GSA Annual Meeting in Phoenix, Arizona, USA - 2019

Paper No. 65-7
Presentation Time: 3:20 PM


SUNDELL, Kurt E., Department of Geosciences, University of Arizona, Tucson, AZ 85721, BLUM, Mike D., Geology Department, University of Kansas (KU), 1414 Naismith Drive, Room 254, Lawrence, KS 66045, GEHRELS, George, Dept. of Geosciences, University of Arizona, Tucson, AZ 85721, PECHA, Mark, Department of Geosciences, University of Arizona, Gould-Simpson Building #77, 1040 E 4th St, Tucson, AZ 85721 and PETTIT, Bridget S., ExxonMobil, Upstream Business Development Company, Spring, TX 77389

Detrital zircon (DZ) U-Pb analysis by laser ablation inductively coupled plasma mass spectrometry (LA-ICP-MS) is a powerful technique for application to sediment provenance analysis and ascribing maximum depositional ages (MDAs). However, there are many complicating factors that result in age distribution bias. Among these, depositional environment holds a significant control on resulting DZ age distributions. Specifically, studies of modern and ancient fluvial systems commonly call on hydrodynamic sorting in fluvial systems as a primary cause of bias in DZ U-Pb age distributions, and have questioned provenance interpretations from them.

We developed a method of acquiring DZ U-Pb ages by multicollector LA-ICP-MS at a rate of 3s/analysis to increase sample throughput and decrease analytical cost, with the goal of increasing sample size (n) to produce more robust age distributions. Round-robin testing of this new rapid acquisition method using zircon reference materials that range in age from ~28 Ma to ~3465 Ma yields 0.5–3% accuracy and 2–10% precision; this allows for acquisition of very large-n data sets (n>1000) at a rate of 1,200 analyses/h. Automated grain selection directly ties individual grain characteristics (e.g., grain size, color, shape) to each age date, and removes all user bias from the acquisition process.

We apply this new method to sandstone samples from Campanian Book Cliffs strata, UT. Results show that both small-n (n≈100) and large-n (n≈300) data are biased by grain size, and that quantitative comparison among different methods are inconsistent. Grain-size effects are still present in very large-n sample sets (n≈1,000), and easily identified with our new method, but quantitative comparisons of these data yield consistent results for different comparison methods and an overall increase in similarity among samples that approaches a common underlying mixture of upstream detritus farther downstream. Further analysis of sample pairs that bracket chronostratigraphic boundaries holds significant implications for sequence stratigraphic interpretations via the calculation of robust MDAs. Increasing sample size in this fashion requires no additional cost or time investment, and comes with minimal sacrifice in accuracy and precision to yield more robust estimates of DZ age distributions.