Data Processing:
During the isotope ratio analysis, the data were collected in 20 cycles. The isotope ratios were edited for outliers, and averaged with internal standard errors estimated from the multiple cycles. This should be understood as the lower limit to the true error: the reported value cannot be any better than the internal reproducibility of the isotope run. Standard Ocean Data View flags were used (reference all flags at https://www.bodc.ac.uk/data/codes_and_formats/odv_format):

1: Good Value: Good quality data value that is not part of any identified malfunction and has been verified as consistent with real phenomena during the quality control process.

2: Probably Good Value: Data value that is probably consistent with real phenomena but this is unconfirmed or data value forming part of a malfunction that is considered too small to affect the overall quality of the data object of which it is a part. [Used when only one replicate confirmed the reported value.]

3: Probably Bad Value: Data value recognized as unusual during quality control that forms part of a feature that is probably inconsistent with real phenomena.

4: Bad Value: An obviously erroneous data value.

Blank Values and Detection Limits:
Two different methods were used to determined Pb concentrations due to very low Pb concentrations. See Boyle et al. (2020) for more detail. Pb concentrations are already intercalibrated and were included in IDP2017.

Average blanks for the 1.5 mL method: 4.3 – 7.0 (± 1.9 SD) pmol/kg (based on 10 analysis days and 12 blanks per day)

Average blanks for 30 mL method: 0.2 - 3.9 (± 0.9 SD) pmol/kg (based on 20 analysis days and 12 blanks per day)

Blanks: There are three main sources of blanks in this analysis. The major contribution comes from the Nobias Chelate PA1 resin, whereas a minor contribution comes from the Eichrom AG-1x8 resin. A significant but minor contribution also comes from the ICPMS hardware from Pb retained from previous analyses. The latter blank is determined by an "On Peak Zero" measurement on high purity 0.2M HNO₃. We cannot measure the PA1 blank directly because the organic matter released by the resin interferes with the isotopic analysis. So we measure one blank value for PA1 + AG1-x8, and then another one for AG1-x8 alone, determining the PA1 blank by difference.

Because the AG1-x8 blank is so low, we cannot make an accurate measurement of its Pb isotope ratios. So we measure the most intense signal (208Pb) and then estimate the blanks for the other isotopes assuming an isotope composition for "typical U.S. Pb" assumed from the ratio of our U.S.- purchased laboratory Pb standard. Because the PA1 blank is higher, we averaged the isotopic compositions observed from several high-blank runs and assumed these applied to all runs using the same batch of PA1.

All of the signals from our large-volume samples were well above Pb concentration detection limits, but the precision of the measurement degrades relative to the Faraday cup Johnson noise (and Daly ion counter statistics) as the signal becomes weaker. We think that our isotope ratio measurements for 206Pb/207Pb are good to the second decimal place for samples with 1-2 pmol/kg Pb.

Internal Consistency:
Seawater Pb concentration samples were analyzed in triplicate on each analytical day. If at least two of these don’t agree within expectations, they are re-run again in triplicate on another analytical day. Internal lab seawater Pb concentration standards were run every analytical day to ensure long-term consistency.

BCO-DMO Processing:
- replaced '#N/A' with 'nd' as missing data value;
- renamed fields;
- added date/time fields in ISO8601 format.