Mesoscale modulation of air‐sea CO₂ flux in Drake Passage

Abstract

We investigate the role of mesoscale eddies in modulating air‐sea CO2 flux and associated biogeochemical fields in Drake Passage using in situ observations and an eddy‐resolving numerical model. Both observations and model show a negative correlation between temperature and partial pressure of CO2 (pCO2) anomalies at the sea surface in austral summer, indicating that warm/cold anticyclonic/cyclonic eddies take up more/less CO2. In austral winter, in contrast, relationships are reversed: warm/cold anticyclonic/cyclonic eddies are characterized by a positive/negative pCO2 anomaly and more/less CO2 outgassing. It is argued that DIC‐driven effects on pCO2 are greater than temperature effects in austral summer, leading to a negative correlation. In austral winter, however, the reverse is true. An eddy‐centric analysis of the model solution reveals that nitrate and iron respond differently to the same vertical mixing: vertical mixing has a greater impact on iron because its normalized vertical gradient at the base of the surface mixed layer is an order of magnitude greater than that of nitrate.

Response to P.R. Gent

“…As regards the magnitude of the response on air-sea carbon fluxes, however, we find that differences between contemporary approaches to define the thickness diffusivity of Gent and McWilliams (1990) are small. Our results are apparently at odds with simulations from Swart et al (2014) who consider a more limited area. This may, or may not, be related to their choice of regional averaging or differences in their biogeochemical modules: Swart et al (2014) do not resolve iron limitation explicitly and feature a retarded formulation of light limitation (Schmittner et al, 2009).…”

Correlations of surface ocean pCO2 to satellite chlorophyll on monthly to interannual timescales

“…This may be why pCO 2 ‐nonT has a smaller seasonal amplitude in the Southern Ocean SPSS biome than in northern SPSS biomes (Figure d). As in the northern SPSS, MLDs are positively correlated with pCO 2 (Figure d), with the scattered negative correlations at 1 × 1° (Figure c) potentially attributable to mesoscale activity [ Song et al ., ]. Chlorophyll and pCO 2 are strongly negatively correlated, due to pCO 2 ‐nonT (Figure ).…”

“…The Southern Ocean is to first-order light limited and is also a high-nutrient, low-chlorophyll (HNLC) region where iron availability limits productivity [Song et al, 2016;Gervais et al, 2002;Falkowski et al, 1998]. This may be why pCO 2 -nonT has a smaller seasonal amplitude in the Southern Ocean SPSS biome than in northern SPSS biomes ( Figure 4d).…”

Simulating natural carbon sequestration in the Southern Ocean: on uncertainties associated with eddy parameterizations and iron deposition

“…Second, although we showed that GM's parameterization is, in terms of the carbon uptake in the Southern Ocean, rather robust -still -all of our coarseresolution simulations could be biased. To this end, the increase in computer power is about to provide some guidance now that recent configurations can afford to resolve much of the mesoscale (e.g., Dufour et al, 2015;Bishop et al, 2016) in the Southern Ocean, and are starting to include biogeochemical and/or modules that comprise carbon (Song et al, 2016). Hence a comparison of the sensitivity of carbon uptake to increasing winds between coarse-resolution models (like the configurations tested here) and configurations that explicitly resolve mesoscale processes, is to come.…”