the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 20 May 2025
Response of Northern Hemisphere Rossby wave breaking to changes in sea surface temperature and sea ice cover
Abstract. Although well-researched in the present climate, it is poorly understood how Rossby wave breaking (RWB) may change in a warmer future climate. In this study, we examine how large changes in sea ice cover (SIC) and sea surface temperature (SST) affect the frequency and spatial distribution of Rossby wave breaking in the Northern Hemisphere during the boreal winter (December–February) and summer (June–August) seasons. Our experiment setup consists of eight 40-year atmosphere-only simulations from two models (OpenIFS and EC-Earth) that use different combinations of prescribed present-day and future SIC and SST values under the SSP5-8.5 scenario.
We find present-day RWB frequencies that correspond well with previous literature. Our models are generally in good agreement with regards to the spatial distribution of RWB. The effects of SSP5-8.5 SST on RWB are substantial, while simulations using future SIC and present-day SSTs do not exhibit statistically significant changes compared to the present. In simulations with SST changes, anticyclonic wave breaking (AWB) frequencies show large decreases during both winter and summer, while the primary change to cyclonic wave breaking (CWB) are small increases of varying magnitude in winter. The winter changes are notably collocated with changes in the strength and location of jet streams. The largest changes occur over the North Pacific, where winter AWB decrease by 60–70 % over the East Pacific and summer AWB decrease by roughly 50 % over the West Pacific and East Asia. Over the western North Atlantic, decreases of 10–30 % in winter AWB are collocated with a stronger eddy-driven jet, which may suggest an eastward shift in AWB. In summer, AWB decreases by about 50 % over North America but increases slightly over Europe. As with related previous studies of future changes in blocking and jet stream waviness, there are uncertainties in our results, and especially determining the impact of SIC changes likely requires longer simulations than those used in this study. This study demonstrates that particularly SST changes are an important component for changes to RWB in future climates.
- Preprint
(11814 KB) - Metadata XML
- BibTeX
- EndNote
Status: final response (author comments only)
-
RC1: 'Comment on egusphere-2025-2212', Anonymous Referee #1, 03 Jun 2025
Review of manuscript egusphere-2025-2212 by Tahvonen et al.
The authors investigate the effect of reduced sea ice extent and higher SSTs separately or together using global climate simulations from two models on the change in zonal wind and Rossby wave breakings. They find that the impact of warmer SSTs is stronger than the impact of reduced sea ice especially on wave breaking frequencies.
The study is interesting and the manuscript well written. Therefore, I recommend minor revisions. My comments follow in the order of the manuscript.
Minor comments:
- Lines 69-72: The authors could also mention that absolute vorticity fields on isobars have also been used to detect Rossby wave breaking (e.g., Rivière, 2009, and Barnes and Hartmann, 2012) as it is easier to get from climate models outputs.
- Lines 130-131: Concerning the simulations set-up: For the SIC_SSP585 simulation, what are the values of the historical SSTs where sea ice has disappeared? For the SST_SSP585, are the SST along the sea ice edge larger (due to future ocean warming) than in the historical simulation? In other words, is the surface temperature gradient much larger at the sea ice edge and if yes, what is the expected impact of this “artificial strong gradient”? Please add a bit more explanation on these points in the manuscript.
- Lines 174-178: What is the point of the DBSCAN step and how can it implies a discrimination with fixed strength and spatial extent ? Can’t simple criteria of strength and spatial extent be used instead?
- Lines 227-229: Could the authors explain in a few words the “false discovery rate control method” in the context of their study?
- Line 275: It is not clear to me that “over the central and East Pacific, the maximum in zonal wind appears to weaken and shift east”. Weakens, fine, but the maximum shifting east, how can you know?
- Lines 285-286: “the eastward shift of the local jet is slightly weaker”: What do the authors mean with “local jet”? And the shift is weaker than what? Please precise.
- Have the authors considered using the zonal wind at a lower pressure level than 250 hPa in order to see how the eddy-driven jet changes are linked with wave breaking changes? The zonal wind at 250 hPa sometimes mixes the sub-tropical jet with the eddy-driven jet such as the North Pacific and Western North Atlantic. It seems that previous studies found a better agreement between low-level (eddy-driven) jet changes and wave breaking changes. See, e.g., Rivière (2009), Fig. 10 in Barnes and Polvani (2013), who used a pressure-weighted average of the 850 and 700-hPa zonal wind, and the last supplementary figure in Michel et al. (2021).
- Lines 302-303: I find that the CWB frequencies changes in SIC_SSP585 look closer to the changes in SSP585 than in SST_SSP585. Don’t you think so?
- Lines 316-317: Could the authors rephrase this sentence? I do not understand it.
- Lines 380-382: Please precise the season in which there is “a reduction in the frequency of blocking”.
- Line 422: “by about .”: Please finish the sentence.
Technicalities:
- Line 11: the primary change → the primary changes
- Line 39: is → are
- Line 139: conditions to which → conditions which
- Line 153: modify → modified
- Line 281: there is are no … differences → there are no … differences
- Line 370: SSP585_SST → SST_SSP585
- Line 394: of → on
- Line 425: SSP should not be in italic.
Citation: https://doi.org/10.5194/egusphere-2025-2212-RC1 - AC1: 'Reply on RC1', Sara Tahvonen, 22 Aug 2025
-
RC2: 'Comment on egusphere-2025-2212', Anonymous Referee #2, 17 Jun 2025
Review of manuscript egusphere-2025-2212: “Response of Northern Hemisphere Rossby wave breaking to changes in sea surface temperature and sea ice cover” by Tahvonen et al.
In this manuscript the authors assess the impact of changes in STTs and Sea Ice on RWB frequencies in the Northern Hemisphere using an appropriate experiment design to test these. Using this, the authors show that the SSTs have a more profound impact on the jet streams in the NH but the results are inconclusive in as far as the impacts of sea ice is concerned because the models do not seem to agree.
The manuscript is succinct and it is well written. I have some comments that the authors should/could take into consideration before it is published.
Major comment:
The main concern that I have as a reviewer of this manuscript, as well written as it is and as impressive the model experiments as they are, is that it disagrees with established results with regards to the poleward migration of the jet and therefore the impact that this might have on RWB events. Whilst the DJF eastward shift in the jet is an interesting finding, is it realistic? Is it observed in the observations in the early 21st century? For instance, do we see this behaviour when we compare 1950 – 1969 vs 2000 – 2019 but weaker in strength.
Minor comments
Lines 25 – 35: Whilst one understands that the study focuses on AWB and CWB types, but why was the equatorward/poleward RWB neglected in this discussion. As we well now know, Thorncroft et al (1993) identified LC1 and LC2 which are equatorward but Peters and Waugh (1996) then showed that these two have poleward counterparts. This should at least be acknowledged here. Also, there are several studies which have shown that these 4 types actually exist in “observations” (reanalyses products).
Lines 40 to 45: The role of barotropic shear in influencing the morphology of RWB appears to be missing here.
Lines 45 – 50: COLs can also be viewed from the traditional synoptic meteorology point view (in geopotential height fields) and some studies have shown that RWB events precede COL formation. Perhaps the point of view of the synoptic meteorologist should be considered here and the role of the jet streaks that arise as the waves break and the transverse ageostrophic circulations that materialise here leading to vertical ascent. This is important to raise because there are several studies that have created climatologies of COLs from the traditional point of view and papers such as Wernli and Sprenger (2007), Portmann et al (2020) that have done the same from a PV perspective yet how these two meteorological worlds link is not considered in the lit review in this paper.
Lines 90: There are some studies in the SH that might be relevant here, even though they focus on the ozone depletion/recovery response during the summer season there (DJF) there.
Section 3.1: (I have some general comments for this section to help improve it)
- May I suggest presentation of the composites of AWB and CWB here to show that the categorisation method employed in this study actually work. These should include the isotachs please so that the climatologies in Fig 3 can then be better explained.
- I also strongly suggest that the authors consider presenting ERA-5 versions Figure 3, either as additional panels to that Figure or separately so that the models can be quantitatively compared with the “observations”.
- In the caption of Figure 3, please change the order in which AWB and CWB are presented and specify the years for these simulations
Lines 190 – 195; 210: I strongly suggest that the authors perhaps consider Peters and Waugh (2003) who looked at jet configurations in the SH to explain some of the morphologies of RWB events identified there. This might help to explain some of the location of the surf zones. For instance, in Fig 3f I see AWB events on the cyclonic barotropic shear side of the jet in the Pacific Ocean, which kind of goes against the grain, but if one considers the 10 m/s isotach once can see why that AWB centre is there. This seems to be easily explained by Peters and Waugh (2003), see their schematic in Figure 3.
Lines 240-245: Can the authors provide some reflection on the jet that does not seem to be migrating poleward with increasing GHGs. This issue should also be attempted to be addressed in current climate in reanalysis.
As discussion on the changes in the flow and its impacts during JJA shown in Figure 5 e – f is not provided in the manuscript or it is very thin and therefore needs to be given some attention.
Lines 325: The direction of eddy momentum fluxes should be explicit here (for in stance the fact that AWB is associated with poleward momentum fluxes)
Citation: https://doi.org/10.5194/egusphere-2025-2212-RC2 - AC2: 'Reply on RC2', Sara Tahvonen, 22 Aug 2025
-
RC3: 'Comment on egusphere-2025-2212', Anonymous Referee #3, 17 Jun 2025
Overview: This manuscript explores both how well two atmosphere-only GCMs capture Rossby wave breaking (RWB) in the historical period and how RWB surf zone will change in response to SSP5-8.5 forcing. They further break down the future change into changes broadly due to sea surface temperature (SST) changes and sea ice cover (SIC). They use a dynamic tropopause-based algorithm, which is the perfect application for a future changes paper given the projected varying changes in the height/pressure level of the tropopause in response to anthropogenic forcing. Their results show that the model reasonably replicates RWB with fidelity, and that future changes in RWB occurrence are far more sensitive to SST changes rather than SIC changes. The paper is well written and the figures are generally clear. There are some minor changes I’d suggest to the authors prior to final publication, but I commend them on a concise and clear study and manuscript.
General Comments:
- Discussion and differentiation of North Pacific and North Atlantic jet: I thought the authors generally did a good job of focusing on changes in the jet and RWB across the two dominant surf zones/ocean basins (North Atlantic and North Pacific). I felt that both the Introduction (eg. around lines 87-98) and Discussion would benefit from a bit more nuanced discussion of the differences between the jet mechanisms and interpretations across the two basins. For example, they made clear that the future changes for the two basins have different levels of confidence (eg. more confidence in jet shifts in the N. Pacific rather than the N. Atlantic) but didn’t get into much detail about why this is the case. Given that, particularly in the winter, the North Pacific jet often acts as a superposed subtropical and eddy-driven jet (that is more zonal in nature), while the North Atlantic is generally an eddy-driven jet (that tilts with latitude), I felt this discussion warranted a bit more careful detail on differentiating the two (and how proposed mechanisms may be impacted by the differences).
- Discussion of model experiments: I completely understand (and support) the author’s decision to reference the Naakka et al. 2024 paper for details on the experiments. This said, I think it would benefit the readers to have a bit more detail here on one particular part of the experimental design: How the SSTs were handled in the SICSSP585 experiment. Though one could dig into the provided reference for detail, I think it would be helpful to throw a sentence or two into your manuscript about how the SSTs evolved/were prescribed from the historical period (presumably under ice cover) when the model was in a reduced/removed ice scenario (seasonally dependent). I think this would help because it’s generally easy to visualize how future SSTs project (because it's provided in figure 1) but much harder to visualize what the baroclinic zones look like with future SIC but historical SSTs.
- Possible supplemental figures: It might be beneficial to try and create a set of maps that show the difference between the experiments and the full future model as a supplemental to help clarify some of the discussion. For example, a 6-panel that is essentially fig. 6 subtracted from fig. 4 (or fig. 7 – fig. 5), etc. I’m don’t think it’s necessary in the main body of the paper, but I think it would help clarify some of the differences you’ve discussed.
Specific comments:
- Lines 29-39: It may be helpful here to cite the recent work by Tamarin-Brodsky and Harnik (DOI: 10.5194/wcd-5-87-2024) in this section. It’s new (and understandable you didn’t have it here), but it’s a nice extension of the discussion you have here.
- Lines 48 (and elsewhere): In general, best practice for citations is to list in chronological order.
- Lines 64-66: I found the start of this sentence a bit hard to interpret – you may want to rework.
- Line 140: When you state that ‘this improves the detection …’, it’s unclear if the ‘this’ refers to the 4.4 K or 2 K. Also, a brief clarifier on how this improves signal detection would be helpful.
- Figure 1: Given that the focus of this paper is the Northern Hemisphere, it might be helpful to cut this figure to only the hemisphere of focus.
- Section 3.1: There were a few 1-2 sentence paragraphs here. I would consider trying to collapse these short ‘paragraphs’ into other ones. For example, lines 190-191 could be combined with the next paragraph, and the paragraph ending on line 221 could be combined with the paragraph starting at line 222.
- Lines 203-205: Does this result match to the climatologies of other studies?
- Line 231: I understand using the 250 hPa wind (a lot of studies do) – but if you have the dynamic tropopause wind in your dataset, why not use that instead to more perfectly match your RWB identification level? This in particular could be beneficial in the JJA analysis given the elevated warm season tropopause.
- Lines 237-239: I may have missed this later in the manuscript, but if you haven’t discussed a bit why this difference occurs between the two models, it would be helpful to do.
- Lines 242-243: It appears the zonal wind also has an equatorward shift (in addition to the eastward shift) for the North Pacific. This has ramifications for the occurrence of AWB (equatorward shift less AWB). It might also be linked to the enhanced CWB (either more CWB due to the equatorward shift, or an equatorward shift due to nudging via momentum flux by the CWBs).
- Line 272: I’m not entirely sure I see the increase in speed over East Asia – in particular in EC-Earth – but I do see the strong signal over the western North Pacific. I might suggest focusing more on that.
- Lines 273-274: It’s unclear to me whether this is actually an eastward shift in the exit region (the differences on the southern flank of the jet extend just as far east as the jet), but instead may be more due to a shift in the entrance region (eastward and southward) coupled with a more equatorward jet (which would reduce AWB). I think the eastward (and southward) shift in exit region is more confined just to the winter months.
- Lines 298-301: It may be beneficial to shift the discussion to changes in RWB in the peripheral Arctic seas, where we see the greatest shifts in RWB (ie. a localized changes). This also plays right into the discussion point on the extremely localized impacts of SIC changes on lower tropospheric temperatures, which may be impacting local baroclinicity, vertical wave propagation, etc. in just these regions.
- Figures 4-9: I had a hard time with the color bar for panels e and f. It might be beneficial to use a cold to warm color bar rather than a cold to cold bar. At times it was hard to identify increases or decreases.
- Lines 304-306: More an observation than something that needs changing (in particular given the lack of statistical significance) – it almost looks like a standing wave response in the summer CWB for the SIC experiment. Interesting given the importance of some of these seas for generating standing wave patterns that could be interpreted as RWB!
- Lines 325-328: I would lean more into the discussion on changes in the jet and causality here (and possibly in the introduction). There is an extensive body of literature exploring changes to the jet in a variety of models and experimental designs that would be beneficial to lean into here.
- Lines 372-373: Consider looking into Woollings and Hoskins 2008 (DOI: 10.1002/qj.310) here to link the weakened flow over the North Atlantic to benefiting CWB. Their study was for winter rather than summer, but there may be helpful information there.
- Lines 379 and 388: I found the starts of both of these paragraphs to be a bit informal – consider reworking.
- Lines 384-385: There are a few more things that play a role in blocking representation in models (eg. orography) – you may want to consider adding a bit more to the discussion here.
- Line 388: ‘contested’ is fine here, though I always find it make it sound a bit more negative in nature. Consider ‘the past and future trends of which are an area of diverging perspectives’.
- Lines 421: I’m not sure you can entirely say this first sentence with the experimental design. I think you’ve shown that SST changes, relative to SIC changes, are the dominant part of the total signal, but I’m not sure you’ve shown that the changes in boreal winter can be exclusively attributed to SST (and nothing else in the system).
Technical corrections:
- Handling of spaces after certain values – this might just be the format for this journal (every journal is different), but I found the gaps between degree symbols and directions (eg. 120° W) as well as the gaps between values and percentage signs (eg. 20 %) to be too wide/awkward in places given the text formatting. I would consider removing the spaces.
- Line 422: You’re missing the end of your sentence here.
Citation: https://doi.org/10.5194/egusphere-2025-2212-RC3 - AC3: 'Reply on RC3', Sara Tahvonen, 22 Aug 2025
Model code and software
Rossby wave breaking detection algorithm Sara Tahvonen https://doi.org/10.5281/zenodo.15357272
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 327 | 64 | 22 | 413 | 12 | 27 |
- HTML: 327
- PDF: 64
- XML: 22
- Total: 413
- BibTeX: 12
- EndNote: 27
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1