Future snowfall in the Alps: projections based on the EURO-CORDEX regional climate models

Frei, Prisco; Kotlarski, Sven; Liniger, Mark A.; Schär, Christoph

doi:https://doi.org/10.5194/tc-12-1-2018

Articles | Volume 12, issue 1

https://doi.org/10.5194/tc-12-1-2018

© Author(s) 2018. This work is distributed under
the Creative Commons Attribution 3.0 License.

https://doi.org/10.5194/tc-12-1-2018

© Author(s) 2018. This work is distributed under
the Creative Commons Attribution 3.0 License.

Articles | Volume 12, issue 1

Research article

|

05 Jan 2018

Research article |

| 05 Jan 2018

Future snowfall in the Alps: projections based on the EURO-CORDEX regional climate models

Prisco Frei, Sven Kotlarski, Mark A. Liniger, and Christoph Schär

Download

Final revised paper (published on 05 Jan 2018)
Supplement to the final revised paper
Preprint (discussion started on 27 Feb 2017)
Supplement to the preprint

Interactive discussion

Status: closed

AC: Author comment | RC: Referee comment | SC: Short comment | EC: Editor comment

- Printer-friendly version

- Supplement

RC1: 'Review', Anonymous Referee #1, 05 Mar 2017
- AC1: 'Replies to Referee #1 and proposed changes to manuscript', Sven Kotlarski, 30 May 2017
RC2: 'Review of “Snowfall in the Alps: Evaluation and projections based on the EURO-CORDEX regional climate models” by Prisco Frei, Sven Kotlarski, Mark A. Liniger, Christoph Schär', Anonymous Referee #2, 27 Mar 2017
- AC2: 'Replies to Referee #2 and proposed changes to manuscript', Sven Kotlarski, 30 May 2017
RC3: 'Review original manuscript', Anonymous Referee #3, 05 Apr 2017
- AC3: 'Replies to Referee #3 and proposed changes to manuscript', Sven Kotlarski, 30 May 2017

Peer-review completion

AR: Author's response | RR: Referee report | ED: Editor decision

AR by Sven Kotlarski on behalf of the Authors (12 Jun 2017) Author's response Manuscript

ED: Referee Nomination & Report Request started (15 Jun 2017) by Xavier Fettweis

RR by Anonymous Referee #2 (04 Jul 2017)

Suggestions for revision or reasons for rejection

Dear Authors,
the revised manuscript has been improved a lot, all the major points have been adequately addressed in the reply to the Reviewers and eventually integrated in the main text. After the correction of the few typos listed below, I think this version of the manuscript is suitable for publication.
Congratulations for this nice paper!
Kind regards
*******************
Technical corrections:
Page 3
Line 96: throughout the annual cycle -> throughout the year
Line 105: “on an RCM grid” -> on “the” RCM grid

Page 4
Line 129: “the very focus” -> “the focus”
Line 130: “on the snow projection aspect” -> “on snow projections”
Line 138: “To estimate observation-based snowfall” sounds like an oxymoron, I suggest: “To estimate the reference fine-scale snowfall”

Page7
Line 230: “not all RCM simulations available through EURO-CORDEX provide raw snowfall” -> “not all EUROCORDEX RCMs provide raw snowfall”

Page 12: missing reference to Figure S5.

Page 13, Line 452: “partly explain the partly substantial overestimation”

Page 19, Line 713: “In” -> in

Page 20:
Line 731: “the an”
Line 735: “on an” -> “on the”

Page 21
Line: 783: “This effect … may be strong under conditions (depending upon location and season) where the current climate is well below freezing. Such conditions may experience a shift towards a temperature range more favourable to snowfall”. Maybe the word “conditions” is not the best choice here, can you please rephrase?

Figure 2: the caption is unclear, you could rephrase “Ratio of a) mean snowfall Smean, b) heavy snowfall Sq99 obtained with the Binary and the Richardson methods to the corresponding values obtained with the Subgrid method …”

Hide

RR by Anonymous Referee #3 (12 Jul 2017)

Suggestions for revision or reasons for rejection

MS No.: tc-2017-7, revised manuscript
Review, 12 July 2017

Title: Snowfall in the Alps: Evaluation and projections based on the EURO-CORDEX regional climate models

Authors: Prisco Frei, Sven Kotlarski, Mark A. Liniger, Christoph Schär

Recommendation: [Revision]

GENERAL COMMENTS:

In my review of the original manuscript a couple of months ago I raised a couple of major points which in my opinion needed attention. Having read the revised manuscript and the accompanying rebuttal letter I notice that the authors have done a considerable effort to address all issues, also those from the other reviewers. Overall, the revisions have resulted in an improved manuscript.

Unfortunately, not all comments have been adequately addressed, therefore some further attention is needed before the manuscript is suitable for publication.

MAJOR COMMENTS

1. (also point 1 in first review) The sentence added by the authors to point out that some works also utilized output from high resolution (12km) RCM output “It thereby complements …(partly originating from EURO-CORDEX) … but with a reduced ensemble size and/or not specifically targeting the entire Alpine region” I don’t understand both phrases.
Firstly, EURO-CORDEX has a marginal role in both papers, and certainly nothing of the material and methodology originates from it. Just leave it out.
Secondly, these are both quite different studies, the Piazza et al. paper studies a variety of models in their reponse for a region in the French Alps., while the de Vries et al. paper utilizes a single-model 8-member ensemble considering a region including the entire Alpine region. I would advise to just explicitly phrase in what these papers primarily differ from your paper (i.e. region of analysis in Piazza et al. paper, and nature of model ensemble studied in de Vries et al.) and avoid using “and/or” constructions

2. (point 3 first review). I do not disagree that HadGEM2-RACMO has an issue with snow accumulation over the Alps, but this is rather the manifestation from a missing process (no glacier model) in combination with not setting an artificial limit on maximum snow depth (as is done in some of the other RCMs), than too much snow fall. Very large numbers in snow depth (or snow mass if available) are also seen in the model combinations EC-EARTH-HIRHAM, EC-EARTH-RACMO, and IPSL-WRF (in the first two even larger than in HadGEM2-RACMO), yet these combinations have been retained.
The authors furthermore mention in their rebuttal that the “obvious feedbacks on temperature” led them to disregard the HadGEM2-RACMO simulation. “. I would expect this to hold primarily for the summer season, and far less relevant in the winter season when sub-zero temperatures still dominate at the higher elevations where the anomalously high snow accumulation is found.
But even when it also affects the September-May season studied by the authors I don’t’ see why the other three model combinations, mentioned above, don’t suffer from this problem. After all, the sensitivity to temperature bias is to a large extent taken care of by the (bias-adjustment) shift in snow fractionation temperature T* to Ta*.
In my opinion the reason to disregard HadGEM2-RACMO is formulated rather ad hoc and could have been equally well been applied to disregard other model combinations which have now been retained. That leaves three options: 1) include HadGEM2-RACMO in the analysis, 2) disregard the other three combinations as well, or 3) provide a plausible explanation why this specific model combination was disregarded and other combinations not.

3. (point 7 first review).The way the authors have addressed this point only makes things worse. I repeat that the line “Previous studies … with this theory (e.g. Allen and Ingram, 2002; Ban et al. 2015)” is out of context, for reasons the authors agree on but refuse to act upon. What the authors do is adding a phrase which underlines that the material in their paper (daily scales, no convection, cold season) is incompatible with the conditions studied in the Ban et al paper (sub-daily or hourly scales, convection, warm season), yet they state that there finding is consistent with the Ban et al..
An additional reason why I am emphasizing this point is that the literature on the projected behavior of hourly precipitation extremes under climate change is quite controversial, and far from settled at the moment (see also literature cited in Ban et al. 2015). That aspect is entirely ignored in the current formulation, which is unacceptable.
I therefore stick to my point made in the first review that this sentence including the corresponding references must be taken out.

MINOR COMMENTS:

1. (point 28 in the first review) I meant to add an ancillary line showing the frequency of occurrence of the elevation intervals across the region of interest (either Switzerland or the Alpine region or both). Evidently the higher elevations are (far) less abundant than the lower elevations – there will be only very few grid cells with elevation beyond 3000 m a.s.l. , making the analysis for high elevation less robust. It would help to give an impression of the pdf of the elevations, you could use the upper horizontal axis to label the percentage. The attached png-file shows what I meant to say (only intervals above 250m are counted), the authors could plug in such line in Fig 3a, or, if they prefer in Fig 2, or Fig 4a, that is for them to choose.

2. (major point 4 first review) Indicate in Table 1 which EC-EARTH realizations are used (like you do for MPI-ESM-REMO) in the various EC-EARTH-RCM combinations: r1i1p1 for EC-EARTH-RACMO; r3i1p1 for EC-EARTH-HIRHAM; r12i1p1 for EC-EARTH-SMHI and EC-EARTH-CCLM.

3. Line 267: “leads a”  “leads to a”

4. Line 300: rephrase “and as we condere σh as a” as “while σh is regarded as a …”

5. Line 328: rephrase “under current and future … climate conditions” as “under changing … climate conditions

Referee Report: PDF

Hide

ED: Reconsider after major revisions (13 Jul 2017) by Xavier Fettweis

AR by Sven Kotlarski on behalf of the Authors (07 Sep 2017) Author's response Manuscript

ED: Referee Nomination & Report Request started (08 Sep 2017) by Xavier Fettweis

RR by Anonymous Referee #3 (02 Oct 2017)

ED: Publish subject to minor revisions (Editor review) (12 Oct 2017) by Xavier Fettweis

Dear authors,

Your responses to the 3 major points raised by the reviewer #3 was convincing for him except that he remains not happy with your choice of the RCMs used in your ensemble mean. His arguments to discredit your selection are convincing (in particular about your choice of showing the summer season only in your Fig R. 1 and not the winter season ?) and, according to his Figure B, it is true that other models (e.g. RCA4) could be also removed in your selection for the same argument. However, your selection seems for me to be honest (as you have not interest to exclude some models I think) and using the same selection than CH2018 is a good point in favour of your choice of models. For this reason, I have decided to accept your paper for publication without additional review round (except by me).

However, coming back to your initial selection of RCMs, as suggested by reviewer #3, is for me the best knowing that the addition of unselected models does not change the main results of your paper. When I compare Figs 9-13 using an ensemble of 12 models vs the ones using an ensemble of 14 models, the results are quasi identical for me.

Therefore, I give you two options:

1. keeping the CH2018 selection in your manuscript but you need to provide in Supplementary, robust and scientific proves of excluding several models (why RACMO and not RCA4 ?) and to provide in Supplementary the same results/fig using you initial selection of RCMs by clearly mentioning in your manuscript that several RCMs has been discarded (by listing them) and that the same figures than Figs. 9-13 are provided in Supplementary using this time all the RCMs.
2. coming back to your initial selection of RCMs and providing in Supplementary the Figs. 9-13 using the CH2018 selection. As the results are the same (for me), this shows that the use of models with biases (e.g. RACMO) does not impact on your results and increases the robustness of your results.

I am clearly in favour of showing more models (as you have already preprocessed these outputs) but I am opened to Option 1 if you have good reasons.

Best regards,
Xavier Fettweis

Hide

AR by Sven Kotlarski on behalf of the Authors (27 Oct 2017) Author's response Manuscript

ED: Publish as is (02 Nov 2017) by Xavier Fettweis

AR by Sven Kotlarski on behalf of the Authors (10 Nov 2017) Author's response Manuscript

Short summary

Snowfall is central to Alpine environments, and its future changes will be associated with pronounced impacts. We here assess future snowfall changes in the European Alps based on an ensemble of state-of-the-art regional climate model experiments and on two different greenhouse gas emission scenarios. The results reveal pronounced changes in the Alpine snowfall climate with considerable snowfall reductions at low and mid-elevations but also snowfall increases at high elevations in midwinter.