the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 25 Apr 2024
North Atlantic Subtropical Mode Water properties: Intrinsic and atmospherically-forced interannual variability
Abstract. This study investigates the contributions of the ocean’s chaotic intrinsic variability (CIV) and atmospherically-forced variability on the interannual fluctuations of the North Atlantic Eighteen Degree Water (EDW) properties. Utilizing a 1/4° regional 50-member ocean/sea-ice ensemble simulation driven by an original surface forcing method and perturbed initially, the forced variability of EDW properties is estimated from ensemble mean fluctuations, while CIV is determined from deviations around the ensemble mean within each member. The model successfully captures the main features of EDW, showing good agreement with observation-based ARMOR3D data in terms of location, seasonality, mean temperature and volume, and interannual variance of its main properties. CIV significantly impacts EDW, explaining 10–13 and 28–44 % of the interannual variance of its geometric and thermohaline mean properties, respectively, with a maximum imprint on EDW temperature. Observed and simulated intrinsic-to-total variance ratios are mostly consistent, dispelling concerns about a signal-to-noise paradox. This study also illustrates the advantages of ensemble simulations over single simulations in understanding oceanic fluctuations and attributing them to external drivers, while also cautioning against overreliance on individual simulations assessments.
-
Notice on discussion status
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
-
Preprint
(1972 KB)
-
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(1972 KB) - Metadata XML
- BibTeX
- EndNote
- Final revised paper
Journal article(s) based on this preprint
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2024-1146', Anonymous Referee #1, 30 May 2024
A review for egusphere-2024-1146
Title: North Atlantic Subtropical ModeWater properties: Intrinsic and atmospherically-forced interannual variability
Authors: Olivier Narinc, Thierry Penduff, Guillaume Maze, Stéphanie Leroux, and Jean-Marc Molines
Overall: The present study investigates intrinsic variability of the Eighteen Degree Water (EDW), the subtropical mode water of the North Atlantic using a 50-member ensemble simulation with 1/4-degree horizontal resolution, eddy-permitting ocean general circulation model. The new method of estimating the surface heat flux for each ensemble member is introduced to avoid artificially damping the intrinsic variability. This new method is interesting. However, I have several questions and comments, as mentioned below, that should be addressed.
Recommendation: major revision
Main comments/questions:
1. The method for estimating surface heat flux of each member and its influence
As shown in Figure A1, this method of surface heat flux strongly affects the amplitude of intrinsic variability, which is the main topic of the present paper. Then I think that meaning of the method and its influence should be discussed further.
1-1. Although it is not clearly mentioned, I guess that this ensemble simulation with the same time-varying air-sea fluxes is original of the present study. If it is not the original of the present study, please add the reference(s) in the description of the method.
1-2. I think Appendix A should be included in the main text.
1-3. “as would be expected (i.e. with no excessive damping) in coupled ocean-atmosphere simulations” (line 109) I agree that it would be expected if the atmosphere has enough time to respond to SST sufficiently. However, it should be noted that observational data show enhanced upward surface heat flux over warm meso-scale eddies (e.g., Tomita et al. 2019, doi:10.1007/s10872-018-0493-x), implying that oceanic intrinsic variability might modify surface heat flux, as meso-scale eddies could be expected as intrinsic variability. Then I think that the actual situation is between the two methods of ensemble simulations, and this new method might overestimate strength of intrinsic variability as it is not damped by surface heat flux. As this discussion can directly relate to the main topic of the present study, the meaning of the method and this possibility of overestimation should be discussed carefully.
2. Reliability of the observational data product
2-1. If ARMOR3D is not reliable, as discussed around line 218, I think it would be better for the authors to use other observational data product(s) to evaluate the model result.
2-2. If ARMOR3D is not reliable for its amplitude of variability, I seriously wonder if the phase of variability in ARMOR3D is reliable. Please add some discussion on it.
3. Influence on the annual cycle
Line 195: “the large control exerted by the atmospheric annual cycle” As mentioned here, the annual cycle of forcing is exceptional. It would be interesting to investigate the time-scale dependence of CIV strength excluding the annual cycle.
Specific comments/questions:
Line 15: “a notable role in regional and global climate” It would be better to explain more explicitly.
Line 52: “EDW” should be STMW, I think.
Line 117: It would be good if the authors can add a brief comment on how the gridded T and S fields are dynamically consistent with the velocity field.
Line 147: Although I know that model simulations always have biases, it is usual to adjust the parameters to define the simulated EDW for comparison with the observed EDW. The authors may want to add some more explanation why they tried to adjust the parameters for the observed EDW.
Line 162: “latter two sets of” It would be better to describe more specifically.
Figure 4: Please improve the labels of Figure 4. Some of the labels on the panels in the right column are overlapped and cannot be read well.
Figure 5: As only correlations are discussed in this paragraph, it might be more appropriate to plot only correlations rather than using the Taylor diagram. As sometimes STDs are very different, it is difficult to compare the distribution of correlations in Figure 5.
Line 255-256: Although I agree that the range is overlapped, the distributions seem very different, and the discussion here seems not objective. The authors may want to add some more objective and quantitative discussion.
Line 291-296: Although it is good to mention here, I do not think the discussion in this paragraph is a new finding of the present study.
Citation: https://doi.org/10.5194/egusphere-2024-1146-RC1 - AC1: 'Reply on RC1', Thierry Penduff, 02 Aug 2024
-
RC2: 'Comment on egusphere-2024-1146', Anonymous Referee #2, 10 Jun 2024
Title: North Atlantic Subtropical ModeWater properties: Intrinsic and atmospherically-forced interannual variability
Authors: Olivier Narinc, Thierry Penduff, Guillaume Maze, Stéphanie Leroux, and Jean-Marc Molines
Note: The validation of the model set up and model experiments is outside of my expertise and hence I cannot provide a review of those.
Summary
In the manuscript “North Atlantic Subtropical Mode Water properties: Intrinsic and atmospherically-forced interannual variability” the authors investigate the contribution of the two different types of variability on the total interannual variability of the eighteen degree water (EDW) using a 50-member ocean/sea-ice ensemble simulation with a horizontal resolution of ¼°. They validate the model results against a gridded product based on observations. The authors use a combination of potential vorticity, density, and latitude/longitude criteria to define EDW which differ for the model output and observation-based product to account for differences in the datasets. The authors define the ensemble mean as the atmospheric-forced variability and the ocean’s chaotic intrinsic variability as differences of each member from the ensemble mean. Six properties of the EDW are investigated and the authors find that between 10-44% of the interannual variability can be explained by the ocean’s intrinsic variability depending on the property, with 44% found in temperature.
This paper is interesting and well written. However, I have questions regarding the analysis which I detail below. I hence recommend major revision for the manuscript.
General comments:
- The authors compare their model results to one gridded data product compared to observations. I cannot see in the presented evidence that the modelled and observation-based results agree as well as the authors claim (e.g. L5-7, L210, L243, L276-278). Fig 1 and 2 simulated and observed sections at 65°W water masses with low potential vorticity occur shallower and warmer to observations and simulated EDW seems to deepen and densify toward the east which is not visible in observations. Fig 3-5 show that the simulations seem to clearly overestimate the interannual variability compared to the observation-based product. Second, I am concerned about the short-comings of the observational-based product which is known – as the authors state – “to substantially underestimate the actual interannual ocean variability” (L218). I appreciated the authors reasoning to use an observation-based product not depending on an underlying model, however given the short-comings of the used product with respect to its interannual variability, which is the time-scale of interest in this manuscript, I would highly recommend to include a few ocean reanalysis products like ECMWF ORAS5, CMCC C-GLORS or GLORYS2V4 to enable a more robust model validation.
- The authors mention the arbitrary definition of EDW and I think their approach to based it on criteria of three different properties (potential vorticity, density and region) is good. However, it would be great if the authors could provide more information about why they choose the criteria as they are. The criteria differ notably for their simulation and the observation-based product and based of Fig. 1 and 2 it is not clear to me, why for the simulations the density range (1.2 kg/m^3) is so large compared to the observation-based product (B: 0.72 kg/m^3). In the abstract and throughout the manuscript the authors mentioned the good agreement between simulation and observation-based product for the mean EDW volume. However, they choose their criteria for observation-based EDW to match the ensemble EDW volume mean (L178-180, L210-211), so it is designed to match. As the mean EDW properties seems to depend on the choice density range and max PV, it would be good to show and discuss this dependency for model and observation-based product for a fairer comparison.
- The authors stating in the abstract and throughout the manuscript that the simulations are in good agreement with the observation-based product in terms of location, seasonality, mean temperature and volume. However, section 2.3 is to brief and from my understanding does not provide the evidence for their statements. In lon-depth space the simulated EDW is clearly shallower compared to the observation-based product. No maps of the spatial (lat/lon) distribution of EDW in simulations and in the observation-based product are shown. It would be also great to show the spatial distribution (lat/lon) of temperature, because from Fig. 1 and 2 it looks like it varies with longitude in the simulations and is not constant. How does the spatial variance of temperature compare between model and observation-based product? How does this impact the temporal variability of the spatially averaged EDW temperature? A section/figures about the seasonal cycle for the different properties is missing.
Figures
I would suggest to add a,b,c labels to any figure, as they all consist of several subpanels and it would make referencing easier.
Figure 1 and 2: As sensitivity B was chosen for the comparison it would be better to show the B limits for the observation-based product instead of the A-limits.
Figure 4 and 5: The RMSE contours and numbers are too fade to be readable. Also the correlation labels of the right hand side panels overlap so that they are not readable. Please adjust this. Think about to change either the red or green to a different color as these are not colorblind friendly in one plot.
Minor:
L125: Add reference for Ertel PV definition
Throughout section 2.5: Units displayed as cursive inconsistent with other text.
Citation: https://doi.org/10.5194/egusphere-2024-1146-RC2 - AC2: 'Reply on RC2', Thierry Penduff, 02 Aug 2024
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2024-1146', Anonymous Referee #1, 30 May 2024
A review for egusphere-2024-1146
Title: North Atlantic Subtropical ModeWater properties: Intrinsic and atmospherically-forced interannual variability
Authors: Olivier Narinc, Thierry Penduff, Guillaume Maze, Stéphanie Leroux, and Jean-Marc Molines
Overall: The present study investigates intrinsic variability of the Eighteen Degree Water (EDW), the subtropical mode water of the North Atlantic using a 50-member ensemble simulation with 1/4-degree horizontal resolution, eddy-permitting ocean general circulation model. The new method of estimating the surface heat flux for each ensemble member is introduced to avoid artificially damping the intrinsic variability. This new method is interesting. However, I have several questions and comments, as mentioned below, that should be addressed.
Recommendation: major revision
Main comments/questions:
1. The method for estimating surface heat flux of each member and its influence
As shown in Figure A1, this method of surface heat flux strongly affects the amplitude of intrinsic variability, which is the main topic of the present paper. Then I think that meaning of the method and its influence should be discussed further.
1-1. Although it is not clearly mentioned, I guess that this ensemble simulation with the same time-varying air-sea fluxes is original of the present study. If it is not the original of the present study, please add the reference(s) in the description of the method.
1-2. I think Appendix A should be included in the main text.
1-3. “as would be expected (i.e. with no excessive damping) in coupled ocean-atmosphere simulations” (line 109) I agree that it would be expected if the atmosphere has enough time to respond to SST sufficiently. However, it should be noted that observational data show enhanced upward surface heat flux over warm meso-scale eddies (e.g., Tomita et al. 2019, doi:10.1007/s10872-018-0493-x), implying that oceanic intrinsic variability might modify surface heat flux, as meso-scale eddies could be expected as intrinsic variability. Then I think that the actual situation is between the two methods of ensemble simulations, and this new method might overestimate strength of intrinsic variability as it is not damped by surface heat flux. As this discussion can directly relate to the main topic of the present study, the meaning of the method and this possibility of overestimation should be discussed carefully.
2. Reliability of the observational data product
2-1. If ARMOR3D is not reliable, as discussed around line 218, I think it would be better for the authors to use other observational data product(s) to evaluate the model result.
2-2. If ARMOR3D is not reliable for its amplitude of variability, I seriously wonder if the phase of variability in ARMOR3D is reliable. Please add some discussion on it.
3. Influence on the annual cycle
Line 195: “the large control exerted by the atmospheric annual cycle” As mentioned here, the annual cycle of forcing is exceptional. It would be interesting to investigate the time-scale dependence of CIV strength excluding the annual cycle.
Specific comments/questions:
Line 15: “a notable role in regional and global climate” It would be better to explain more explicitly.
Line 52: “EDW” should be STMW, I think.
Line 117: It would be good if the authors can add a brief comment on how the gridded T and S fields are dynamically consistent with the velocity field.
Line 147: Although I know that model simulations always have biases, it is usual to adjust the parameters to define the simulated EDW for comparison with the observed EDW. The authors may want to add some more explanation why they tried to adjust the parameters for the observed EDW.
Line 162: “latter two sets of” It would be better to describe more specifically.
Figure 4: Please improve the labels of Figure 4. Some of the labels on the panels in the right column are overlapped and cannot be read well.
Figure 5: As only correlations are discussed in this paragraph, it might be more appropriate to plot only correlations rather than using the Taylor diagram. As sometimes STDs are very different, it is difficult to compare the distribution of correlations in Figure 5.
Line 255-256: Although I agree that the range is overlapped, the distributions seem very different, and the discussion here seems not objective. The authors may want to add some more objective and quantitative discussion.
Line 291-296: Although it is good to mention here, I do not think the discussion in this paragraph is a new finding of the present study.
Citation: https://doi.org/10.5194/egusphere-2024-1146-RC1 - AC1: 'Reply on RC1', Thierry Penduff, 02 Aug 2024
-
RC2: 'Comment on egusphere-2024-1146', Anonymous Referee #2, 10 Jun 2024
Title: North Atlantic Subtropical ModeWater properties: Intrinsic and atmospherically-forced interannual variability
Authors: Olivier Narinc, Thierry Penduff, Guillaume Maze, Stéphanie Leroux, and Jean-Marc Molines
Note: The validation of the model set up and model experiments is outside of my expertise and hence I cannot provide a review of those.
Summary
In the manuscript “North Atlantic Subtropical Mode Water properties: Intrinsic and atmospherically-forced interannual variability” the authors investigate the contribution of the two different types of variability on the total interannual variability of the eighteen degree water (EDW) using a 50-member ocean/sea-ice ensemble simulation with a horizontal resolution of ¼°. They validate the model results against a gridded product based on observations. The authors use a combination of potential vorticity, density, and latitude/longitude criteria to define EDW which differ for the model output and observation-based product to account for differences in the datasets. The authors define the ensemble mean as the atmospheric-forced variability and the ocean’s chaotic intrinsic variability as differences of each member from the ensemble mean. Six properties of the EDW are investigated and the authors find that between 10-44% of the interannual variability can be explained by the ocean’s intrinsic variability depending on the property, with 44% found in temperature.
This paper is interesting and well written. However, I have questions regarding the analysis which I detail below. I hence recommend major revision for the manuscript.
General comments:
- The authors compare their model results to one gridded data product compared to observations. I cannot see in the presented evidence that the modelled and observation-based results agree as well as the authors claim (e.g. L5-7, L210, L243, L276-278). Fig 1 and 2 simulated and observed sections at 65°W water masses with low potential vorticity occur shallower and warmer to observations and simulated EDW seems to deepen and densify toward the east which is not visible in observations. Fig 3-5 show that the simulations seem to clearly overestimate the interannual variability compared to the observation-based product. Second, I am concerned about the short-comings of the observational-based product which is known – as the authors state – “to substantially underestimate the actual interannual ocean variability” (L218). I appreciated the authors reasoning to use an observation-based product not depending on an underlying model, however given the short-comings of the used product with respect to its interannual variability, which is the time-scale of interest in this manuscript, I would highly recommend to include a few ocean reanalysis products like ECMWF ORAS5, CMCC C-GLORS or GLORYS2V4 to enable a more robust model validation.
- The authors mention the arbitrary definition of EDW and I think their approach to based it on criteria of three different properties (potential vorticity, density and region) is good. However, it would be great if the authors could provide more information about why they choose the criteria as they are. The criteria differ notably for their simulation and the observation-based product and based of Fig. 1 and 2 it is not clear to me, why for the simulations the density range (1.2 kg/m^3) is so large compared to the observation-based product (B: 0.72 kg/m^3). In the abstract and throughout the manuscript the authors mentioned the good agreement between simulation and observation-based product for the mean EDW volume. However, they choose their criteria for observation-based EDW to match the ensemble EDW volume mean (L178-180, L210-211), so it is designed to match. As the mean EDW properties seems to depend on the choice density range and max PV, it would be good to show and discuss this dependency for model and observation-based product for a fairer comparison.
- The authors stating in the abstract and throughout the manuscript that the simulations are in good agreement with the observation-based product in terms of location, seasonality, mean temperature and volume. However, section 2.3 is to brief and from my understanding does not provide the evidence for their statements. In lon-depth space the simulated EDW is clearly shallower compared to the observation-based product. No maps of the spatial (lat/lon) distribution of EDW in simulations and in the observation-based product are shown. It would be also great to show the spatial distribution (lat/lon) of temperature, because from Fig. 1 and 2 it looks like it varies with longitude in the simulations and is not constant. How does the spatial variance of temperature compare between model and observation-based product? How does this impact the temporal variability of the spatially averaged EDW temperature? A section/figures about the seasonal cycle for the different properties is missing.
Figures
I would suggest to add a,b,c labels to any figure, as they all consist of several subpanels and it would make referencing easier.
Figure 1 and 2: As sensitivity B was chosen for the comparison it would be better to show the B limits for the observation-based product instead of the A-limits.
Figure 4 and 5: The RMSE contours and numbers are too fade to be readable. Also the correlation labels of the right hand side panels overlap so that they are not readable. Please adjust this. Think about to change either the red or green to a different color as these are not colorblind friendly in one plot.
Minor:
L125: Add reference for Ertel PV definition
Throughout section 2.5: Units displayed as cursive inconsistent with other text.
Citation: https://doi.org/10.5194/egusphere-2024-1146-RC2 - AC2: 'Reply on RC2', Thierry Penduff, 02 Aug 2024
Peer review completion
Journal article(s) based on this preprint
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 208 | 83 | 30 | 321 | 18 | 16 |
- HTML: 208
- PDF: 83
- XML: 30
- Total: 321
- BibTeX: 18
- EndNote: 16
Viewed (geographical distribution)
| Country | # | Views | % |
|---|---|---|---|
| United States of America | 1 | 166 | 43 |
| France | 2 | 59 | 15 |
| China | 3 | 34 | 8 |
| Netherlands | 4 | 16 | 4 |
| Germany | 5 | 14 | 3 |
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
- 166
Olivier Narinc
Guillaume Maze
Stéphanie Leroux
Jean-Marc Molines
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(1972 KB) - Metadata XML