the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
Antarctic climate response in Last-Interglacial simulations using the Community Earth System Model (CESM2)
Abstract. We examine results from two transient modelling experiments that simulate the Last Interglacial period (LIG) using the state-of-the-art Community Earth System Model (CESM2), with a focus on climate and ocean changes relevant to the possible collapse of the Antarctic ice sheet. The experiments simulate the early millennia of the LIG warm period using orbital forcing, greenhouse gas concentrations and vegetation appropriate for 127 ka; in the first case (127 ka) no other changes are made; in the second case (127 kaFW), we include a 0.2 Sv freshwater forcing in the North Atlantic. Both are compared with a pre-industrial control simulation (piControl). In the 127 ka simulation, the global average temperature is only marginally warmer (0.004 °C) than in the piControl. When freshwater forcing is added (127 kaFW), there is surface cooling in the NH and warming in the SH, consistent with the bipolar seesaw effect. Near the Antarctic ice sheet, the 127 ka simulation generates notable ocean warming (up to 0.4 °C) at depths below 200 m compared to the piControl. In contrast, the addition of freshwater in the North Atlantic in the 127 kaFW run results in a multi-millennial sustained cooling in the subsurface ocean near the Antarctic ice sheet. We explore the physical processes that lead to this new result and discuss implications for climate forcing of Antarctic ice sheet mass loss during the LIG.
- Preprint
(3442 KB) - Metadata XML
- BibTeX
- EndNote
Status: final response (author comments only)
-
RC1: 'Comment on cp-2024-19', Anonymous Referee #1, 08 May 2024
The comment was uploaded in the form of a supplement: https://cp.copernicus.org/preprints/cp-2024-19/cp-2024-19-RC1-supplement.pdf
-
RC2: 'Comment on cp-2024-19', Anonymous Referee #2, 13 May 2024
The comment was uploaded in the form of a supplement: https://cp.copernicus.org/preprints/cp-2024-19/cp-2024-19-RC2-supplement.pdf
-
RC3: 'Comment on cp-2024-19', Joel B. Pedro, 24 May 2024
Berdahl et al. examine CESM2 model runs under pre-industrial and 127ka orbital parameters. The focus is mostly on the degree and mechanisms of ocean warming adjacent to the Antarctic ice sheet (AIS) in 127ka simulations with and without fresh water (FW) forcing. In particular, the paper proposes the hypotheses “that warming in the Southern Ocean, owing simply to the insolation anomalies during the LIG, may have been sufficient to cause substantial WAIS collapse”. This is certainly a worthwhile topic to explore and the paper raises some new and interesting insights.
In my view there are two major results.
- There has been a common view that North Atlantic FW forcing of the magnitude of a H event (ca. 0.2 Sv) drives warming against the Antarctic ice sheet. However, Bridahl et al. show modest ocean cooling against the AIS in the their 127ka FW simulation and give a sound treatment of what is driving the cooling. They also identify similar, largely overlooked, subsurface cooling in previous simulations where FW is applied to the NA.
- They identify modest sub-surface warming (0.4C), in the 127ka run without FW forcing and suggest that the warming is orbitally forced. Here, I was less clear on the specific processes driving the subsurface warming in the model and their link to orbital forcing.
In my view the paper makes some good and new advances on understanding of impacts of orbital and FW forcing on high southern latitudes. Congratulations on this large, detailed and well written study. I recommend publication after addressing the major and technical points below.
Major points:
- Some more attention is needed to clarifying significance of anomalies in the text and in figures, refer to technical points below for specific examples.
- Concerning analogs.. Line 366: “Given that we do not expect a large freshwater forcing of comparable scale to the H11 event in the future, our 127ka simulation, without freshwater forcing, may be a more relevant analog for the future”. I don’t think the analog point is valid as written here and elsewhere in the text (e.g. line 411). The future has different orbital parameters and CO2 levels to 127ka. Also, there is growing evidence that an AMOC collapse this century is plausible, due to NA warming and freshening (e.g. Ditlevsen & Ditlevsen, 2023) and previous modelling indicates that the far-field impacts of AMOC collapse are similar irrespective if the trigger is FW or some other mechanism (e.g. Brown and Galbraith, 2016). Also relevant to the analog argument is the different (and changing) thermal structure of the ocean now compared to 127ka. The impact of changes in SO windstress and upwelling on subsurface tempertatures near the AIS is sensitive to the temperature and properties of the upwelled waters, e.g. in the present climate where CDW is warming (e.g. Auger et al., 2021), there is evidence the poleward shifted westerlies are driving sub-surface warming, not cooling (e.g. Herraiz-Borreguero & Naviera Garabato, 2022). These factors need to be considered in commenting on the results and relevance of the current simulations for future AIS melt.
- Line 127 to 139: the text here begins to discuss but does not close out why global ocean heat content is higher in the 127ka simulation than in the PI. Similarly, Section 3.1.2. is compelling that there is subsurface warming close to the AIS in 127ka compared to PI, but I miss a clear explanation of what is driving the warming. Is it reduction in upwelling in the Tight SO? I think this is a major point because it goes to a key conclusion of the paper that its orbital forcing driving the sub-surface warming – some more explanation is needed to clarify the connection from orbital forcing to sub-surface warming near the AIS.
Technical Points:
- Global mean SAT in the 127ka is described as 0.004°C warmer than PI. Please clarify if this difference is significant using an appropriate test. I suspect it may not be, either way it should be made clear in the text. If the answer is that it’s not significant, then ‘marginally warmer’ should not be reported in the abstract. Instead report the significant seasonal temperature anomalies.
- Line 29: Maybe it’s a common view, but I don’t think it is canonical.
- Line 36: Give the estimated timing of the FW discharge. Also give the timing of the LIG sea level maximum (ca. 125ka?).
- Line 40: But you should also not that AMOC collapse has been predicted this century, and the far field impacts are similar irrespective of trigger. See e.g. Ditlevsen & Ditlevsen (2023) and Brown and Galbraith (2016).
- Line 55: Give the volume in m. sea level equivalent for context.
- The model set up and experimental design are appropriate and well described.
- Table 2, denote which anomalies are significant at 95%CI.
- Line 99: 3 to 8 thousand years before..
- Excellent figures throughout!
- Line 75 to 180 and elsewhere, clarity about significance of anomalies is needed. Probably all these anomalies are significant given the 100 year run times. But this needs to tested and stated. Similarly, Fig 4. marks significance with hatching for the snowfall rate but not for the other panels.
- Line 178. Figure 2 misses a. b. etc labels.
- Line 189, and similar to Major Point 3):. “As noted, the Southern Ocean generally warms as a result of the orbitally-forced 127ka”. If it's to be accepted that the warming is *the result* of orbital forcing then some more explanation and justification is needed closing the link between orbital forcing and the warming. If you can’t get to the bottom of what processes are driving the warming that’s ok, but it needs to be tackled head on (by comparison the explanation of how FW forcing drives cooling against the AIS is more thorough and complete).
- Fig 3. There is no colorbar here for the shading, presumably SST.
- Line 237: Quantify the cooling here.
- Line 248: all ocean basins accumulated heat..
- Line 249: a ca. 500-yr cooling..
- Mechanistically the reason for different responses in the SO and ‘Tight SO’ is that the Tight SO is south of the ACC, which is a barrier to meridional heat transport. This point is relevant to the definition and should be made in the text.
- Fig 5f and line 229: From the figure this looks more like no change in OHC than ‘a sustained reduction’. Can you demonstrate more clearly that this is a sustained and a significant reduction? The point is better made by Figure 6 and 8 than 5f. Also, from Fig 7 we see an initial cooling for ca. 500- 1000 years followed by gradual warming. Hence some more nuance is needed than ‘sustained reduction’ and ‘sustained cooling’.
- Line 237. Rather than ‘distinct cooling’ give the delta T.
- Fig 6. Are all shaded anomalies significant at 95%?
- Line 252-255etc: ‘After this multi-century cooling, MOTs rebound slowly for the remainder of the simulation.’ This is a clear description and should come earlier. Some of the earlier descriptions could even be trimmed.
- Line 261: Good point.
- Line 264: I don’t think it's appropriate to describe this as a ‘bulwark of cold subsurface water’; this is an anomaly with respect to 127Fw and not showing absolute water temperature and readers could miss the important distinction. Delete or revise.
- Lines 285 to 290 etc: Great analysis.
- Line 297: Good to hear. But the phrasing ‘is at least partly responsible for the slow rebound in the initial wind-driven cooling at depth near the AIS’ may be confusing. Do you mean that the eddy driven transport is at least partly responsible for the longer-term warming trend that replaces the initial wind-driven cooling? (I think it is explained more clearly later at line 360).
- Line 322: The argument here is that 127ka orbital forcing results in warmer subsurface temperatures against the AIS than we see in PI and therefore orbital forcing plausibly drove LIG mass loss from the AIS. An alternative way to test this idea would be to examine how subsurface temperatures evolve using a time slice before the LIG with respect to the 127ka run. This could diagnose if there is an orbitally driven sub-surface warming *trend* across the LIG. I don’t expect you to run new simulations, but please address this point in the text, and if your approach is the better way of diagnosing forcing by sub-surface warming then explain why.
- Line 375-380: Add a citation to Herraiz-Borreguero & Naviera Garabato, Nature Clim. (2022), which explores these factors in modern observational data, including the relationship between sub-surface warming and the strength and position of the westerlies. Their observations link poleward shifted (DJF) westerlies to *warming* of mid-depth CDW at depths and locations consistent with AIS ice melt. This appears opposite to the trend in subsurface temperatures in response to stronger westerlies in 127kaFW simulation. I expect this is because in the present day the increased wind stress is drawing up anomalously warm CDW. I think it’s worth to discuss this, given the papers intent that results are relevant for understanding future forcing on the AIS.
- Line 407: A key result that deserves a place in the abstract in my view.
- Line 411: Relevant to explore for process understanding, but I don’t think ‘analog’ is the right word here given the different orbital parameters, and CO2 levels and also the point above that AMOC collapse has been predicted by some studies this century.
- Line 415: model*s*, the only typo I found in the whole manuscript!
References
Auger et al., Nature Commun., 2021, https://www.nature.com/articles/s41467-020-20781-1
Brown and Galbraith, Clim. Past, 2016: https://doi:10.5194/cp-12-1663-2016).
Ditlevsen & Ditlevsen, Nature Clim. Change, 2023: https://doi.org/10.1038/s41467-023-39810-w
Herraiz-Borreguero & Naviera Garabato, Nature Clim. Change, 2022 (DOI:10.1038/s41558-022-01424-3)
Citation: https://doi.org/10.5194/cp-2024-19-RC3
Viewed
HTML | XML | Total | BibTeX | EndNote | |
---|---|---|---|---|---|
385 | 98 | 30 | 513 | 26 | 20 |
- HTML: 385
- PDF: 98
- XML: 30
- Total: 513
- BibTeX: 26
- EndNote: 20
Viewed (geographical distribution)
Country | # | Views | % |
---|
Total: | 0 |
HTML: | 0 |
PDF: | 0 |
XML: | 0 |
- 1