Analyses from a validated global U-Pb detrital zircon database: Enhanced methods for filtering discordant U-Pb zircon analyses and optimizing crystallization age estimates

Determining the best age of individual grains from U-Pb detrital zircon analyses remains problematic, with the current system generally relying on an arbitrary cut-off age for determining when to switch from Pb-206/U-238 ages to Pb-207/Pb-206 ages. Recently, some found that iteratively calculated IsoplotR single grain concordia ages provide estimates with increased precision and have turned to using these estimates as the preferred age. Here, we compile a new 604,553 record global detrital zircon database containing U-Pb ratios, ages, uncertainties, laboratory settings, among other variables. The new database provides a means for investigating four best age models from the standpoint of accuracy. The models are (1) 1700 Ma cut-off for best age, (2) 1000 Ma cut-off for best age, (3) iteratively calculated IsoplotR single grain concordia ages, and (4) a new non-iterative model developed by comparing U-Pb ages of zircon standards dated by both ID-TIMS and LA-ICP-MS spectrometry. We then classify all analyses by their degree of concordance and calculate unweighted age distributions. The age distributions are nearly identical for all models when highly concordant analyses are used. However, as discordance increases, the models with arbitrary cut-offs produce age distributions with artificial depressions at the cut-off age. After segregating database records into five concordance classes, cross-correlation analyses show the five non-iterative, class-based, age distributions are aligned and correlated better than those from the other age models. Rather than abruptly transitioning from Pb-207/Pb-206 ages to Pb-206/U-238 ages at a specific cut-off age, the results support a cumulative distribution model that defines a gradual transition from the opposing geochronometers over an interval of similar to 1200-myr, primarily encompassing the span from 1600 Ma to 400 Ma. Additionally, a supplementary template is developed to provide geochronologists with a means to validate U-Pb related data, calculate the new concordance classifications, and calculate non-iterative ages.

Automated summary

This study compares the accuracy of four different best age models using a new global detrital zircon database, finding that the age distributions based on the five non-iterative class-based models are better correlated.