PAPER SUMMARY AND RECOMMENDATION

This study looks at how Greenland's peripheral glaciers may change in the future due to climate change. It emphasizes the distinction between solid ice discharge (calving of icebergs) and liquid freshwater runoff (melting and rain). The study employs the Open Global Glacier Model (OGGM), which simulates the evolution of glaciers using climate data from the Coupled Model Intercomparison Project Phase 6 (CMIP6) under four emission scenarios. Key findings include:

• Projected declines in area and volume: By 2100, the glaciers are projected to lose 19-44% of their area and 29-52% of their volume, contributing 10-19 mm to sea level rise.
• Shifting freshwater contributions: Projections show solid ice discharge decreasing over the century, especially after 2050, as marine-terminating glaciers retreat inland
• Peak water timing: Peak water, the time of maximum annual freshwater release, is projected to occur around 2050 under low emissions (SSP126) and shift to around 2082 under high emissions (SSP585)

The article explores an innovative and relevant topic for The Cryosphere and is well-written and easy to follow. The approach is sound, and the study’s core findings offer insights that will be beneficial for future research efforts. To further strengthen the manuscript, I recommend that the authors incorporate some additional analyses and provide further clarifications. These improvements will help make the article ready for publication.

MAJOR COMMENTS

1. Despite the improved modeling of marine-terminating glaciers, it would be useful to include a comparison with outputs from multiple global models (e.g., PyGEM, GloGEM, OGGM v1.6). Although these model results are briefly discussed, I suggest including a summary table with results (e.g., volume and area) and comparing them statistically (methods in Section 2.5). This comparison is valuable given the similarity of the GCMs (L114).
2. Section 2 could benefit from a clearer separation of data, models, and calibration methods. For instance, the climate section mentions a precipitation factor without explaining its calibration. The OGGM model section includes all data related to glaciers, so model calibration could be placed in its own subsection under modeling.
3. Reproducibility: There is no mention of the version of OGGM used in this study. According to the documentation, the most recent version was not used due to issues with the proposed frontier ablation (https://tutorials.oggm.org/stable/notebooks/tutorials/kcalving_parameterization.html). It would be helpful to specify the version used.
4. Oceanic Forcing: The study frequently notes the lack of oceanic forcing in the OGGM model, but the potential implications of this limitation are not discussed. I recommend that the authors explore how incorporating oceanic forcing might affect their projections, particularly for solid ice discharge and regional variability. Could the current projections underestimate or overestimate the actual contribution?
5. Regional Variability: The study mentions significant regional variability in projected glacier loss, sea level rise contributions, and freshwater runoff. I encourage the authors to expand their analysis of these regional differences, examining how factors such as glacier size, geometry, elevation, and local climate contribute to the observed variability.
6. Figures Quality: Please check the quality of the figures, particularly Figures 3 and 4. It is unclear whether the glaciers are represented as points, polygons, or grid interpolations. Additionally, several panels show mismatched axis ranges, which makes the analysis difficult to interpret. Consider using color palettes that are more accessible for colorblind readers (e.g., ColorBrewer palettes: https://www.datanovia.com/en/blog/the-a-z-of-rcolorbrewer-palette).
7. Precipitation Factor Calibration: The manuscript does not explain how the precipitation factor is calibrated, which is important for analyzing freshwater contributions. Please consider adding a table listing all parameters, their values, and any calibrations involved (Section 2.4).

MINOR COMMENTS

1. Abstract: Clarify how the values (e.g., 19-44%) were calculated—are they multi-model means or medians for extreme scenarios?
2. Fig 1: Define "MT glaciers" and consider adjusting the size or transparency of the triangles for better clarity.
3. Fig 1c: Change "Numbers" to "Number of glaciers" and verify the caption.
4. L125: Please elaborate on the delta method and explain how anomalies are calculated.
5. L234: Could one variable be analyzed using different methods? Please clarify.
6. Fig 3: Was smoothing applied to the confidence intervals? Consider sharing the vertical axis across subplots to increase the size of the lower subplots. This applies to subsequent figures as well.
7. Section 3.2: The section heading suggests a comparison, but there is little temporal and spatial comparison between variables. Please expand this comparison.
8. References: Some references are listed with full journal titles, while others use abbreviations. Please ensure consistency throughout the reference list.
9. Data: Clarify whether the data on Zenodo is available at the subregional scale or glacier-ID scale.

General:

The study projects the future glacier evolution of peripheral glaciers of Greenland with focus on glacier area, volume and runoff evolution. The main goal is to understand how the freshwater contribution from these glaciers will change over the 21^st century. Therefore, the authors differentiate between solid ice discharge at marine-terminating glaciers, and liquid glacier runoff. The runoff is further differentiated between glacier meltwater, rain and melting of snow.

The motivation and relevance of this study are well outlined in the introduction. The study is well structured over largest parts. Overall, the study adds valuable knowledge to the understanding of changes in glacial freshwater contribution in the 21^st century in Greenland. I have one larger and two smaller major comments and several specific comments which need addressing and revision before publishing the article. I would rate the article as in between major and minor revisions, depending on the effort taken to address major comment 1, which is why I advise reconsidering after major revisions.

Major comments

1. Calculation of liquid runoff composition

One major goal of this study is to analyze the liquid glacier runoff composition, and how it is changing over the 21^st century. To do so, you consider an area given by the RGI outline of the respective glacier. When the glacier is retreating, glacier-free areas are increasing. This way, it is not surprising that the contribution from off-glacier snow melt will increase from <1 % to 15 % at the end of the 21^st century since the off-glacier area was basically non-existent at the beginning. With this approach, you are first comparing the runoff of a (quasi) glacier-only area with the runoff of a mixed are, which is not consistent. To keep consistency and really allow a comparison of how the runoff composition is changing, I would advise to include the entire hydrological catchment in the runoff calculation, not only the glacier outline. This way, you could give a realistic analysis of the runoff composition entering the ocean and on its evolution over the 21^st century.   

Hydrological catchments are available globally from the Hydrobasins dataset on different levels (https://www.hydrosheds.org/products/hydrobasins).

Lehner, B., Grill G. (2013). Global river hydrography and network routing: baseline data and new approaches to study the world’s large river systems. Hydrological Processes, 27(15): 2171–2186. https://doi.org/10.1002/hyp.9740

2. Lack of detail in results and discussion

The results and discussion section lack detail in my opinion. The analysis and discussion of differences across Greenland and the forcing behind these differences stay very short and superficial. I would suggest looking into these subregional differences in more detail. I think there is potential for many more interesting results from such an analysis.

For example, the evolution of Flade Isblink Ice Cap is very stable for all scenarios, which is not mentioned in the results section and not discussed in the discussion section, although this area is pointed out in the methods part. Do you have an explanation for this behavior?

There are also subregional differences in how big the spread between the scenarios is, indicating that for some subregions the emission path is even more important than for others (e.g., Central-East has a very large difference between SSP126 and SSP585). Why is this the case?

3. Figure quality

The formatting of the figures needs to be significantly improved. Respect consistency in style and font size, and arranging and dimensions of subpanels. The subpanels of the subregions in the line plots are not in the same order as in the pie diagrams, which could be changed for consistency. Specific details per figure are given below.

Specific comments

L57:     What does “a significant mass loss process” mean? Please add a number from these two references that you refer to, stating how much it is roughly.

L70:     Why would glaciers shift to a more cold-based regime in a warming world? Please explain this contra-intuitive statement briefly.

L74ff:  I would like to read a bit more details about the methods and the approach in this paragraph. For example: What kind of data is the cited frontal ablation dataset, is it observed or modelled? Which projections are being used (low emission to high emission scenarios)? What is the covered period of the geodetic mass balance? Just add a few details so that readers, who don’t want to read the entire paper, get the most essential information from this paragraph.

L92:     How did you subdivide the drainage basins of FIIC? Are there no subdivisions at all in the RGI6 inventory, or different ones? If the latter is correct, could you show the RGI subdivision in Fig. 1b?

L94:     What do you mean with “active calving basin”? Are there also “inactive calving basins”? If so, how do you differentiate?

L105:   You write that you apply a precipitation correction to the ERA5 data. Based on which data do you do this correction? Do you use observations for that?

L113ff: I do understand your argument that using the same GCMs as previous studies makes comparison and interpretation of results easier. However, those other studies all had a global (or entire northern hemisphere) focus while this study has a clear focus on Greenland. Therefore, it is important to apply GCMs that are well suited for Greenland. How do you guarantee this?

L125:   Can you explain this in more detail. GCM data is bias-corrected towards the ERA5 data, which has also been corrected somehow (see comment to L105).

How is the climate data being interpolated over the topography coming from the GCM resolution?

L147f:  What is a “resolution suitable for the glacier size”? Please explain.

L161ff: The first two sentences of this paragraph rather describe the creation of the climatic input dataset. Therefore, I would move it so section 2.2.1. What lapse rate are you using for the temperature data? Is it a constant value, is it different per grid point but constant in time, or is it fully variable?     

L171:   Why is there another precipitation correction applied? Do I understand it correctly that first the ERA5 data is corrected, then the GCM data is corrected to the ERA5 data and finally the precipitation is corrected a third time with this precipitation factor? If so, why do you do step 1 and 2 at all? Please explain. Also see comments to L105 and L125.

L197ff: You describe the “annual freshwater runoff from the glacier” but include off-glacier snowmelt and rain. This sounds contradictory.

L204ff: If you want to analyze the total freshwater runoff, including areas where the glacier has melted of, you have to include the runoff not only from the RGI outline but from the entire catchment. Your approach right now is inconsistent, where you first ignore the not-glaciated areas at the beginning of the simulation, but then include the areas where the glacier has melted off during the 21^st century. This way, it is also not clear what kind of peak water you are calculating. It is not the peak water of glacier runoff, as you also include off-glacier areas. But it is also not the peak water of the entire catchment, as you exclude areas that had been glacier-free before the start of your simulation.

L222:   The title of this subsection does not fit well in my opinion, as you are mainly describing the significance tests. Consider reformulating.

Fig. 3:  (a) The FIIC not losing area for all four SSPs. Why is this the case?

L253:   Add the number of volume loss for the Central-West here.

L256:   (And all similar text passages before and after) Are the exact numbers to the ANOVA really important to be communicated in the text? I am not familiar with this particular significance test, thus maybe I am wrong. But I would suggest defining once (maybe in the methods section 2.5) how you define a test to be significant (what level etc.), then stick to this definition throughout the rest of the manuscript and just say that it is a significant result (or not). This way, you wouldn’t have to give these numbers every time you mention a significant result, which would improve the text flow in my opinion.

L290ff: Remove the sentence starting with “These trends were supported by …”. You already mentioned that the trends are statistically significant in the sentence before. See also comment to L256.

L293ff: See comment to L256.

L303:   Maybe you could add one sentence at the beginning of this paragraph saying that in general the freshwater runoff clearly is the dominant term of mass loss compared to solid ice discharge. Or add one short paragraph at the end of this section, where you compare the contribution of the two mass loss terms and how the importance is projected to change over the 21^st century.

L312ff: Not surprising that the off-glacier rainfall and snowmelt are below 1% at the beginning of the study period, since the off-glacier area is basically non-existent at that time. See major comment 1.

Fig. 7:  (c) Why is the off-glacier rain missing here?

Fig. 8:  (a) How do you explain the results for the South-East? Why is the peak water reached earliest for SSP126, later for SSP245, earlier for SSP370 and latest for SSP585?

L345f.: Do you refer to the same study site as yours in these other studies you mention?

L357ff: You mention the regionally different behavior here, and also mention that the glaciers in the North-East are more resilient. Can you go a bit more into detail here. What is the reason for this behavior? Why is the FIIC so resilient to all emission scenarios?

L376:   This sentence implies that Greenland’s peripheral glaciers are calving-dominated at the moment, which is not correct. Please rephrase.

L384:   You explain the increase in ice discharge in the North-East with a more extensive coverage of marine-terminating (MT) glaciers compared to the other subregions. What is the ratio of MT glaciers here and compared to the other subregions/entire Greenland? Another interesting information would be to show how this ratio is changing over time in the different subregions. This way, you could show how many percent of the MT glaciers are becoming land terminating over the 21^st century, which would be an interesting information in my opinion (potentially to be added in the results section 3.2 as well).

L385:   You also mention that a delayed response to ocean forcing might be a reason for the increase in ice discharge. However, ocean forcing is not considered in your model.

L394f:  Give more details here. What are the regional differences and where do they come from? Why would it increase the freshwater contribution from the North-East glaciers by 2100?

And is this even the case? The contribution from North-East glaciers was in a similar order of magnitude in 2020, wasn’t it?

L403:   I rather see a prolonged glacier melt season in 2100, with higher glacier melt through September.

L413:   “capacity of glaciers to potentially regain a new equilibrium”  How do you see a new equilibrium from this data?

L422:   Oceanic feedbacks are not considered in your model. Thus, they can’t explain any modeled differences in peak water.

L459f:  See comment to L394f.

Section 4.4:  I would suggest restructuring this section:

Keep the first paragraph where you describe the uncertainties from the future forcing.

Then describe the uncertainties related to the climate data from the GCMs (currently paragraph 2 and 4).

Move L495 (“When comparing …”) to 503 (”… our findings.”) to the discussion section 4.1 as this is not related to uncertainties/limitations of the study. Remove the sentence starting L503f (“While OGGM …”).

Keep paragraph 5 about neglect of oceanic forcing (L509-518).

Keep paragraph 6 about neglect of (near-)surface processes (L519-525).

Make an additional paragraph about limitations due to calving observations instead of mentioning it in one sentence only at the end of paragraph 6 (L525f).

Keep the last paragraph.

L537:   Add a brief paragraph (1-2 sentences) on your approach and methods before diving into the results.

L540f:  Again, consider the entire catchment and not only the RGI outline if you want to discuss the runoff composition of a catchment (see major comment 1).

L544:   How do you know that glaciers would reach a new equilibrium under a low emission scenario?

L545:   How do you know that there would be a complete glacier loss and loss of glacier-fed runoff under a high emission scenario? This is neither seen in the volume evolution nor in the freshwater runoff data, where still 72% of runoff come from glacier melt at the end of the 21^st century.

L551ff: I would not only mention the significant different in SLR between low and high emission scenario. You also show significant differences in glacier area, volume and runoff amounts. Also for these variables your results demonstrate that greenhouse gas emission control is crucial.

Technical comments

L34:     Agree on how many decimal places you will show and keep it consistent throughout the paper

Fig. 1:  (a) The triangle symbols for the MT glaciers cover a large part of the glacier areas. In the SE basically nothing of the glacier area is visible. You should try to find a better way to displaying this information. Maybe you can use a smaller symbol, or use a dark and light version of the same color for land-/marine-terminating, or hatching, …

I would avoid using grey two times in this figure (for NO and CL2 glaciers) as it is confusing. It also took me a moment to understand which grey you picked up in panels (c) and (d).

(b) In the text you write “Flade Isblink Ice Cap” with capital C.

Tab. 1: Seems like the font sizes of the references are not all consistent.

L150 & L176:  Please use a consistent spelling for shallow ice approximation. The abbreviation “SIA”, that you introduce here, is not being used throughout the rest of the manuscript.

L172:   Remove “a”.

L214:   I suggest to add the variable name of  once more to make it easier for the reader: “We use the following (…) for calibration of air temperature sensitivity .”

L245f.: I suggest rephrasing to “Projections suggest a decrease in total glacier area of 19 ± 6 % under SSP126 and 44 ± 15 % under SSP585 by 2100”.

Fig. 3:  (a) The colorbar strongly highlights the upper and the lower end of values while it kind of draws the attention away from the middle part (40-60% remaining area). I would advise to use a perceptually uniform sequential colormap. Same applies for Fig. 4.

(b) Remove the number 2100 in the first subplot. It is not needed as you describe this already in the figure caption.

In the text, the SSPXYZ is always written in capital letters. Please adjust in the legend here.

Remove the legend in the last subplot. It is enough to have it one time in the first subplot (like you have it in Fig. 4).

It is hard to read the values of the 2100 result on the right side. Maybe you can duplicate the y-axis on the right side of the plot, or use horizontal lines every 25% step etc.

L254:   The abbreviation “HSD” for the “Tukey’s Honestly Significant Difference” has already been introduced before. Remove the long name here.

Fig. 4:  (a) adjust colorbar as in Fig. 3a

(b) The region names are crashing the line plots. Please avoid this. Potentially extent the y-axis to 125% as you did in Fig. 3b (in any case, be consistent with Fig. 3).

L266:   Be consistent with writing of exponents. In L241 you use “% yr^-1” while here (and many more times throughout the text) you use “mm/yr”.

L266ff: To make it clearer when you are talking about the trend and not absolute values, I would advise to add a “+” in front of the positive trends here.

L268 & 270:    Why is the trend given in mm/yr² here? Trends before were given in mm/yr.

L269:   Missing open bracket here.

L275:   Rephrase to “… SLR from Greenland’s peripheral glaciers will substantially increase …”

Fig. 5:  (a) Some region names are crashing the y-axis. Align these consistently in the subplots.

            (b) Some of the numbers are crashing. Place them further outside and/or make them smaller to avoid this.

Fig. 6:  Align regions names, see comment to Fig. 5a.

Some of the subpanels are narrower than others, be consistent and align subpanels.

The y-axis of the subpanels does not always start at zero as it seems. Make this consistent.

Move the legend with the SSP colors into the first subpanel, as in Fig. 5. Remove the two legends below the plot.

Move the year information for the two boxplots in the first subpanel, not in the fourth.

Fig. 7:  (b) Move this panel on the right side next to panel (a), like you did in Fig. 5; and move the legend outside of the pie.

Make the numbers smaller and/or move them further out to avoid crashing, as in Fig. 5b.

With panel (b) being moved to the first row, you can align (c) over (e) and (d) over (f). The legend could be set in the center between (c) and (d).

(e) and (f) What is the unit, is it Gt per month?

Fig. 8:  Are the colors in this figure the same as in the other figures? If not, please adjust.

L351:   Suggest rephrasing to: “The projected mass loss from Greenland’s peripheral glaciers …”

L403f:  Suggest rephrasing to: “The seasonal analysis (..) reduced glacier melt season and an increased and earlier contribution from snow melt as well as rainfall throughout the year (..)”

L414ff: Suggest rephrasing to: “However, (…) until exhausted, potentially becoming unable to support freshwater runoff.”

L450ff: Suggest removing this last sentence of this paragraph as it does not bring any new information and is no discussion.

L468:   Replace “researchers suggested” with “studies suggest”

L468:   Remove “lead to”

L470f:  Suggest removing the last sentence of this paragraph as it does not bring any new information and is no discussion.

L487:   Replace “melt-dominant” with “melt dominated”