the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 08 Oct 2024
Quantifying the feedback between Antarctic meltwater release and subsurface Southern Ocean warming
Abstract. The subsurface ocean around Antarctica is one of the primary drivers of mass loss from the Antarctic ice sheet through the basal melting of ice shelves. The resultant meltwater flux into the surrounding ocean can, mainly through reduced vertical mixing, further enhance subsurface ocean warming, inducing a positive feedback that amplifies mass loss and sea-level rise. Here, we quantify this feedback between Antarctic meltwater release and ocean warming using Linear Response Functions in the Earth System Model EC-Earth3. Increased meltwater release from five individual Antarctic ice sheet regions is found to unambiguously warm the subsurface Southern Ocean at centennial time scales. This warming response is quantified in terms of Linear Ocean Response Functions. Combining these with Linear Response Functions of the Antarctic mass loss and sea level rise because of ocean warming, allows for the quantification of the meltwater–ocean-warming feedback. Here, this feedback is calculated for ocean temperature projections from 14 CMIP6 Earth System Models, and Linear Response Functions from 8 Ice Sheet Models. Using a fixed basal melt relation to ocean temperatures, the feedback enhances 21st century projections of the Antarctic sea-level contribution by approximately 80 %. Calibration of the basal melt relation on observed historical icemass loss reduces this effect to 5 %. This calibration is achieved by reducing the basal melt parameter by 35 % on average. We propose that a similar reduction in the basal melt parameter must be applied in ice-sheet model forcing when transitioning from a stand-alone setup to an ice-sheet–ocean coupled setup in which the meltwater–warming feedback is explicitly simulated.
-
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
(5016 KB)
-
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(5016 KB) - Metadata XML
- BibTeX
- EndNote
- Final revised paper
Journal article(s) based on this preprint
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2024-2257', Anonymous Referee #1, 12 Nov 2024
Please find attached a pdf with my review of “Quantifying the feedback between Antarctic meltwater release and subsurface Southern Ocean warming” by Lambert et al.
- AC1: 'Reply on RC1', Erwin Lambert, 16 Jan 2025
-
RC2: 'Comment on egusphere-2024-2257', Anonymous Referee #2, 15 Dec 2024
Review of “Quantifying the feedback between Antarctic meltwater release and subsurface Southern Ocean warming” by Lambert et al.
The manuscript by Lambert and colleagues uses output from Earth system models, ice sheet models, and observations to estimate the impact of an ocean-ice feedback on sea level projections. The majority of Earth system models do not yet include coupled ice sheets and ice shelves, and thus ice-ocean feedbacks are not currently included in sea level projections. Additional meltwater from Antarctica increases stratification, causing surface cooling and subsurface warming at the depth of ice shelf basal melting, which has the potential to accelerate melting. Making use of existing model output, the authors calculate linear response functions to estimate the impact of this feedback on sea level projections. They find at first pass, that this feedback could increase sea level projections for 2100 by 80%, which is a substantial effect. However, they also note that the basal melt parameter for this estimation has not been calibrated on observed historical melt rates. When accounting for calibration, they revise their estimate to suggest that the feedback could increase sea level projections for 2100 by 5%.
The findings of this manuscript will be useful to the scientific community working towards coupling dynamic ice sheets and shelves into Earth system models. The manuscript is well written, and the methodology is clear described. I have several minor comments specified below, which I expect will be straightforward to address. I also have one major comment, regarding bias / potential bias in the Earth system model ocean component that will influence the ocean response function and results – sensitivity to this should be investigated and presented – currently, it is not clear how dependent the results are too possible ocean-model bias in both the mean state and trends. I thus recommend major revisions for this manuscript.
Specific comments
Abstract line 10-14: The three sentences with reported percentages read a bit as unconnected listed items, and this has the effect of making the percentages seem unrelated. Rewording the calibration methodology and revised sea level enhanced projection sentences would help here.
Line 40: “SOFIAMIP project” would be better described as “Southern Ocean Freshwater Input from Antarctica (SOFIA) Imitative”.
Line 40-42: Can you add citations for these modelling studies in the last sentence – as it is currently written, this seems like it is referring to the SOFIA model output from Swart et al. (2023), which is not the case.
Line 50-51: Do any of models or the multimodel mean agree with the observations?
Line 89: Suggest “redistributed” here.
Line 99-102: It is implicit here that the Antarctic runoff generated in the CMIP6 version of the model is no longer generated in this version, but it would be helpful to describe this explicitly, so that it is clear you aren’t double counting this runoff plus your more realistic representation of meltwater release.
Line 142: Does “bias correct” mean bias correct the mean state, or to bias correct the trends, i.e. de-drift? I think it means the latter. How is this bias correction / de-drifting done? E.g. have you removed linear grid point trends from the piControl in the corresponding historical and SSP scenarios?
Further, the historical CMIP6 ocean trends around the Antarctic margins themselves are biased (Purich and England 2021) – have you corrected for this? For example, multimodel mean historical trends around the margins show more warming than observed in most regions, except the Amundsen Sea, over 1975-2012. The historical ocean trends will presumably exert a strong influence on your projections and gamma calibration. Sensitivity to model bias in both the mean-state and trends should be investigated and reported in the manuscript.
Purich, A., & England, M. H. (2021). Historical and future projected warming of Antarctic Shelf Bottom Water in CMIP6 models. Geophysical Research Letters, 48, e2021GL092752. https://doi.org/10.1029/2021GL092752
Line 169: Computation.
Line 204-207: It would be useful if you list the ice sheet models and ESMs used in this study in a table.
Line 220-222: It would be useful here to specify that you are using the observed historical ice mass loss, and the modelled ocean warming. What is the uncertainty in the historical ice mass loss? What is the uncertainty in the ocean warming?
Line 232-235: What does it mean that some model pairs require a negative gamma to reproduce the positive sea level contribution? Is the model representation of ocean temperature around the Antarctic margins suitable for this use?
Fig. 4 caption: Capitalisation should be “western Peninsula” and “eastern Peninsula”
Line 275: In this section it would be useful here to comment on this subsurface warming relative to other studies (particular in the Amundsen Sea), and comment on the resolution of this model / limitations to representation of high-resolution processes that may affect the response to meltwater.
Line 293-295: This saturation effect has been reported previously, e.g. in Schloesser et al. (2019).
Schloesser, F., T. Friedrich, A. Timmermann, R. M. DeConto, and D. Pollard, 2019: Antarctic iceberg impacts on future Southern Hemisphere climate. Nat. Climate Change, 9, 672– 677, https://doi.org/10.1038/s41558-019-0546-1.
Line 296-298: The upper limit seems reasonable. Did you experiment with other (non-linear) fits?
Line 319: Specify “subsurface” warming.
Line 353-354: Specify “the last decades of our projections”, so it is clear that you are not referring to the last decades of the observations. Also, here and elsewhere, the date that you are presenting projections for (i.e. 2100) should be specified.
Line 363: Specify “surface” cooling.
Line 364-365: Can you elaborate on why it could lead to an overestimation of sea-level projections?
Line 375-377: Have you shown the warming around Antarctica in EC-Earth3, compared to other models? Can you support this statement with references?
Line 380: “Contrast”
Conclusions: Some of the conclusions text seems like it is written in brief draft format. I recommend revising the text.
Citation: https://doi.org/10.5194/egusphere-2024-2257-RC2 - AC2: 'Reply on RC2', Erwin Lambert, 16 Jan 2025
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2024-2257', Anonymous Referee #1, 12 Nov 2024
Please find attached a pdf with my review of “Quantifying the feedback between Antarctic meltwater release and subsurface Southern Ocean warming” by Lambert et al.
- AC1: 'Reply on RC1', Erwin Lambert, 16 Jan 2025
-
RC2: 'Comment on egusphere-2024-2257', Anonymous Referee #2, 15 Dec 2024
Review of “Quantifying the feedback between Antarctic meltwater release and subsurface Southern Ocean warming” by Lambert et al.
The manuscript by Lambert and colleagues uses output from Earth system models, ice sheet models, and observations to estimate the impact of an ocean-ice feedback on sea level projections. The majority of Earth system models do not yet include coupled ice sheets and ice shelves, and thus ice-ocean feedbacks are not currently included in sea level projections. Additional meltwater from Antarctica increases stratification, causing surface cooling and subsurface warming at the depth of ice shelf basal melting, which has the potential to accelerate melting. Making use of existing model output, the authors calculate linear response functions to estimate the impact of this feedback on sea level projections. They find at first pass, that this feedback could increase sea level projections for 2100 by 80%, which is a substantial effect. However, they also note that the basal melt parameter for this estimation has not been calibrated on observed historical melt rates. When accounting for calibration, they revise their estimate to suggest that the feedback could increase sea level projections for 2100 by 5%.
The findings of this manuscript will be useful to the scientific community working towards coupling dynamic ice sheets and shelves into Earth system models. The manuscript is well written, and the methodology is clear described. I have several minor comments specified below, which I expect will be straightforward to address. I also have one major comment, regarding bias / potential bias in the Earth system model ocean component that will influence the ocean response function and results – sensitivity to this should be investigated and presented – currently, it is not clear how dependent the results are too possible ocean-model bias in both the mean state and trends. I thus recommend major revisions for this manuscript.
Specific comments
Abstract line 10-14: The three sentences with reported percentages read a bit as unconnected listed items, and this has the effect of making the percentages seem unrelated. Rewording the calibration methodology and revised sea level enhanced projection sentences would help here.
Line 40: “SOFIAMIP project” would be better described as “Southern Ocean Freshwater Input from Antarctica (SOFIA) Imitative”.
Line 40-42: Can you add citations for these modelling studies in the last sentence – as it is currently written, this seems like it is referring to the SOFIA model output from Swart et al. (2023), which is not the case.
Line 50-51: Do any of models or the multimodel mean agree with the observations?
Line 89: Suggest “redistributed” here.
Line 99-102: It is implicit here that the Antarctic runoff generated in the CMIP6 version of the model is no longer generated in this version, but it would be helpful to describe this explicitly, so that it is clear you aren’t double counting this runoff plus your more realistic representation of meltwater release.
Line 142: Does “bias correct” mean bias correct the mean state, or to bias correct the trends, i.e. de-drift? I think it means the latter. How is this bias correction / de-drifting done? E.g. have you removed linear grid point trends from the piControl in the corresponding historical and SSP scenarios?
Further, the historical CMIP6 ocean trends around the Antarctic margins themselves are biased (Purich and England 2021) – have you corrected for this? For example, multimodel mean historical trends around the margins show more warming than observed in most regions, except the Amundsen Sea, over 1975-2012. The historical ocean trends will presumably exert a strong influence on your projections and gamma calibration. Sensitivity to model bias in both the mean-state and trends should be investigated and reported in the manuscript.
Purich, A., & England, M. H. (2021). Historical and future projected warming of Antarctic Shelf Bottom Water in CMIP6 models. Geophysical Research Letters, 48, e2021GL092752. https://doi.org/10.1029/2021GL092752
Line 169: Computation.
Line 204-207: It would be useful if you list the ice sheet models and ESMs used in this study in a table.
Line 220-222: It would be useful here to specify that you are using the observed historical ice mass loss, and the modelled ocean warming. What is the uncertainty in the historical ice mass loss? What is the uncertainty in the ocean warming?
Line 232-235: What does it mean that some model pairs require a negative gamma to reproduce the positive sea level contribution? Is the model representation of ocean temperature around the Antarctic margins suitable for this use?
Fig. 4 caption: Capitalisation should be “western Peninsula” and “eastern Peninsula”
Line 275: In this section it would be useful here to comment on this subsurface warming relative to other studies (particular in the Amundsen Sea), and comment on the resolution of this model / limitations to representation of high-resolution processes that may affect the response to meltwater.
Line 293-295: This saturation effect has been reported previously, e.g. in Schloesser et al. (2019).
Schloesser, F., T. Friedrich, A. Timmermann, R. M. DeConto, and D. Pollard, 2019: Antarctic iceberg impacts on future Southern Hemisphere climate. Nat. Climate Change, 9, 672– 677, https://doi.org/10.1038/s41558-019-0546-1.
Line 296-298: The upper limit seems reasonable. Did you experiment with other (non-linear) fits?
Line 319: Specify “subsurface” warming.
Line 353-354: Specify “the last decades of our projections”, so it is clear that you are not referring to the last decades of the observations. Also, here and elsewhere, the date that you are presenting projections for (i.e. 2100) should be specified.
Line 363: Specify “surface” cooling.
Line 364-365: Can you elaborate on why it could lead to an overestimation of sea-level projections?
Line 375-377: Have you shown the warming around Antarctica in EC-Earth3, compared to other models? Can you support this statement with references?
Line 380: “Contrast”
Conclusions: Some of the conclusions text seems like it is written in brief draft format. I recommend revising the text.
Citation: https://doi.org/10.5194/egusphere-2024-2257-RC2 - AC2: 'Reply on RC2', Erwin Lambert, 16 Jan 2025
Peer review completion
Journal article(s) based on this preprint
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 396 | 157 | 193 | 746 | 27 | 56 |
- HTML: 396
- PDF: 157
- XML: 193
- Total: 746
- BibTeX: 27
- EndNote: 56
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
Cited
Dewi Le Bars
Eveline van der Linden
André Jüling
Sybren Drijfhout
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(5016 KB) - Metadata XML