EVALUATION OF COMMONLY USED METHODS FOR CALCULATING DETRITAL ZIRCON MAXIMUM DEPOSITIONAL AGES

Maximum depositional ages (MDAs) calculated from detrital zircon populations are commonly used to investigate the age of strata for the correlation of stratigraphy, the calculation of sedimentary rates, and correlation of sedimentary deposits to the geologic timescale. In these applications the youngest zircon, or sub-sample of zircon grains, from a rock can be used to constrain its maximum age as the rock must be younger than its component parts. Methods to calculate MDAs range from taking the youngest date of the sample, to Monte Carlo simulations of individual date uncertainties. The calculation of accurate MDAs is subject to many geological uncertainties and measurement-method uncertainties. Although many studies use these data (>400 publications in 2017) and multiple studies have strived to assess the accuracy of these methods in natural datasets, no direct comparison of MDA methods has been done in a numerical simulation where all parameters that impact the accuracy of calculated MDAs can be controlled.

To understand the impact that calculation method, as well as sample size (n), abundance of ages that approximate true depositional age (TDA), and measurement method uncertainty have on MDA accuracy, we test 10 MDA methods in two simulation scenarios. In the first simulation scenario we vary the abundance of TDA-equivalent ages between 1 and 10% of the total population. In the second we vary the uncertainty attributed to each synthetic date holding TDA-equivalent ages static at 3% of the total population. In each simulation a variable amount of synthetic dates (50 – 1000 in increments of 50) are drawn from the parent population (N≈25 000) 500 times at each sample size and the MDAs of the 10 methods are calculated. We found that increasing sample size to large-n (n>300) greatly reduces the residual error and variance of calculated MDAs. Additionally, the calculation of MDAs from high precision dates and parent populations with abundant TDA-equivalent ages also decreased the residual error of calculated MDAs. No single MDA method performed best in all scenarios. Lastly, as MDA uncertainties scale with strata age, we use our results to test what geological processes can be resolved using MDAs. The results of this study can help shape how datasets are acquired for MDA analysis and what geological questions can be answered by them.