Review of the manuscript

The future North Atlantic jet stream and storm track: relative

contributions from sea ice and sea surface temperature changes

Daniel Köhler, Petri Räisänen, Tuomas Naakka, Kalle Nordling, and Victoria A. Sinclair

submitted to WCD (WCD-2024-3713)

Using coordinated simulations with four different global atmospheric models, 

the study explores the roles of future sea surface temperature changes and 

sea-ice loss for changes of the wintertime North Atlantic jet streams and 

storm tracks. 

The simulated changes are extensively described with a focus on the 

dynamical drivers of the North Atlantic jet stream changes, namely 

the changes in the baroclinicity and momentum flux convergence.

However, differences between the model responses with respect to the 

position and strength of jet stream and storm track shifts, point to 

uncertainties and leaving the European climate's future response uncertain.

Given that there is still a debate within the scientific community about 

the impact of Arctic sea ice loss in the mid-latitude westerlies and storm 

tracks (with the possibility that Arctic sea ice loss weakens the 

North Atlantic jet stream  favoring more severe cold winters), the topic 

of this study is timely and relevant. As reasons for the disagreement in 

the response to sea ice loss found in modelling studies include differences 

in the forcing fields (pattern and magnitude of forcing), coordinated 

experiments are an appropriate approach to tackle this question.

The Polar Amplification Model Intercomparison Project (PAMIP, Smith et al., 

2019) as part of CMIP6 provided a large set of coordinated experiments 

which allows to study the relative roles of local sea ice and remote sea 

surface temperature changes in response of the global climate system  

to changes in Polar sea ice.

This study is in my view complementary using a smaller set of models and 

different forcing fields.

The manuscript is well-structured, but especially the part describing the results 

is lengthy, and, at some places, provides too much details. 

In comparison, the discussion and conclusion part is too short, and 

misses links to relevant other studies.

Overall, the submitted manuscript needs careful and major revision. 

---------------------------------

Major comments:

(1) I appreciate the careful description of the figures in the results sections. 

Nonetheless, I recommend a shortening of the description part of the manuscript, 

especially sections 4 to 6, to allow for clearer main messages. In addition, 

I strongly recommend more discussions of the results in comparison to 

other studies. This could be done either in the respective sections or 

as an additional section.

As one example, I would like to mention the role of deep Arctic warming 

for mid-latitude atmospheric circulation changes which is discussed e.g. 

in Cohen et al. (2020), Xu et al. (2023), Kim et al. (2021).

Since the model results presented here do show significant differences 

in the vertical extent of Arctic warming in response to se-ice loss (figure 10),

the results obtained here have to be discussed in the context of other 

studies which will certainly provide interesting insights!

Such discussion are needed also for the other interesting results to place them 

in the context of other studies.

(2) Based on the above mentioned extended discussions, the conclusions in section 7 

have to be improved to highlight the new insights from this study and to elaborate 

on future research.

(3) In the introduction, the benefits of coordinated model experiments should be 

discussed  more in depth, in particular the PAMIP approach (Smith et al., 2019), 

and the differences to the approach applied in this study. It would be also good to

explain a bit more the role of the coordinated model experiments for the whole

EU project CriceS.

(4) The relative small sizes of the model runs with 40 years and what does this 

mean for the robustness of the results should be more critically discussed, 

given that, e.g. Peings et al. (2021) showed that even 100 years might be not 

enough to capture the remote atmospheric response to future Arctic sea ice loss 

against the large internal variability.

(5) The applied statistical testing has to be critically reviewed. For all testings, 

a two-sided t test has been applied. For some of the shown distributions (e.g. 

ETC life time in fig. 4b) it is obvious that the assumptions for the t-test, in 

particular normally distributed samples, are not fulfilled. 

For testing differences in fields, the tested fields are highly 

spatio-temporally correlated. Thus the t-test has to be controlled for 

field significance (e.g. by applying the false

discovery rate (FDR; Wilks, 2016). 

(6) I recommend an extended descriptions of the model experiments. I recognized that 

more information is given in Naakka et al. (2024), but it would help the reader to have 

more information also in this manuscript. For the description of the experiments, a table 

would be good. Since the model simulations are atmosphere-only simulations, I suggest to

mention the similarities in the models, too:

e.g. NorESM atmospheric component is based on CESM atmospheric component, and EC-Earth 

atmospheric component is based on an earlier version of OpenIFS. This should be also 

included into the discussion of the results.

----------------------------------------------------

Minor comments:

(1) Introduction, L36-37: References are missing.

(2) Section 2, L112-115: Beside the E-vector, the Plumb flux (Plumb, 1985)and the 

localized EP flux (Trenberth, 1986) allows local diagnosis 

of the three-dimensional propagation of wave activity. Could you, please comment, why 

you have chosen the E-vectors?

(3) Section 2, L123-124: For frictionless motion!

(4) Section 3.1, L164-165: I would appreciate to see a corresponding figure to show the 

differences in zonal wind and baroclinicity to ERA5, e.g. in the appendix. 

(5) Section 3.2, L199-200: Please explain, why the focus is on the jet exit region? 

(6) Section 3.2, L220: Isn't the North-Atlantic jet an eddy-driven jet in all models?

(7) Section 3.3, L223:  "ETCs affecting Europe (30° N - 70° N, 15° W - 35° E)," Does that 

mean those ETC's which have their endpoint in that area?

(8) Section 5.1, L353: "to increase the jet speed located above the maximum in the Baseline simulation"

For CESM, the increase in the jet speed is clearly southward of the maximum in the Baseline simulation, 

which is not so obvious in the other models (Fig. 8g). Could you, please comment on this?

 

-------

Kim, D., et al. (2021). Atmospheric circulation sensitivity to changes in the vertical

structure of polar warming. GRL, 48, e2021GL094726.

https://doi.org/10.1029/2021GL094726

Xu, M., et al. (2023). Important role of stratosphere-troposphere coupling in the 

Arctic mid-to-upper tropospheric warming in response to sea-ice loss. npj Clim Atmos Sci 6.

https://doi.org/10.1038/s41612-023-00333-2

Cohen, J., et al, (2020). Divergent consensuses on Arctic amplification influence on midlatitude 

severe winter weather. Nat. Clim. Chang. 10, 20-29 (2020). 

https://doi.org/10.1038/s41558-019-0662-y

Smith, D. M., et al. (2019). The Polar Amplification Model Intercomparison Project (PAMIP) 

contribution to CMIP6: investigating the causes and consequences of polar amplification. 

Geosci. Model Dev., 12, 1139-1164.

https://doi.org/10.5194/gmd-12-1139-2019 

Wilks,D. (2016). The stippling shows statistically significant grid points-

How Research Results are Routinely Overstated and Overinterpreted, and What to Do about It. 

BAMS 97, 2263-2273.

https://doi.org/10.1175/BAMS-D-15-00267.1

Peings, Y., et al. (2021). Are 100 ensemble members enough to capture the remote atmospheric 

response to+ 2° C Arctic sea ice loss?. Journal of Climate, 34, 3751-3769.

https://doi.org/10.1175/JCLI-D-20-0613.1

Plumb, R. A. (1985). On the three-dimensional propagation of stationary waves. 

J. Stm. Sci., 42, 217-229.

https://doi.org/10.1175/1520-0469(1985)042<0217:OTTDPO>2.0.CO;2

Trenberth, K. E. (1986). An assessment of the impact of transient eddies on the zonal 

flow during a blocking episode using localized Eliassen-Palm flux diagnostics. 

J. Atm. Sci., 43, 2070-2087.

https://doi.org/10.1175/1520-0469(1986)043<2070:AAOTIO>2.0.CO;2 

Summary: This paper examines the North Atlantic and European response of the jets and storm tracks across four atmospheric GCMS with sea ice, SSTs, or both together, as the surface forcing. The differences across the four models are discussed for their present day climatologies and for end their 21st century climate. The relative contributions of each SST and SIC make to the full response is examined. One model is an outlier for many of the responses.

Overview: the paper is mostly well-written and clear, but is also too long and wordy in places. The figures are of good quality. I find the mechanistic approach to understanding inter-model differences to be illuminating and a real strength of this work compared to past ones. However, I have some concerns that the differences being discussed are an artefact of the short time period being used (39 winters), and the potential impact of seasonal-to-decadal atmospheric variability hampering a meaningful comparison between the models, as well as in their responses. As the experiments themselves are not the topic of the paper, I'll restrict my comments to suggestions for ways that the analysis presented and discussion around it could be framed to create a stronger paper. For example, there is too much detail in the results sections and the end result is that the reader is left without a clear view on what are the important results. I'd suggest reframing the results around a few salient points rather than listing off all the results. Another suggestion is to show inter-model differences in the figures themselves,  alongside significance of the difference. This way only relevant differences, even if they are influenced by variability due to the short time series, are discussed. 

Specifics:

There are some references missing from the paragraph around line 25-30. E.g. Ronalds et al 2019, Hay et al 2022

L34: McKenna paper uses a more primitive model (intermediate complexity) which you might want to make clear, as it's not directly comparable with GCM results.

Paragraph beginning L45 and Section 2.1: As this is not a new approach, there are missing references here, from early work on sea-ice loss AGCM experiments (e.g. Deser et al 2010), to similar but coupled experiments from McCusker et al 2017 and Oudar et al 2017. Finally, how these experiments differ from PAMIP and what is added by performing these needs more discussion. Here is where the discussion on how the short time series relative to noise and variability could be important. What is the change in surface mean temperature, SSTs, sea-ice loss and pattern of loss, and could these be relevant for understanding differences with PAMIP?

L85: If you used ssp126, could you scale to combine and get 78 winters to boost sample size?

Combine Sect 2.2 with 2.3 as it isn't really used as a standalone metric.

L126: What is meant by this last sentence?

Sect 2.5 could be combined with 2.1, and I'd leave the NL term out and just discuss it where relevant. There is also an inconsistent use of the nomenclature used here and terms like 'Future' and I'd suggest ensuring consistent use of one or the other throughout.

Sect 3.3: are the ETC metrics calculated across the entire track of cyclones passing through the box, or just for the portion of the track within the box? It could change interpretation of the results. 

Sect 4.2 When discussing the contribution of SIC or SST to FT, taking a pattern correlation and then giving a % spatial variance explained by each can give a quantitative way of expressing the relative importance of each.

Sect 6.2 The non-linearity of the ETC response is quite interesting and a novel result. I'd suggest spending a bit of time discussing it.

Throughout: Discussion and placing results within context of existing literature is lacking.

Details/typos:

L1 with -> from

Last two sentences of abstract are kind of saying the same thing

L17 & elsewhere: majorly is an informal/slang word and shouldn't be used in a scientific paper

L17: enhancing -> enhanced

L22: Move 'together' to after 'jet stream

L31 remove 'and' before necessitating

Starting at L36 there seems to be a new paragraph as it is unrelated to part before.

L47: with-> from

L70 repeating 'atmosphere'

L101-103: This sentence is hard to understand

L124: 'are interpreted by the zonal mean denoted' -> 'in the zonal mean, denoted'

L140: was computed -> is computed

L190:is EC-Earth3 here a typo?

L211: You used both variable names and acronyms here and elsewhere, which makes the paper longer.

L227: remove last sentence

L243: 'significantly the most equatorward' is awkward wording

L270: as EC-Earth -> to EC-Earth

L287: Repeating OpenIFS (Perhaps you might want to shorten the four variables name you use, for readability)

Section 5: I'd remove the whole paragraph to shorten the paper.