Exciting news! We're transitioning to the Statewide California Earthquake Center. Our new website is under construction, but we'll continue using this website for SCEC business in the meantime. We're also archiving the Southern Center site to preserve its rich history. A new and improved platform is coming soon!

Dear Prudence: how many surface clasts are required to yield an accurate exposure date?

Veronica B. Prush, & Michael E. Oskin

Published August 15, 2018, SCEC Contribution #8786, 2018 SCEC Annual Meeting Poster #231

Prush and Oskin (submitted) propose a new statistical model for determining the exposure age of offset fluvially deposited surfaces dated using surface clasts. Our model requires fitting a generalized Pareto distribution (GPD) to these datasets, rather than the commonly assumed Gaussian distribution, to model the inherited component of exposure age. In our study we use a Monte Carlo Markov Chain algorithm to fit the GPD model to published Beryllium-10 datasets that include 8 or more surface clasts. Such small sample size reflects current sampling practices and allows us to analyze 64 published sites. Unfortunately, such small sample sizes are inadequate to recover the true distribution of inheritance, and thus limits prediction of surface age. To determine the number of clasts required to adequately sample the GPD, we use our algorithm to recover the age and other fit parameters from populations of clast ages randomly drawn from a known distribution. Our results suggest that approximately 20 clasts are required for the expected value from 50% of the synthetic tests to fall within 1 kyr of the true surface age. Out of all published datasets surveyed, only 6 10Be surface clast datasets meet this rigorous requirement, two of which are from California. However, only ~15 clasts are required to determine the surface age with 95% confidence. Though it is preferable to sample and date as many clasts as possible depending on the unique time and financial constraints of a study, we show that there are diminishing returns to including additional samples, and that statistically meaningfully ages can be determined with ~15 clasts.

Citation
Prush, V. B., & Oskin, M. E. (2018, 08). Dear Prudence: how many surface clasts are required to yield an accurate exposure date?. Poster Presentation at 2018 SCEC Annual Meeting.


Related Projects & Working Groups
Earthquake Geology