In general, procedures followed those of Reuer et al. 2007. Briefly, 1 and 2-L bottles were pre-poisoned with 200 \u03bcL and 400 \u03bcL saturated HgCl2 solution, dried, and then evacuated to a pressure of < 10-3 mbar before the gas-tight valve was closed. The Louwer-Hapert 9mm valve sidearm was then filled with distilled water and capped, making sure to remove all visible bubbles, to retard any atmospheric leaks. During sampling, water from the bottom of the Niskin flask was allowed to flow into the sidearm for several seconds (to flush out the distilled water) before the vacuum valve was opened to allow water to be siphoned into the sampling bottle. After sampling\u2014each bottle typically being half full\u2014the sidearm was filled again with seawater and capped until analysis. All gas samples were analyzed within 4 months of sampling.
\n
Before gas analysis, bottles were gently agitated on a shaker table for 48h to equilibrate dissolved gases with the\u00a0
\nheadspace at room temperature (25C). Next, bottles were inverted and the seawater pumped out. Finally, the headspace gases were extracted and purified according to Li et al. (2019), Yeung et al. (2016), and Yeung et al. (2018), in which O2 is separated from Ar and N2 by gas chromatography. O2/Ar ratios are standardized relative to atmospheric O2 and air dissolved in distilled water. Bulk \u03b418O and 17\u0394 values are standardized against atmospheric O2 and O2 derived from laser-fluorination of San Carlos Olivine (Yeung et al. 2018). Clumped-isotope \u039435 and \u039436 values are standardized against laboratory-generated photolytic and heated-gas standards (Yeung et al. 2014).
\n
O2/Ar ratios were determined by integrating the peak areas for each species on an Aglient 7890B gas chromatograph utilizing thermal conductivity detection. Dissolved O2 saturation was determined from these O2/Ar measurements and checked independently through manometric determination of O2 yield and weighing the seawater from which it was extracted. Isotopologue analyses were performed on a custom Nu Instruments Perspective IS high-resolution isotope ratio mass spectrometer at Rice University.
\n\u200b
\nData was processed as was done in Yeung et al. (2018) for \u03b418O and 17\u0394 values and Yeung et al. (2016) for clumped-isotope values. Dissolved oxygen saturation is calculated using potential temperature and salinity and the O2 and Ar saturation calculated using MATLAB programs by Roberta Hamme (O2sol and Arsol from https://web.uvic.ca/~rhamme/download.html). These calculated values were compared to measured values to determine O2/O2sat.
Seawater was sampled from Niskin bottles associated with CTD casts on each cruise. Water for these dissolved gas isotope samples was the first to be sampled from a given Niskin bottle. When possible, bottles from the same cast were sampled, but depth profiles often came from separate casts at the same site.\u00a0
BCO-DMO Processing Notes:
\n
- added date and description column
\n- added conventional header with dataset name, PI name, version date
\n- modified parameter names to conform with BCO-DMO naming conventions
\n
\u00a0