| Summary: | We compared field measurements of the biological O 2 saturation anomalies, ΔO 2 /Ar and ΔO 2 /N 2 , from simultaneous oceanographic deployments of a membrane inlet mass spectrometer and optode/gas tension device (GTD). Data from the Subarctic Northeast Pacific and Canadian Arctic Ocean were used to evaluate ΔO 2 /N 2 as an alternative to ΔO 2 /Ar for estimates of mixed layer net community production (NCP). We observed strong spatial coherence between ΔO 2 /Ar and ΔO 2 /N 2 , with small offsets resulting from differences in the solubility properties of Ar and N 2 and their sensitivity to vertical mixing fluxes. Larger offsets between the two tracers were observed across hydrographic fronts and under elevated sea states, resulting from the differential time-response of the optode and GTD, and from bubble dissolution in the ship’s seawater lines. We used a simple numerical framework to correct for physical sources of divergence between N 2 and Ar, deriving the tracer ΔO 2 /N 2 ′. Over most of our survey regions, ΔO 2 /N 2 ′ provided a better analog for ΔO 2 /Ar, and thus more accurate NCP estimates than ΔO 2 /N 2 . However, in coastal Arctic waters, ΔO 2 /N 2 and ΔO 2 /N 2 ′ performed equally well as NCP tracers. On average, mixed layer NCP estimated from ΔO 2 /Ar and ΔO 2 /N 2 ′ agreed to within ∼2 mmol O 2 m –2 d –1 , with offsets typically smaller than other errors in NCP calculations. Our results demonstrate a significant potential to derive NCP from underway O 2 /N 2 measurements across various oceanic regions. Optode/GTD systems could replace mass spectrometers for autonomous NCP derivation under many oceanographic conditions, thereby presenting opportunities to significantly expand global NCP coverage from various underway platforms.
|
|---|