LA-Quadrupole-ICP-MS is at the extreme end of cost-effectiveness but the method has low precision due to low ion transmission and duty cycle losses. These characteristics are problematic for young zircons (≤ 5Ma) that have ²⁰⁷Pb counts and ²⁰⁷Pb/²⁰⁶Pb within error of zero or negative values. This calls into question the assumption of Gaussian statistics (normal distribution) that are typically used to characterize ²⁰⁷Pb/²⁰⁶Pb. In this situation the zero-inflated Poisson (ZIP) or truncated normal (TN) models can be used to circumvent these issues because they are bound at zero. Furthermore, ZIP accounts for both sources of zeroes inherent to time-resolved data close to zero: The possibility that no counts are present (‘true’ zero) and zeroes generated from random variability.

Two zircon splits were separated from samples and measured by LA-Q-ICP-MS and the more precise LA-MultiCollector-ICP-MS method. Intrasample comparison indicates that while the quadrupole data are imprecise they agree with LA-MC-ICP-MS data. ZIP and Poisson distributions were used respecitvely on ²⁰⁷Pb and ²⁰⁶Pb to determine a ²⁰⁷Pb/²⁰⁶Pb ratio. Errors are reduced by c.a. 2% compared to Gaussian statistics, confirming that limiting precision is lower than what is produced by the normal distribution. Likewise, the TN produces lower uncertainty. We suggest use of either ZIP or TN for more accurate descriptions of the ²⁰⁷Pb/²⁰⁶Pb ratio in young zircons.

Previously collected SIMS U-Pb ages from one sample are offset relative to both LA-ICP-MS methods. There is a dearth of xenocrysts or antecrysts in the SIMS ages as well. Quantitative bias models show choice of spot selection based on CL-texture affected U-Pb ages in the SIMS study. Accounting for these biases may drastically alter interpretations of magmatic systems including but not limited to lack of assimilation inferred by absence of xenocrysts and antecrysts, incremental emplacement of discrete melt batches inferred from temporally disparate ages in differing pluton lithologies, or short-lived large volcanic systems. We suggest interpretation of U-Pb zircon data should account for bias in spot selection, be the bias accidental or from targeted analyses.