the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 26 Feb 2026
Impacts of Simulated Coastal Ocean Alkalinity Enhancement on the Seasonal Cycle of the Air-Sea CO2 Flux and Surface Ocean pCO2 in European Waters under a Low- and a High-Emission Scenario
Abstract. One potentially scalable method to remove CO2 from the air is ocean alkalinity enhancement (OAE), which works to lower surface ocean partial pressure (pCO2) and accelerate CO2 sequestration and durable storage. This study explores how OAE might affect the seasonal carbon cycle, which plays a key role in the ocean’s annual CO2 uptake. By analysing earth system model simulations of OAE implemented continuously at the European coastline until 2100 under low and high climate forcing, it was found that: when carbon cycle seasonality is temperature-driven, a) OAE enhances CO2 uptake in winter, when it is naturally strongest, and it reduces ocean pCO2 in summer, when it is naturally highest; b) higher CO2 emissions increase the sensitivity of the seasonal carbon cycle; c) a region with a shallow bathymetry and well-mixed waters may be ideal for implementing OAE due to fast air-sea CO2 equilibration.
- Preprint
(1468 KB) - Metadata XML
- BibTeX
- EndNote
Status: final response (author comments only)
- RC1: 'Comment on egusphere-2026-926', Anonymous Referee #1, 24 Mar 2026
-
RC2: 'Comment on egusphere-2026-926', Anonymous Referee #2, 27 Mar 2026
Summary
The manuscript presents a coastal OAE study in which ALK is added to European coastal grid cells of a medium resolution model. The model is simulated for two emissions scenarios, and analysis focuses on the seasonal effects on the carbonate system. The main results include that OAE enhances uptake in the winter over the summer and that the North Sea represents a good region for future OAE deployment.
Assessment
Overall, I found the manuscript generally well-written but with a number of omissions that reduce its value to the literature. My major comments are as follows:
- The manuscript focuses on a regional sea but does not validate the performance of the model at this scale. This is particularly important in the case of the coastline release areas, as the manuscript essentially takes the results from these areas at face value, despite potential issues around poor representation due to resolution, missing physical / biogeochemical processes or riverine inputs.
- Presentation of the manuscript’s results is slender with a focus purely on the 2090s with nothing about how the system then came into being, arguably absent spatial plots on changes in seasonality (e.g. maps of seasonal amplitudes), and confusing in-line use of important metrics that would be much more useful if organised into a table.
- Although the work is performed globally, the analysis is strictly a regional affair, and it omits consideration of factors such as OAE performance (e.g. extra mol DIC per added eq ALK), where this happens, loss (to the region) of added ALK and extra DIC, and impacts on the model’s natural ecosystem (NPP, CaCO3 production, etc.). These affect the assessment of requirements for MRV and the cost-benefit ratio.
- Main conclusions, such as the suitability of the North Sea for OAE, fail to note the dependence of this on the specific OAE approach here (i.e. addition of dissolved ALK), while the manuscript’s caveats overlook model limitations in shelf seas areas, and potentially oversell this region.
My assessment is Major Revisions to address these gaps so that readers can both better understand this study and better assess its findings in the context of the wider OAE literature.
Specific comments
Ln. 29-30: tweak to “… to lower its sea surface partial pressure” and “durable ocean storage” for clarity
Ln. 34-35: “OAE enhances CO2 uptake in winter … reduces ocean pCO2 in summer” is potentially confusing as lowering pCO2 in the summer might also be read as enhancing the conditions for CO2 uptake
Ln. 36-37: grammar – switch “a region with a shallow bathymetry” to “regions with shallow bathymetry”; also, perhaps change “fast” to “faster” on Ln. 38
Ln. 130: correct the Chien et al. (2022) citation to the published paper and not the Discussions manuscript:
Chien, C.-T., Durgadoo, J. V., Ehlert, D., Frenger, I., Keller, D. P., Koeve, W., Kriest, I., Landolfi, A., Patara, L., Wahl, S., and Oschlies, A.: FOCI-MOPS v1 – integration of marine biogeochemistry within the Flexible Ocean and Climate Infrastructure version 1 (FOCI 1) Earth system model, Geosci. Model Dev., 15, 5987–6024, https://doi.org/10.5194/gmd-15-5987-2022, 2022.
Ln. 130: see my comment below too, but the validation here is at a global scale while your analysis is at the regional scale; so it does not especially help with ground-truthing your North Sea study
Ln. 141: O’Neill et al. (2016) is the protocol for the ScenarioMIP simulations, not for general guidance on spin-up or Historical simulations; you should also add Eyring et al. (2016)
Ln. 147: is the addition region really 50km or is it the width of a single grid coastal cell?; and if the latter, you may find that the width of the addition area changes with latitude because of grid distortion; I assume that, where the amounts of alkalinity added are quantified later that proper account is taken of varying grid cell areas
Figure 1: Could this plot be improved to either show the full land mask (it’s currently cut off on the east and west margins by unexplained curves) or by cropping the domain shown to remove these missing portions; more generally, you’re showing a lot of domain that aren’t part of your later analysis; you might also want to note in the caption that the x and y axes are not longitude and latitude – instead this plot seems to use the model’s i-j grid
Figure 1: thank you for clearly including the British Isles as part of European waters – it helps us Remainers keep the flames of hope alive! ;-)
Ln. 152-157: you don’t say at what depth the alkalinity is added; and, while you make what you do clear, you might want to note how this compares to other studies – either here or in the discussion; it sounds a little like CDRMIP and different from the likes of Palmieri & Yool’s 2024 study where ALK was added on the seafloor
Ln. 157: as you model alkalinity prognostically, you might want to express the quantity of alkalinity added in equivalents as well as carbon units
Ln. 165: the explanation for slicing to a European region feels a bit vague; you will need to be clear later on what the effect of cropping to this region is; in the context of the cited study, Palmieri & Yool (2024), you might be adding ALK in this region but the DIC is only absorbed outside this region
Ln. 174: you say that “ocean pCO 2 and the CO 2 flux are written out only for the surface”, but they’re actually only CALCULATED for the surface too. They’re air-sea interface terms not 3D ocean terms.
Ln. 175: “alkalinity is calculated” isn’t quite right – it’s more *represented*
Ln. 182: some validation of the model seems in order; as already noted above, the Chien et al. (2022) paper deals with large scale validation and is not focused in the region here; and certainly not focused in the thin coastal strip that you use for parts of your analysis; given the importance of temperature SST seems a good start; ditto pCO2 (if observational resolution permits it); I’m not sure how good DIC and ALK datasets are in this specific region, but if they’re available, definitely validate against these
Ln. 201: you note a North Sea section here – any chance it could be plotted on Figure 1?
Figure 2: consider using a diverging palette (e.g. classic blue-white-red; or the one used in Figure 4) for the CO2 flux so that zero flux is easier to spot
Figure 2: I like that you kept the graininess of the model grid here – much more honest than the interpolated contour plots people use
Figure 2: why show surface NPP when, due to mixing, it’s surface layer productivity that controls CO2 uptake?
Figure 2: the ordering of the subplots seems a little arbitrary; by default, I would have gone SST, MLD, NPP, CO2 flux to reflect some sort of process-ordering, but in reading your results, I might have at least put SST and CO2 flux closer together so it’s easier to see the common patterns
Ln. 217: a general comment I’d make about this section is that there are lots of numbers given in the text and it would be better to consolidate these into a table so that readers can clearly see the major features identified both geographically and between scenario runs
Ln. 227: vertical processes are invoked here but no results are presented in support of these; perhaps a comparison of Hovmollers of decadally-averaged ALK from the two scenarios (or differences of scenarios from their control twins) would assist?
Figure 3: is the “coastline average” the values in the grid cells indicated in Figure 1?
Figure 3: might it be an idea to have the upper and lower plots sharing the same vertical axes so it’s easy to compare them?; it looks like it should generally work
Figure 3: looking at this figure’s results from the 2090s makes me wonder if there’s any value in exploring the same (or similar) quantities across the whole of the simulation period?; perhaps amplitude of seasonality would be good to see given the two scenarios clearly diverge on properties like this
Figure 4: one can imagine that the calculation of CO2 flux on a much coarser grid that’s implied here might affect the results; it might be helpful to articulate how exactly the CO2 flux calculation is done – i.e. are ocean properties (SST, DIC) averaged to the atmospheric grid and then pCO2, CO2 flux is calculated, or is ocean pCO2 calculated on the ocean grid, averaged to the atmospheric grid and then CO2 flux calculated, or is everything done on the ocean grid, and the CO2 flux is integrated to the lower-resolution atmospheric grid?; most models I know do the latter; in any case, and explanation – probably in the methods – is needed to explain what’s done and what the implications of this are
Figure 4: is it informative to plot the point-by-point correlation between delta-ALK and delta-pCO2?; this might help clarify the text around different regions responding differently; and use of different coloured points for the two scenarios could allow them to be plotted on the same axes, and this might pull-out the divergence in the two runs seen in the region of the North Sea east of Scotland and west of Denmark
Ln. 293-303: per my earlier remarks, consolidating the figures here in a table that perhaps separated the numbers between regions as well as between the scenarios might help the manuscript articulate the important differences better; text descriptions are less clear than even a simple table
Ln. 339: “becomes more sensitive” – I’m unclear on what becoming more sensitive means for carbonate chemistry; perhaps illustrating with a clear example (“for instance, …”) would help
Ln. 352: “stronger chemical leverage” – again, I’m not sure what exactly is meant by this
Ln. 353: “exacerbated by OAE implementation” makes it sound like the specifics of how OAE is implemented are important when you might just mean the use of OAE in general
Ln. 352-353: this concluding sentence is similarly difficult to follow without a concrete example that would help readers follow; do, for instance, you mean that
Ln. 360-361: same point I made about the abstract; i.e. this sentence, which reads like it’s making a nice compare-and-contrast statement, is kind-of talking about apples and oranges by flipping between CO2 flux for the winter but pCO2 for the summer; looking at Figure 3, net flux is lowest in the summer when air-sea delta pCO2 is actually the greatest; which points to perhaps other factors (winds and gas transfer velocity, mixing depth) as being important (but less acknowledged here)
Ln. 362: “sensitivity” – would this perhaps be better as “size of the response” or similar?; it’s not clear to me that readers will understand what “sensitivity” means here
Ln. 368: given the work focuses on a regional experiment in a major shelf sea, this list of limitations omits reference to the model’s resolution and representation of both shelf seas physics and biogeochemistry; there’s a mention of riverine inputs of ALK, which is good, but I thought MOPS was more an open ocean model than a shallow shelf seas model (I say this because the model I use is somewhat similar)
Ln. 364: the statement about locations like the southern North Sea may be dependent upon the mode of ALK addition used for OAE; where dissolution of particulate minerals is used to add the ALK, temperature effects may favour warmer equatorial waters; this is specifically examined in Palmieri & Yool (2024) where the effectiveness of different shelf seas at the global scale is evaluated
Ln. 368: a style point, but I’d be inclined to move this whole paragraph into the Discussion rather than Conclusions; keep the Conclusions to your main findings (and maybe bullet-point them – but that really is a style point)
Ln. 384: on the subject of style points, maybe trim people to initials only in this paragraph
Citation: https://doi.org/10.5194/egusphere-2026-926-RC2
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 149 | 73 | 17 | 239 | 13 | 28 |
- HTML: 149
- PDF: 73
- XML: 17
- Total: 239
- BibTeX: 13
- EndNote: 28
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
Please see the attached file.