Quartz purity was tested using inductively coupled plasma optical emission spectrometry (ICP-OES).

Nine in situ 14C measurements, and four in situ 14C replicate measurements were made. In table AMS_14C.csv aliquot A indicates initial in situ 14C concentrations measured in each sample and aliquot B indicates repeat in situ 14C measurements on the same sample.

A long-term average 14C blank used to correct the initial measurements of in situ 14C (n = 9) is 4.53 +/- 0.24 x 10^4 atoms g-1. For replicate measurements (n = 4), a higher short-term average 14C blank of 7.14 +/- 0.30 x 10^4 14C atoms g-1 was used. A 6 % (1 sigma) uncertainty was assigned to each in 14C measurement concentration reported by AMS before calculating exposure ages. The 6 % uncertainty exceeds the reported analytical uncertainty for all in situ 14C measurements made for this study and reflects the repeatability of measurements of CRONUS-A extracted at Tulane. The in situ 14C concentration of CRONUS-A extracted at Tulane is reported with a long-term value of 6.12 +/- 0.32 x 10^5 (n = 10) (Goehring et al., 2019). The distribution of the long-term 14C measurement background of samples from the Tulane University extraction facility is long-tailed and non-time dependent (Balco et al., 2022).

Blank correction using a long tailed distribution is, therefore, appropriate for low concentration samples such as subglacial bedrock cores (Balco et al., 2023) where blank variability constitutes the dominant source of uncertainty in the in situ 14C measurements (Balco et al., 2022). Blank correction procedures for low concentration samples are discussed at length in Balco et al., (2022) and can be viewed directly via the following link: https://doi.org/10.5194/tc-2022-172-AC2