Data Processing:
The height of the Co reduction peak is linearly proportional to the amount of total dCo present in the sample. Peak heights were determined by NOVA 1.10 software. A standard curve was created with 4 additions of 25 pM dCo to each sample, and a type-I linear regression of the addition standard curve performed by the LINEST function in Microsoft Excel allowed for the calculation of the initial amount of Co present in the sample.

Analytical blank measurements for each reagent batch (a unique combination of DMG, EPPS, and NaNO2 reagent batches) were measured to determine any Co contamination due to reagent impurities. Blanks were prepared in triplicate with UV-irradiated surface seawater passed through a column with Chelex 100 resin beads (Bio-Rad) to remove metal contaminants, then UV-irradiated again. Chelex beads were prepared as described in Price et al. (2013) to remove organic impurities from leaching into the eluent. For the 5 batches of reagents used on this expedition, the analytical blanks were found to be 2.3 pM, 4.0 pM, 10.1 pM, 15.6 pM, and 8.6 pM dCo, with an average of 8.1 pM Co. The analytical blank detected for the laboratory-run total dCo samples was 1.0 pM. It should be noted that blank values above 10 pM are considered high for this method. Analytical blank values were subtracted from the measured Co values determined with the respective reagent batch. The average standard deviation within each triplicate batch of blanks (1.3 pM) was used to estimate the analytical limit of detection (3 * blank standard deviation) of 4 pM. When detectable dCo concentrations were found below the 4 pM detection limit, their values were preserved in the dataset and flagged as below the detection limit (<DL).

BCO-DMO Processing:
- removed "NaN" as a missing data indicator (represented as blank/no value in the .csv data file);
- replaced "n.d." with "not detected";
- renamed fields to comply with BCO-DMO naming conventions;
- converted date/time field to ISO 8601 format.