the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 30 Jan 2026
A case for a pragmatic oxygen-based approach to quantifying the biological contribution to the marine carbon sink
Abstract. AOU, the ‘apparent oxygen utilization’, is a widely used concept in ocean biogeochemistry to assess and interpret ocean deoxygenation and changes of the marine carbon cycle. It provides an estimate of oxygen (O2) used and dissolved inorganic carbon (DIC) released during organic matter degradation in the ocean interior. AOU is calculated from observations of temperature, salinity and O2. This calculation relies on the assumption that surface water O2 is in equilibrium with the atmosphere when water masses form and sink into the interior ocean. Using specifically designed idealized model tracers, we here provide an evaluation of the reliability and uncertainty of this approach. Global AOU reliably estimates total oxygen debt, defined as the sum of ‘true’ oxygen utilization (TOU) since last contact with the atmosphere and biotic oxygen disequilibrium (O2dis,bio). AOU and TOU + O2dis,bio agree to within about 10 %, both for the preindustrial state estimate and the transient climate change situations explored. Taking differences of the biotic components of the O2 and DIC disequilibrium (DICdis,bio) into account, we find that, for the pre-industrial state estimate, the carbon equivalent of AOU is an about 15 % underestimate of the sum of DICremin and DICdis,bio. For the climate change transient we find that š„AOU underestimates our estimates of the disequilibrium corrected change in Csoft (DICremin + DICdis,bio,COU*-BGC*) by 25 % . In summary, we suggest that AOU in particular can be used to assess causes of ocean deoxygenation, and is a very useful proxy for changes of storage of biologically processed carbon under various climate change scenarios.
- Preprint
(1845 KB) - Metadata XML
-
Supplement
(3049 KB) - BibTeX
- EndNote
Status: final response (author comments only)
-
RC1: 'Comment on egusphere-2026-304', Benoit Pasquier, 01 Apr 2026
- AC1: 'Reply on RC1', Wolfgang Koeve, 22 May 2026
-
RC2: 'Comment on egusphere-2026-304', Anonymous Referee #2, 04 May 2026
Ā Ā The authors revisit the concept of āApparent Oxygen Utilization (AOU)ā in this manuscript. They assess the usefulness of the AOU approximation for the preindustrial steady state and for decadal-to-centennial climate change projections using the UVic Earth System Model (ESM). Moreover, they have extensively evaluated the usefulness of AOU as an estimate of the soft-tissue carbon pump.
Ā Ā It has been widely discussed that uncertainties in AOU arise from air-sea disequilibrium and from the interpretation of AOU relative to āTrue Oxygen Utilization (TOU)ā. The authors have designed a suite of sensitivity experiments to quantify uncertainties under various boundary conditions (i.e., scenarios). The AOU (and its extension to quantify the biological carbon pump) has been widely used for observational and model diagnostics (e.g., deoxygenation studies), and it is important to assess its uncertainties and interpretation. The authors introduced (revisited) the concept with a nice introduction, methods, and discussion based on sensitivity experiments. I think it is important to discuss the widely used AOU concept to better understand the ocean oxygen and carbon cycles. I enjoyed reading the paper, and the study will be a great contribution to advancing our understanding of ocean deoxygenation and carbon uptake under climate change. I still have several comments for the paper, and I hope this helps to improve the manuscript.
Major Comments
1. I see this study is focusing on further (or re-) interpreting the concept of āAOUā. I believe, in general, we think of AOU as an approximation of TOU (under the assumption of perfect equilibration at the surface with given T and S, as the authors also stated), but the authors introduced further interpretation, including ādisequilibriumā (more specifically, biologically driven disequilibrium) to interpret AOU from a different point of view.
Ā Ā I understand what the authors are trying to do in this paper, but this needs clarification in the main message (especially in the abstract and conclusion). Moreover, I am wondering whether the authors are arguing that AOU and TOU provide slightly different information (by including the impact of air-sea disequilibrium stemming from both physical and biological processes). I think there is room to revisit the presentation style to highlight this rather than just introducing the interpretation of the AOU and TOU. This relates to the minor/specific comment I made for L247, which I think needs clarification on the difference between AOU and TOU from the author's perspective (and whether the authors are arguing both useful concepts, but interpretation needs to be made with caution). Particularly, the meaning and difference between āā¦ accumulated O2 utilization of interior ocean waters since last contact with the atmosphere (for TOU)ā and ā the total accumulated O2 debt associated with the degradation of organic matter in the interior ocean (which I think points to the AOU)ā in the main text were not entirely clear to me (I usually think the former is the basic interpretation of both AOU and TOU).
2. It is interesting to see how AOU bases estimates of soft tissue carbon pump diverge from the DIC-based estimates. In climate change conditions, the authors explain this by the impact of the Canth-effect (L374-ā¦), but I am still not sure why the transient increase in atmospheric pCO2 impacts DICdis, bio. I understand the physical-chemical invasion of anthropogenic carbon could blind things, but I would like to ask for further explanation and clarification on why this impacts the DICdis, bio, not the DICdic, phys.
Ā Ā In addition, when authors conducted no biological pump experiments, did you also turn off the carbonate pump (calcium carbonate formation)? (I see the statement turning off the impact of biological carbon pumps on alkalinity and DIC is disabled for noBioPumps). I understand the AOU (Csoft) mainly focuses on soft-tissue pumping, so it is a separate topic, but I am wondering how this could affect the carbon in your suite of sensitivity experiments, especially the carbon disequilibrium, which may impact the results.Ā
3.Ā Finally, it would be helpful to see statements on how we can use the concept of AOU (and TOU) for future model diagnostics, such as CMIP model diagnostics, given the diverse treatment of ocean biogeochemical components in the model and differences in circulation and mixing. This is somewhat general, but it will still be nice to see the author's opinion and comments on facilitating the future analysis for better understanding of the ocean deoxygenation and changes in the carbon cycle.
Minor/Specific Comments
How could the difference in air-sea gas exchange and equilibrium timescales between O2 and CO2 affect the results, given that the authors highlight the disequilibrium argument in this study?
L42: āWilson et al., 2022ā font should be fixed (from italic to non-italic).
L57: I think this is a good point; the assumption is not only from the perfect surface O2 saturation but also stems from mixing and subsurface warming. Could you also clarify how subsurface warming could occur (are you thinking of shortwave radiation penetrating beyond the surface)?
L122: How does wind speed, sea state, sea ice cover, etc., affect the biological part of disequilibrium? (O2dis, bio)
L198: How much of a difference does it make to run CO2-emission-based simulations (compared to the prescribed CO2 simulations as in Arora et al., 2020)
L205: āor the rate with which the sinking speed increases over depthā, this should be clarified more by adding sinking āorganic particleā speed.
L222:Ā It is about the style, the sentence "Globally, most of this difference, about 80%, is explained - rather than by the, primarily, thermal effect of solubility increase due to cooling - by upwelling of undersaturated waters that carry the O2 debt stemming from organic matter." is not easy to follow. I suggest rephrasing it. (I think the authors are arguing that most of the difference (about 80%) between AOU and TOU could be explained by upwelling of undersaturated waters ... correct?)
L247: ā(TOU) ā¦ it is in fact not a good measure of the total accumulated O2 debt associated with the degradation of organic matter in the interior ocean.ā
This is an interesting point (and somewhat a bit confusing point to me as mentioned in the major comments), I usually think of AOU as an approximation of TOU, which means AOU also represents the accumulated O2 utilization of interior ocean waters since last contact with the atmosphere (i.e. depends both on water mass pathways, age of water mass, and degradation of organic matter along the pathways), but the authors claim that AOU provides a good estimate of ātotal biological effect TOU+O2 dis,bioā. Do authors think we should treat the AOU and TOU differently for interpreting the O2 changes (rather than approximation of TOU)?
The authors also argue that under various circumstances (tested through a suite of sensitivity experiments), the uncertainties between AOU and āTOU + O2 dis,bioā are ~ 10%, so I believe it is important to clearly state the reinterpretation of AOU in the manuscript (again, as mentioned in major comments).
L258: āā¦ in year 100ā
Is this true? Looking at Fig. 2, the magnitude of ĪAOU and ĪTOU differs significantly by year 800 (we do see differences by year 100, but it is hard to see a 30% difference from the graph, perhaps because of the scale). It will be nice to include the actual numbers (magnitudes).
L298, 303-304: Minor point, but āĪAOU is robustly able to estimate ĪTOU+š„O2 dis,bio to within about 10% uncertaintyā, (under various model setups). Good to see low uncertainties across various scenarios, but where do you think this 10% difference stems from (numerics in the model, or other factors)?
L298: āĪAOU āā font should be fixed.
Fig. 3: I suggest including a legend explaining markers (circle, cross, etc.).
Citation: https://doi.org/10.5194/egusphere-2026-304-RC2 - AC2: 'Reply on RC2', Wolfgang Koeve, 22 May 2026
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 963 | 365 | 113 | 1,441 | 221 | 94 | 192 |
- HTML: 963
- PDF: 365
- XML: 113
- Total: 1,441
- Supplement: 221
- BibTeX: 94
- EndNote: 192
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
See attachment.