the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 28 Oct 2024
Persistent deep-water formation in the Nordic Seas during Marine Isotope Stages 5 and 4 notwithstanding changes in Atlantic overturning
Abstract. Reductions in the extent and formation of North Atlantic Deep Water (NADW) and the expansion of southern-sourced waters in the Atlantic Ocean were linked to enhanced marine carbon storage during glacial and stadial periods and are considered a key mechanism explaining late Pleistocene atmospheric CO2 variations on glacial-interglacial and millennial timescales. However, changes in the formation of deep waters in the Nordic Seas, an important source of NADW, and their influence on the geometry and intensity of Atlantic overturning remain poorly understood, especially beyond the last glacial maximum, leaving possible impacts on atmospheric CO2 changes elusive. Here, we present high-resolution Cibicidoides wuellerstorfi B/Ca-based bottom water [CO32-] reconstructions, alongside with complementary C. wuellerstorfi stable oxygen and carbon isotopes and abundance estimates of aragonitic pteropods in marine sediment core PS1243 from the deep Norwegian Sea to investigate past deep-water dynamics in the Nordic Seas and potential impacts on Atlantic overturning and carbon cycling. Our data suggest continuous formation of dense and well-ventilated (high-[CO32-]) deep waters throughout Marine Isotope Stages (MIS) 5 and 4, alongside a deepening of the aragonite compensation depth by at least 700 m during the MIS 5b-to-4 transition, consistent with sustained Nordic Seas convection. In addition, higher-than-Holocene bottom water [CO32-] during MIS 5e highlight the resilience of Nordic Seas overturning towards a warmer North Atlantic, decreased Arctic sea ice extent and meltwater supply from surrounding ice sheets. A compilation of bottom water [CO32-] records from the Atlantic Ocean indicates that dense waters from the Nordic Seas may have continuously expanded into the intermediate and/or deep (western) North Atlantic via supply of dense water overflows across the Greenland-Scotland Ridge, diminishing the capacity of the North Atlantic to store carbon during MIS 4 and stadial conditions of MIS 5. Our study emphasises differences in the sensitivity of North Atlantic and Nordic Seas overturning dynamics to climate boundary conditions of the last glacial cycle that have implications for the carbon storage capacity of the Atlantic Ocean and its role in atmospheric CO2 variations.
- Preprint
(3784 KB) - Metadata XML
- BibTeX
- EndNote
Status: open (until 23 Dec 2024)
-
RC1: 'Comment on egusphere-2024-3163', Anonymous Referee #1, 18 Nov 2024
reply
Stobbe and co-authors present a novel carbonate ion record from the deep Norwegian Sea, spanning the last 130 kyr and encompassing major climate transitions. The record is interrupted during parts of MIS3 and MIS2 and the authors hence focus on interpreting the last glacial inception as well as millennial-scale climate events during MIS5. The carbonate ion record is primarily interpreted as an indicator of past deep convection in the Nordic Seas, with comparisons drawn to other Atlantic records to assess the export of newly formed deep water into the Atlantic basin. However, the latter analysis is somewhat limited due to the absence of carbonate ion records in key regions, preventing a comprehensive assessment of Nordic Seas deep water expansion during the last interglacial period and glacial inception.
The authors provide a commendable and transparent assessment of the inferences drawn from the new record, clearly acknowledging areas of uncertainty and where additional data is required. The manuscript is well-written and illustrated, although some figures may benefit from simplification to enhance clarity, as they are too crowded for my taste.
The authors interpret the new data as indicative of sustained deep water formation not only during the last interglacial period but extending well into the glacial inception, a conclusion that appears to be supported by the data. However, their interpretation of millennial-scale variations in the Norwegian Sea record and attempts to correlate these with Heinrich Stadials are less convincing. Given the (multi-)millennial-scale resolution of the record and significant internal variability, this particular interpretation appears to lack robust support from the data.
I therefore would like to see specifically this issue addressed by the authors before I can recommend publication of the study. Please find more detailed comments of these issues below.
Major comment:
The authors posit that the new PS1243 record resolves millennial-scale events, with particular emphasis on Heinrich Stadials. However, when compared to the reference record MD95-2039 from the Iberian Margin, PS1243 appears to lack the requisite temporal resolution and coherency for reliable interpretation of variability during Heinrich Stadials. Moreover, the carbonate ion record of PS1243 exhibits substantially higher internal noise than MD95-2039, further complicating the identification and interpretation of smaller-scale changes associated with these events. The low sedimentation rate of 1-4 cm/kyr of PS1243 presents a significant constraint on temporal resolution. This limitation is particularly evident during MIS5c to 4, where the temporal sampling frequency appears to be less than one sample/kyr. Such low resolution is consistent with the observed sedimentation rate but raises concerns about the validity of interpreting millennial-scale events. The interpretation of millennial-scale events is a recurring theme throughout the manuscript, featuring in the abstract and receiving detailed treatment in section 5.5. However, given the multi-millennial resolution of the new record, I strongly recommended that the authors exercise caution in interpreting individual data points that coincide with the discussed events. This is particularly crucial considering that the age model may not provide sufficient precision to confidently associate these data points with specific events. In light of these considerations, I advise to reassess the claims regarding millennial-scale event resolution. The authors should consider either refraining from such interpretations or significantly qualifying their assertions, acknowledging the limitations imposed by the record's temporal resolution.
Minor comments:
L18: Atmospheric CO2 was increasing by up to 15 ppm during Heinrich Stadials suggesting less not more marine carbon storage as noted here by the authors (even though the terrestrial carbon storage may have also played a role). In general, the community shifted away from attributing all carbon storage changes to water mass changes in the Atlantic, now having a stronger focus on marine carbon storage of the SO and Pacific. Maybe, this can also be reflected here in the Abstract.
L57: Please also cite newer studies, including modelling efforts constrained by proxy data (e.g., Muglia and Schmittner 2021, Poppelmeier et al., 2023).
L65: Due to the rapidity of the anthropogenic change there are no real analogues in the past. Maybe this statement should be hence adjusted accordingly.
L71: Please briefly mention the role insolation played in the different conditions of MIS5e vs the Holocene, which explains most of the differences.
L229: Please better justify your estimated age uncertainty.
L231: The additional age constraints of d18O tied to NGRIP is a promising approach, but might also be prone to errors. For instance, there are other instances where the NPS d18O record exceeds the threshold of 1 sigma, but with maybe one data point to little to be identified as an HS. Of course age models are difficult to construct for the Nordic Seas, but can you provide a more thorough assessment of potential age biases due to the employed approach?
L242: Does this imply virtually no bioturbation in the core? Any bioturbation should obscure millennial-scale events at this low sed rate.
L286: Maybe I misunderstand, but 107 +- 7 seems to be well within error of 117 +- 11 µmol/kg for the Holocene and MIS5e. Also, are these 1 sigma uncertainties?
L307: There does not seem to be a clear millennial-scale variability in PS1243 as is seen in the Iberian Margin record. The statement of the anti-correlation therefore seems not fully supported by the data.
L335: This is really hard to see in Fig. 6 as so many lines overlap. Can you visualize this more clearly? Further, the GeoB records have a very low temporal resolution, which makes such statements not supported by the data for these records.
L342: U1313 (and also PS12543) does not really have the temporal resolution to make such statements. Mostly, just a single data point falls into the stadial periods.
L366: As mentioned before, the average CO32- concentrations at MIS5e and the Holocene agree within error. The discussion should therefore be more nuanced on this regard.
L375: Galaasen et al. (2020, Science) suggested centennial to millennial scale NADW variability also during MIS5e. Do you see any such variability, or rather could such events add to some noise the PS1243 record? Or can you exclude these events?
L379: “growth growth” delete one.
L390: Most models than run beyond 2100 show eventually an AMOC recovery often to a stronger state than under PI conditions. See for instance the results of the LongRunMIP (Bonan et al., 2022). Only transiently the AMOC and Nordic Seas deep water formation weakens. On the timescale this study looks at, one would expect a stronger than Holocene circulation at MIS5e.
L430: After HS10 not during.
L483: Maybe phrase more carefully, since PS1243 does not exhibit a MIS2 section.
L505: promoted instead of enforced.
L527 following: As mentioned before, the record of PS1243 doe not really have the resolution to resolve these millennial-scale events. The following paragraph therefore seems too speculative and not well-enough supported by the data.
Fig. 1A: Please highlight the core location a bit more predominantly. The panel generally feels a bit too busy. Maybe some elements can be removed or highlighted differently (e.g., there are a lot of dashed lines).
Fig 3: The Mn/Ca ratio seems to have a consistent downward trend from 130 ka to 70 ka. How can this be explained?
Fig. 8: All three panels are very crowded and it’s hard get a good overview of the records and core sites. I don’t have an obvious suggestions to redesign the figure, but I would greatly appreciate if the authors find a cleaner way to visualize the data.
Citation: https://doi.org/10.5194/egusphere-2024-3163-RC1 -
RC2: 'Comment on egusphere-2024-3163', Thomas Chalk, 29 Nov 2024
reply
The paper from Stobbe et al., presents new foraminifera and pteropod data from the Norwegian Sea, covering most of the last glacial cycle. As far as I am aware this the most extensive high latitude study using B/Ca and I comment the authors for their efforts. Given its location in the central Norwegian Sea the age model seems more-or-less robust, though I would be cautious to over interpret the minutia, especially given the proxies used. I presume this is why most of the discussion is focused on MISs 4–5 where the data quality is best. The authors use the data, in context of other records to show that deep-water formation was likely a persistent feature of this region, even through an interval of major glaciation.
The paper is well presented and the data are of high quality, I have no doubt of the worthiness for publication. My comments here are mostly focused on the context and the presentation of data, which I hope will improve the manuscript.
Throughout the introduction I found the referencing to be a bit limited, this is especially stark when compared to the discussion (which I think is very good). It would be nice if the all papers that the data are compared to in the discussion (e.g. all the B/Ca records from table 1) were at least mentioned in the introduction, as that is really what sets up the discussion from the current state of the art.
I have an issue with the way the B/Ca method is presented. Especially given that many comparisons are made between datasets. The authors use JCt to normalize the B/Ca, with a value of 218 µmol/mol, but other publications have this as low as 190 µmol/mol (Hathorne et al 2013) which would make a ~10% difference. I realize the majority of the data will have used the sensitivity of C. wuellerstorfi to CO32- and calibrated the coretops, meaning the absolute B/Ca matters little, but is that true in every case? And it would be good regardless to make sure that the B/Ca data in its raw form is comparable in an honest way for future studies.
Table 1: Missing a few datasets, such as Kirby et al 2020 and Oppo et al. 2023, is there a reason these (shorter?) records are omitted from the compilation? They are both from useful areas to fill gaps.
When comparing the datasets, I think in general the authors rely too much on their age model and maybe more importantly, those of the other studies, as well as the precision of the proxy to make detail comparisons. Alongside the analytical uncertainty. The B/Ca-CO32- proxy (Yu and Elderfield 2007) has a calibration uncertainty of ~10µmol/kg, so a more statistical approach is warranted between some of the comparisons between sites, especially as in the literature the delta carbonate has been transformed into carbonate ion in more than one way.
Even if you age model is good enough (I am not an expert in this area), are the others good enough to draw the comparisons made on millennial timescales? I see another comment to this effect which covers my sentiments on this well.
Fig4: The presence of pteropods box here should be moved off axes to show the ‘near zero’ values in MIS4. I like the combination of the proxy datasets, and if the aragonite CCD is responsible for the presence and absence which is not shown in the carbonate ion records (e.g. Sulpis et al 2022) that would be very interesting. I question whether it is feasible with an epifaunal benthic species though, and do the authors have any thoughts on what you might expect the overall % aragonite to be pre-dissolution over this interval? A back of the envelope calculation for how it is changing the carbonate system would be really nice. Also, regarding the interpretation of this, is it a local effect or are you implying that pteropod dissolution is a key mechanism for increasing the carbonate ion of all the deep water emanating from this basin? As such overriding the potential of CO2 invasion and biological processes? Or should we not interpret the high CO32- values in MIS4 here outside the local area?
Minor points
Please be consistent with the use of hyphens (-) and ‘n-dashes’ (–) throughout, at the moment they are mixed. ‘N-dashes’ should be used for all ranges.
Line 9: Simplify to say ‘higher CO32- values in MIS4 and 5 than the Holocene’.
28: careful when talking about a ‘transition’ with two non-adjacent stages. I would prefer ‘between’ or similar when talking about an extended time period.
43: denser water masses, it’s all relative.
55: Crocker et al 2016 could be added as a reference here (see point above).
76: please give values as well as the CO2 decline.
116: AF, defined by s=35
165: I might be out of date here, but I thought the value was globally closer to 0.48 (Marchitto et al 2014) is there a reason that locally 0.64 works better here?
196: I always worry about this use of multiple samples to represent a stable late Holocene situation, is it valid? Are the uncertainties propagated through the rest of the record? 1.4µmol/mol seems very low for a measurement error combined with a n=3 grouping.
200: the use of 10±5 seems a bit arbitrary here, is it just somewhere in between the other estimates? Would you explain better the process of finding this number?
284: See above, but stating that PS1243 is ‘consistently’ higher seems like an overstatement. Slightly higher?
291: contextualise “low CO2” MIS 5b (and d) is still an interglacial interval.
304: here you say the variations are similar to MD95-2039 but above you say significantly offset. Explain more precisely the situation.
Fig 5: The colour scale in the middle makes it look related to the y-axis. Would work better as a traditional legend.
329: I’m not sure this point is borne out by the data, at least as presented here. Perhaps some statistics or an additional figure (histogram of values?) could help? U1308 and RC 16-59 data also look like they are in approximately the same area and are both >2km depth.
336: U1313 is in the western Atlantic basin, not the east.
Figure 6: This figure is a bit strangely organized, I’m not sure why the MD95 and PS1243 records both appear twice, and in the case of the latter with different symbols. The mixture of carbonate ion and delta carbonate ion (an estimation of CO32-sat could be made?) as well as the separation of the panels without labelling why they are divided so makes it difficult to read.
362: I’m not sure that the arguments on this are so one-sided? See Gallaasen et al. 2014 and others, the geochemical studies cited here are consistent with multiple potential export strengths, as they are looking at stratification and water sources.
379: growth appears twice.
403: remove transition here.
410: see new temperature estimates in Morley et al 2024, which shows even larger T changes.
487: you have und in place of and.
Fig 8: This figure is a bit bizarre, I think it needs to be bigger and have the map placed above or below the data for clarity. The colour palette could also be used to aid interpretation instead of the (random?) selection currently used. The yellow is difficult to see, and it took me a while to notice the x-axes are not equally scaled. If recalibrating data (e.g. 980) why not use the higher resolution data from Crocker et al. 2016 also? And maybe picked a preferred calibration.
562: more or deeper records required, as U1313 is already W. Atlantic.
References mentioned.
Oppo, D. W., Lu, W., Huang, K.-F., Umling, N. E., Guo, W., Yu, J., et al. (2023). Deglacial temperature and carbonate saturation state variability in the tropical Atlantic at Antarctic Intermediate Water depths. Paleoceanography and Paleoclimatology, 38, e2023PA004674. https://doi.org/10.1029/2023PA004674
Kirby, N., Bailey, I., Lang, D.C., Brombacher, A., Chalk, T.B., Parker, R.L., Crocker, A.J., Taylor, V.E., Milton, J.A., Foster, G.L. and Raymo, M.E., 2020. On climate and abyssal circulation in the Atlantic Ocean during late Pliocene marine isotope stage M2,∼ 3.3 million years ago. Quaternary Science Reviews, 250, p.106644.
Galaasen, E.V., Ninnemann, U.S., Irvalı, N., Kleiven, H.K.F., Rosenthal, Y., Kissel, C. and Hodell, D.A., 2014. Rapid reductions in North Atlantic Deep Water during the peak of the last interglacial period. Science, 343(6175), pp.1129-1132.
Morley, A., de la Vega, E., Raitzsch, M., Bijma, J., Ninnemann, U., Foster, G.L., Chalk, T.B., Meilland, J., Cave, R.R., Büscher, J.V. and Kucera, M., 2024. A solution for constraining past marine Polar Amplification. Nature Communications, 15(1), p.9002.
Hathorne, E.C., Gagnon, A., Felis, T., Adkins, J., Asami, R., Boer, W., Caillon, N., Case, D., Cobb, K.M., Douville, E. and Demenocal, P., 2013. Interlaboratory study for coral Sr/Ca and other element/Ca ratio measurements. Geochemistry, Geophysics, Geosystems, 14(9), pp.3730-3750.
Citation: https://doi.org/10.5194/egusphere-2024-3163-RC2
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 209 | 53 | 11 | 273 | 6 | 5 |
- HTML: 209
- PDF: 53
- XML: 11
- Total: 273
- BibTeX: 6
- EndNote: 5
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1