Order-Statistical Corrections for Inbuilt Age Effects at Paleoseismic Sites

Glenn P. Biasi, Katherine M. Scharer, & Devin McPhillips

Submitted August 30, 2026, SCEC Contribution #15315, 2026 SCEC Annual Meeting Poster #127

We investigate inbuilt age effects in charcoal used to constrain paleoseismic earthquake dates. On the San Andreas fault and most California faults, most estimated earthquake dates are derived from radiocarbon ages from charcoal fragments in alluvial deposits. The inbuilt age (IA) represents the uncertain amount of time from carbon fixing in the plant to its death and entrance to the geomorphic system. Samples may derive from recent growth or an old tree. An established method for estimating IA that combines sampling and radiocarbon uncertainties leads to an IA distribution for Transverse Range sites that is over 400-years wide (Scharer et al., EPSL, 2024). We propose an alternative, strictly positive IA distribution that, with order-statistical methods, significantly improves layer dates in the paleoseismic context.

The IA distribution reflects the age spectrum of materials in the watershed. In the Transverse Ranges, watersheds are relatively small and transport times can be neglected. Scharer et al. (2024) measured the IA distribution on fires from 2013 and 2020 with precision using F14C methods. Half of the materials were younger than 25 years and only 20% are likely to have come from materials older than 80 years. We approximate these data by an exponential probability distribution. In individual layers, samples can be viewed as being drawn at random from the IA distribution determined from recent fires, shifted to the age of the layer. Random samples might be near the date of the fire, in the middle of the IA, or in the tail. This spreads radiocarbon dates in age even though they are from the same paleoseismic deposit. If dates from a layer are combined, for example by using Oxcal Phase function, dating precision suffers. Alternatively, if one keeps just the youngest sample(s), radiocarbon dates may be discarded. Order-statistic distributions can be used when charcoal can be considered a sample from the IA. Specifically, they predict individual age adjustments that would align the ordered sample onto the age of the fire. Applied at the Pallett Creek paleoseismic site, this approach clearly identifies past fire deposits and sharpens many layer age estimates. Date adjustments predicted by the IA also give grounds to consider explanations for date discordance, where reworking of older charcoal or unrecognized geologic depositional conditions may not have been recognized.

Key Words
paleoseismology, radiocarbon, Pallet Creek

Citation
Biasi, G. P., Scharer, K. M., & McPhillips, D. (2026, 08). Order-Statistical Corrections for Inbuilt Age Effects at Paleoseismic Sites. Poster Presentation at 2026 SCEC Annual Meeting.

Related Projects & Working Groups
Earthquake Geology