the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 14 Dec 2023
Contribution of physical processes to variability of dissolved silicate in the Labrador Sea between 1980 and 2015
Abstract. We use output of an eddy-permitting biogeochemical ocean general circulation model to investigate the drivers of the observed decline in the pre-bloom inventory of dissolved silicate (DSi) and its InterAnnual Variability (IAV) over the period 1980–2015. Specifically, the relative impacts of an abrupt decrease in Deep Winter Convection (DWC) and changes in Arctic inflow on DSi concentrations at the regional scale are examined. The IAV of the upper layer DSi inventory covaries with both the Arctic inflow and DWC, however, the pre-bloom decline seems driven primarily by the DWC and associated winter vertical mixing, while the contribution of Arctic inflow is negligible. Our study suggests that the inventory responds to natural decadal variability which is influenced by two major climate modes, the North Atlantic and the Arctic Oscillations, with the former appearing to be the main control.
- Preprint
(3073 KB) - Metadata XML
-
Supplement
(7793 KB) - BibTeX
- EndNote
Status: final response (author comments only)
-
RC1: 'Comment on egusphere-2023-2538', Anonymous Referee #1, 04 Feb 2024
In this study, the authors use output of an eddy-permitting biogeochemical ocean general circulation model to investigate the drivers of the observed decline in the pre-bloom inventory of dissolved silicate (DSi) and its InterAnnual Variability (IAV) over the period 1980-2015. The study results suggests that the inventory responds to natural decadal variability which is influenced by two major climate modes, the North Atlantic and the Arctic Oscillations, with the former appearing to be the main control.Overall, this manuscript is a valuable study that addresses an important and challenging topic. However, it needs moderate improvement and modifications before publication. The suggested comments please see as follows:
(1) The manuscript abstract should be revised and polished. The main analysis results and implications should be included in the abstract.
(2) In the introduction section, the manuscript should review the relevant literature more thoroughly.
(3) In discussion section, the manuscript should discuss more thoroughly about the results, as well as comparison with other studies in discussion section so that readers can better evaluate innovation significance of this study.
Specific comments
- 1.2 Observational dataset, there are many data sources, what are the data sources uncertainty ? Maybe the DSi analysis method is different. In addition, the Interpolated data was also adopted, such as from the Atlantic Repeat Hydrography Line 7 West.
- In the Contribution of the Arctic inflow section, please add the calculation errors.
- In 5 conclusion section:“Our results suggest that the regime shift in DWC did not extend to the DSi inventory or the NPP, whose variability also depends on other factors”. What are the other factors? Please describe in details.
- The conclusion sectionis drawn based on the long-term dateset, and is very interesting. The conclusions should be expanded and focus on the research significance.
Citation: https://doi.org/10.5194/egusphere-2023-2538-RC1 -
RC2: 'Comment on egusphere-2023-2538', Anonymous Referee #2, 24 Feb 2024
“Contribution of physical processes to variability of dissolved silicate in the Labrador Sea between 1980 and 2015” by Alizée Dale, Marion Gehlen, Douglas W. R. Wallace, Germain Bénard, Christian Éthé, and Elena Alekseenko discusses the reasons behind the nutrient trend in the Labrador Sea, and the major conclusion made in the paper that the reason for the decline is weakening of convection rather than advection from the Arctic.
Unfortunately, I cannot recommend this manuscript for publication. The main message is not new. A similar story was published in 2022 and analyzed the connection between convection, nutrients and chlorophyll in detain. There are also many misconceptions and mistakes in interpretations
What advance has been made by this study in comparison to Tesdal et al., 2022, who showed that both nutrient replenishment of the upper layer and phytoplankton bloom are correlated with convection depth? Their figures also show how the correlation changes, or rather flips, with depth, which can be regarded as prove of the link to convection. The authors of the present work do not explain what new does their message add to the previous one. I see mentioning of Tesdal’s discussion of advection, but not the main conclusion about convection (like between lines 20 and 25). The fact that this time the same conclusion comes from model does not add to the point, because there is bias in simulated nutrient profile – nutrients there increase with depth greater than in reality, and model apparently overshoots convection in the 1990s – this is how I explain the fact that the model doubles up the natural change in nutrients, according to Figure 1. Using just a vertical nutrient profile and a time series of convection depths we can come up with the same statement. The question is how much. Considering that these biases are unrealistic, and expect just by mere entrainment of elevated deep nutrients, I cannot recommend this result for publication – a change in either ocean or biochemical model or both will most likely give different results.
I also do not think the authors proved or rejected the contribution of the Arctic outflow. First of all, the series showing Arctic outflow are correlated with convection – and this may be a direct response of the model to atmospheric forcing, which may or may not be realistic – models do not pick all processes right. So, a simple correlation with two dependent processes does not allow to realistically partition the signals by sources. Further, I do not understand, how the Arctic water gets into the central Labrador Sea immediately after entering the region. I thought it mostly follows the shelf and shelf break – getting to the center takes time, and may depend on many factors, including wind. So, I would not make a statement that Arctic outflow rapidly, directly and strongly contributes nutrients to the water column. Then, the northern section shown is not Davis Strait – it is in the northern Labrador Sea, and part of it is influenced by the West Greenland Current. I assume the authors computed the net transport across that line, but how much of nutrients are transferred north, recirculate and return. What would a true Davis Strait section show? There are too many caveats in bringing everything together … What is the role of Denmark Strait outflow? Irminger Current?
I cannot agree with the point that the authors made about NAO and AO being unrelated – these two indices are not independent. In fact, the regions used to compute the two substantially overlap. So, stating that AO controls the Arctic outflow, and NAO convection is essentially stating that all four are correlated. Therefore, we cannot separate the dependent factors and their results using statistical methods.
I am a bit puzzled with the Arctic outflow varying by a factor of two, from year to year, I mean – is it really so? Can the authors support this statement with observations? There is a chance that what they see is a result of pushing that line into the Labrador Sea – by no means I would call it a Davis Strait line.
There is mix-up in pre-bloom and post-bloom data validation as the ship data were collected in random seasons with respect to bloom, an in most cases the euphotic nutrient concentrations are low – as being mostly consumed, so any comparison has this element of uncertainty and seasonality, and not really knowing if the spring bloom already took place or not. Using the model data for pre-bloom removes the comparison aspect as the whole Arctic outflow is focused on the top 200 m – too fresh and low-dense.
Line 68: Why using the OSNAP section to close the Labrador Sea, when OSNAP cuts through convection zone on the Greenland side.
Line 93: Seems like there is too much noise in the observations. Probably because seasonality and sampling issues. I also do not agree that the match is good in Figure 1.Other than general response on convection, as I already mentioned, the agreement is weak.
What is conservative temperature?
Line 108: Compared with a single cruise? I do not think this is enough when the Davis Strait flux interannually changes at factor of two, if not more (although I really doubt that).
Line 116: The 0-450 seasonality is huge – and nutrients are mostly depleted in the top layer, and the longer the more.
Line 166-167: I disagree with that statement, see NAO vs AO above.
Figure 2. I have an issue with the Davis Strait line being inside the Labrador Sea, and therefore affected by the Labrador Sea inventory, hence correlations. Not clean.
Line 225: Where is Tesdal et al., 2022 here? They showed the two correlated, but they excluded the top layer, to minimize the effect of blooms.
Figure 3a – There is a vertical bias in model Si profiles, so the numbers are affected.
Another issue is the convection depths. Figure 3d shows no deep convection in 2008-2015, but there were quite a few. The pattern does not agree with observations.
Lines 290-294: Larger variability can be explained by larger variability - I do not think this explains it. As a matter of fact, I provided the explanation already – bias in convection depth + bias in vertical Si profile.
Line 307: Was convection of 1994 weak?
Line 426: The authors did not convince me that they have a reliable number.
Line 336: This is a conclusion from Tesdal et al. paper! Furthermore, the match is not very clean.
Lines 340-349: A strange explanation without showing the heat atmospheric forcing (NAO does not tell about the actual forcing in some years, and definitely not linearly dependent). Not clear about preconditioning and climate models.
Lines 350-350: The DS line is inside the Labrador Sea, so could be affected by convective mixing, as well as some processes of local dynamics.
Line 355: 2010 was warm year, but I would not connect it to Arctic as it takes time, same as with Beaufort Gyre, even more there.
Overall, the authors do not seem to give any delay to Arctic signal, linking immediate reaction in nutrients to Arctic change. If, as they suggest, AO influences Arctic, it may take years for Si to respond. Overall, I think what they see is a local response of ocean model rather than advection from the north. To prove their point, they need to start from the Beaufort Gyre. At this time, there is no support for any statement, other than AO-NAO twins.
Line 358: Zhang et al., 2021 suggested 10 years for signal transfer.
Lines 260-370: Again, not related to the story as the timing is not zero.
Line 374: I do not see Arctic transport associated with Arctic circulation mode changes, and convincing signal spreading to the Labrador Sea.
Line 410: Again, shown by Tesdal. By the way, there is no nutrient observations before 1987 to tell about 1980-1986.
Line 420: NAO and AO cannot decorrelate – apparently the authors are unaware of how the two are compute, what domains used for that. Absolutely wrong statement. For it to become true, the Icelandic Low should be removed from atmosphere.
Convection depths are totally wrong! By no, by no, by no means convection in 2009 was any deeper than 2008. Same is applied for a long period where convection, according to the authors, hardly change. Moreover, I object using the method they used for the definition of MLD, as in many years the threshold may think that the previous year MLD was formed in the current one.
Overall, in my opinion the authors did not achieve the goal listed in the title. The physical processes were misrepresented, some concepts are incorrect, and the biases in the models and experimental design affect interpretations. A response time of the Labrador Sea on the Arctic change is significant. So, I think, when seeing the correlated convection and Arctic signal, they both are related to convection. As I said, a change in 2 Sverdrups is huge for a narrow shallow strait.
Citation: https://doi.org/10.5194/egusphere-2023-2538-RC2 -
RC3: 'Comment on egusphere-2023-2538', Anonymous Referee #3, 24 Feb 2024
The manuscript by Dale et al. explores the variability of dissolved silicate in the Labrador Sea from 1980 to 2015 using the NEMO-PISCES biogeochemical ocean GCM. They show that the decadal variability of silicate covaries with the extent of winter convection and the Arctic inflow. They link a decline in pre-bloom silicate mostly to changes in the winter convection, while the contribution of Arctic inflow is negligible. They further link this to natural variability as seen from the winter NAO and AO indices.
The manuscript presents a comprehensive work utilizing the NEMO ocean model to explore the underlying drivers of silicate variability in the Labrador Sea. The study as great potential to advance our understanding of biogeochemical dynamics in an important region, particularly in the context of ongoing climate change. However, I think the manuscript requires major revisions to fulfill its potential fully.
The most critical flaw is the absence of a formal nutrient budget analysis. Given the model's capabilities to integrate physical transport with ecosystem dynamics, the study misses a crucial opportunity to dissect nutrient supply mechanisms and uptake processes in detail. This gap limits the manuscript's ability to offer comprehensive insights into the nutrient cycle in the Labrador Sea. I strongly suggest to conduct a detailed nutrient budget analysis, thereby providing a more robust examination of the physical supply and biological utilization processes. Besides this primary concern I have following comments for the authors to consider.
Specific Comments:
(1) Acronyms Usage (lines 12-13 and throughout the manuscript): The use of "InterAnnual Variability (IAV)" and "Deep Winter Convection (DWC)" is unnecessary and also not consistently used in the text. I recommend using common terminology without reliance on acronyms.
(2) Phrasing Issues (e.g., lines 22, 29, 31, etc.): Various phrasing issues require attention for clarity and appropriate wording (see my line-by-line comments below).
(3) Use notation consistent with previous literature (see my comments in Section 2.2.3)
(4) Result/Discussion organization: Results and Discussion sections need clear demarcation and focus. Expand Section 3.4 to provide a clearer link between silicate variability and net primary production (NPP). Also, the analysis presented in Section 4 about the NPP anomalies (Figure 5 and D1) and the role of AO and NAO (Figure B1, C1 and C2) belong into the Result section.
(5) Appendices A and B should be referred to and integrated more clearly in the main text. Consider integrating essential details into the main text and choose between appendices and supplementary material for coherence.
Line-by-line comments:
Line 22 "It’s particularly lies" should be reworded for clarity to something like “This region is particularly characterized by”
Line 24: Just as an example, here DWC is not used for “deep winter convection”, so it is confusing why DWC is not used here. Unless there is a specific process meant with DWC, I would suggest to avoid using this specific acronym. Also if these processes weaken or shut down, is it then still called “deep” winter convection? You might just call it “winter convection“ and classify it based on the decadal variability (i.e., strong versus weak winter convection).
Line 29: “including valuable fisheries, among others, of Atlantic salmon, cod, mackerel, and herring.” - confusing wording. Simplify it to “[…] supports higher trophic levels, including key fisheries such as Atlantic salmon, cod, mackerel, and herring."
Line 31: Diatoms are also also dependent on light and other nutrients. The phrasing does not really reflect that. Maybe change to: “While diatoms, like other phytoplankton groups, rely on light, nitrogen, phosphorus, and iron for growth, they uniquely require …”
Line 42: The term "unbalanced mixing" is unclear. Please specify what is meant by this term.
Line 42, 45. The acronym "SRAW" is introduced but not consistently used. Suggest omitting uncommon acronyms for clarity.
Lines 48-49:Is there a difference between "wintertime deep convection" and "deep winter convection”?
Lines 51-52: NPP is a measure of the rate of primary production, so this sentence confuses me as it seem to conflate NPP with its outcomes but making a distinction between NPP and primary production, which are sort of the same thing.
"A decrease in DSi could significantly reduce ocean primary productivity (i.e., NPP) and alter the phytoplankton community composition, potentially leading to a decrease in diatom biomass. Such changes might also affect sinking fluxes, further influencing the marine carbon cycle."
Lines 53 to 64: Merge the two paragraphs.
Line 55 “we briefly” - It is important to examine the link between DSi variability and NPP. Instead a more in-depth discussion on this aspect due to its critical implications for marine ecosystem. It is essential to answer the question “Why should we care about silicate in the Labrador Sea”?
Line 61: I don’t think “weather regimes” is the right term. As you are referring to climate indices here given the broader climatic implications of the NAO and AO, as "weather" typically refers to short-term atmospheric conditions.
Line 66: “two sections to the north and south of the Labrador Sea” - The map in Figure 1a suggest to me it is northwest and southeast.
Line 67: I believe this study could do a better job in defining it’s sections based on previous work. I am not convinced of a good reason why to choose their wide “northern section” instead of the obvious choice to use Davis strait and Hudson strait. This would have the advantage for direct comparison with existing observations of transport and hydrography
Line 70: Clarify what “shallowest extent of the maximum MLD” means. Shallowest in terms of spatial variation? Maximum in terms of annual maximum?
Lines 101-102: Clarify that the model spin-up period is only from 1958 to 1980. Detail whether this was a single continuous spin-up or if multiple cycles were used. A 22-year spin-up for an ocean model seems very short to have a steady state. So I am skeptical all the details of the spin up process are provided.
Line 106: 'Units were converted as needed' sounds very vague and unclear. Please provide more specifics. It is crucial to specify which units were converted, into what they were converted, and any assumptions made during these conversions (e.g., what density value is used for volume to mass conversions).
Line 112: The World Ocean Atlas (WOA) is a gridded observation-based product from optimal interpolation of observational data. It is no not a reanalysis product.
Line 115: "The phrase 'evaluate the modeled variability' requires further clarification. Please specify the methods and metrics used for this evaluation. Detailing how model variability was assessed, including any statistical analyses or comparison benchmarks.
Line 122: Integrating NPP over time and depth converts it from a rate to a cumulative measure, such as total production. So I would not call it “NPP” as it is now a measure of the total carbon fixed rather than a production rate.
Line 124: "NPP exceeds its yearly median by 5%” The median of what exactly? Yearly median? Spatial median? And 5% of what?
Line 139: The term 'effective transport (Teff)' is used without a reference, which might be unclear to readers as 'T' is commonly used to denote temperature in the literature. To avoid confusion and ensure clarity, I recommend adopting a more conventional notation for transport that aligns with previous literature.
Equations 1 to 3 : I suggest using common notation that is more consistent with established literature in physical oceanography.
Line 148-149: Change phrase to “By convention, northward (southward) and eastward (westward) transports are positive (negative).
Line 156: Rephrase to clarify what diagnostics are evaluated at model time step and which ones have to be derived offline.
Line 161: Use common notation such as u, v for lateral and w for vertical. Instead of “mld” it should be z_{mld} to denote the vertical axis (z).
Line 166: Again “weather regimes” is not the right term to describe conditions represented by climate indices.
Line 169: It would rather say that the NAO is “defined” by the pressure gradient instead of it “represents” the pressure gradient.
Line 174: 'DFS5.2' has not been defined nor a reference is given.
Line 175: Note the notational error with ‘\%.' Are there any references that can be used to set this number into context with previous work?
Line 177: “A similar method is used to quantify the AO.“ Instead of this just specify the method and present the AO along with NAO definition. Or add the same details you provided for NAO (e.g., dipole/EOF pattern, percentage of variability associated with first mode).
Line 184: Consider changing section numbering to include an 'S' in front, like 'Section S1.2,' for supplemental material.
Line 186 "Rephrase 'underestimate at the surface' to provide a specific range or clarify that surface concentrations are underestimated.
Lines 194-198: The sentence structure discussing DSi decline periods is too complex and needs simplification for clarity.
Line 203: Text denotes it as “Northern Sections”, But Figure 2a says “Davis Strait transport"
Lines 207-209: Suggest potential drivers of variability as indicated by the relationship between Arctic inflow and DSi inventory.
Figure 2: Improve the distinguishability of trend lines in panel a by using colors instead of grey shades
Figure 2: Why show panel b in Figure 2? It is not referred to in the main text.
Figure 2 Caption (Line 218) Say “panel b”not figure b
Lines 247-249: Please elaborate on the difference in time scales observed between the DSi transport's overall linear decline and the abrupt shift in annual maximum MLD over 1993-1995, discussing the implications of these temporal dynamics.
Section 3.4 is incomplete and lacks a clear purpose. I suggest extending this to provide clear evidence between linking silicate variability to NPP
Line 269-270: Discuss the NPP anomalies for the period 1994-2007 to provide a comprehensive view across all studied intervals.
Line 279: Clarify the rationale behind selecting the three specific time periods for investigating DSi changes.
Line 293: “than its measured counterpart.” sound confusing. Just say “observed DSi concentration”.Line 297: "The above values are given between 1990 and 2015 for comparison" - Don’t start a new paragraph clarifying something you stated in the previous paragraph. That just confuses the reader.
Lines 298-300: This sentence seems a bit convoluted. Consider rephrase to a more straightforward statement: The current simulation does not extend beyond 2015, leaving it uncertain whether the model would accurately reproduce the observed trends in the most recent period.
There are two sections numerated as 4.2.
Consider relocating the analytical content currently in Sections 4.2 to the Results section. The Discussion section should primarily interpret these findings within the context of existing literature, contrasting and comparing rather than introducing new results.
Figure 5 should be introduced in the Results section.
It would be useful to explore and discuss the roles of other nutrients and environmental factors within the model's analysis. This inclusion could provide a more comprehensive understanding of the system being studied and leverage the model's full capabilities.
Line 326: Please elaborate on why the DSi transport from the Arctic shows a correlation yet is considered negligibleLine 344: The term “warming and cooling seasons” is confusing. I would just call autumn-winter as “cooling season” Please specify if you're discussing a singular transitional season or multiple seasons.
Line 346: ”Replace 'density loss' with 'buoyancy gain’
The last paragraph of Section 4.2 focuses on the role of the Arctic Oscillation (AO) and Arctic transport as factors, but it should be clarified that these are not significant and restate the assessment that was done in the Result section.
Streamline Section 4.3 to better integrate with the overall narrative of the paper. Refine or remove phrases that do not directly contribute to the central thesis or findings.
Line 375: “RCP8.5 scenario and two CMIP5 models (MPI-ESM-LR and IPSL-CM5A-LR)” - Just say previous modeling work suggest… No need to introduce additional terminology/ details that are not associated with this study.
Line 376-382: unclear phrasing.“Intensification of the variability of the pressure dipole over the Pacific ” /““weakening of the variability of the NAO and of the pressure dipole ””Line 385: Unclear phrasing “NPP exhibits a less significant negative” / “This low significance”
Line 410. Not clear what “indirectly” purpose is here
Line 413: sentence structure makes this unclearLine 431: The figure does not show vertical transport
Line 434 Figure AA?
Line 450 . Clarify that T us integrated over the upper 40 m
Figure B1: “ The red line”… you mean the “black line? I don’t see a red line
Figure C1: “ standard deviation of what (temporal or spatial)?
Figure C2: Denote what the purple arrow show, the gyre in the caption. Also clarify what is the difference between yellow and orange arrows. Why use two different colors?
Figure D1. Panel are missing labels (a to h) and titles
Line 502: Latex syntax: $<$Citation: https://doi.org/10.5194/egusphere-2023-2538-RC3
Data sets
Model output Alizee Dale https://doi.org/10.14768/99d68638-53d3-49b6-9af4-c19f21670a8e
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 247 | 57 | 27 | 331 | 31 | 16 | 15 |
- HTML: 247
- PDF: 57
- XML: 27
- Total: 331
- Supplement: 31
- BibTeX: 16
- EndNote: 15
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1