Baselines Matter: Mass Spectrometric Assessments of Biological O2 Supersaturation (ΔO2:Ar) Benefit from Two-Point Calibrations

TL;DR

This paper shows that ignoring baseline measurements in MIMS can lead to significant errors in estimating ocean productivity.

Contribution

The study demonstrates the critical impact of two-point calibrations on accurate ΔO2:Ar measurements in MIMS.

Findings

Higher gas loads increase signal heights but also baseline levels, reducing instrument sensitivity.

Disregarding baselines can lead to underestimations of ΔO2:Ar values by factors of 1.4 to 4.

Two-point calibrations are essential for accurate net community productivity analysis.

Abstract

Membrane-inlet mass
spectrometry (MIMS) based assessments of ΔO2:Ar to
estimate marine productivity are becoming a widely
used tool in biogeochemistry. Especially continuous ship-borne surveys
of dissolved gases allow for high spatial and temporal resolution
in the analysis of surface ocean net community productivity. Depending
on instrument configuration and architecture, however, measurements
may be afflicted with substantial detection baselines for each analyzed
gas. We hypothesized that ignoring these baselines (as seems to be
common practice) can considerably affect the outcomes of the measurements.
Using MIMS data from two cruises, we assessed the impact of calibration
procedures and different pressure regimes, i.e., gas loads, on the
ΔO2:Ar ratios and analyses of net community productivity.
We compared conventional ratio-based one-point calibration approaches
with two-point calibration approaches that include baselines and either
calibrate for the equilibrium ratio of O2:Ar, or calibrate
for the individual gas concentrations. Our data show that higher gas
loads increase apparent signal heights, but also disproportionately
increase the individual gas-associated baseline levels, thereby reducing
instrument sensitivity while increasing a potential bias if data is
used without correction. Depending on the adjusted gas loads, severe
underestimations of final ΔO2:Ar values by factors
of ca. 1.4 and 4 were calculated when disregarding baselines, emphasizing
the critical importance of baseline determinations in MIMS-based methods.

Figures