Characterizing melt ponds on sea ice in the Norwegian Earth System Model (NorESM2)

Wang, Caixin; Debernard, Jens; Wang, Keguang

doi:10.5194/egusphere-2026-1525

Preprints

https://doi.org/10.5194/egusphere-2026-1525

Preprints

30 Apr 2026

| 30 Apr 2026

Characterizing melt ponds on sea ice in the Norwegian Earth System Model (NorESM2)

Caixin Wang, Jens Debernard, and Keguang Wang

Abstract. Melt Ponds (MPs) are pools of open water that formed on Arctic sea ice during the warm months. They significantly affect the surface radiation budget of the Arctic Ocean and play an important role in the Arctic sea ice mass balance, climate and ecological systems. The second generation of the coupled Earth System Model developed by the Norwegian Climate Center, NorESM2, includes an explicit representation of MPs. To characterize the spatial distribution and the seasonal evolution of MPs, we introduced four key variables: melt pond fraction (MPF), relative melt pond fraction (RMPF), MP area (MPA), and pond area fraction (PAF). Using these variables, particularly PAF, we conducted a comprehensive evaluation of MPs in NorESM2 across the Pan-Arctic region and its four sub-regions: the Pacific, Atlantic, Laptev and Canadian sector. Additionally, we classified MPs base on ice type, distinguishing between first-year ice (FYI) and multi-year ice (MYI), over the period from May to September during 2002–2011. This study utilized observational data, including MPF from MODIS and MERIS, sea ice age from NSIDC, sea ice concentration from MODIS and AMSR-E, and surface downward solar radiation from SYN1deg, and reanalysis atmospheric data from ERA5. Our study revealed that while differences existed between MODIS and MERIS, these discrepancies were relatively small during May, June and September, which corresponds to the early melt season and the refreezing season. In contrast, larger discrepancies were observed during July and August, the peak melt season. These significant deviations are likely attributed to the challenges associated with retrieving MPF using coarse-resolution optical sensors. Despite these differences, the overall spatial patterns and temporal evolutions of MPs derived from MODIS and MERIS were largely consistent. MPs were found to form earlier and more extensively on FYI compared to MYI, leading to more abundance of MPs on FYI. NorESM2 successfully reproduced these general characteristics. However, the model consistently simulated the maximum PAF was in August (except for FYI in the Pacific region), which was typically one month later than in the observations. NorESM2 exhibited a systematic underestimation of MPs in May and June, not only in the pan-Arctic but also across the sub-regions, and on both FYI and MYI. Conversely, MPs on MYI in the Pacfic sector were overestimated, likely due to the higher prevalence of MYI in NorESM2 during these months in this region. The systematic underestimation of MPs in NorESM2 during the early melt season can be attributed to the too low surface downward solar radiation in NorESM2 and the too large retrieved MPF from MODIS and MERIS. The former attribution was further confirmed by the experiments conducted with the 1D sea ice model ICEPACK. Giving the critical role of MPs play in Arctic albedo and energy budgets, advancing the representation of MPs in the model has a profound implications for the Arctic system.

Received: 18 Mar 2026 – Discussion started: 30 Apr 2026

Publisher's note: Copernicus Publications remains neutral with regard to jurisdictional claims made in the text, published maps, institutional affiliations, or any other geographical representation in this paper. While Copernicus Publications makes every effort to include appropriate place names, the final responsibility lies with the authors. Views expressed in the text are those of the authors and do not necessarily reflect the views of the publisher.

Download & links

Caixin Wang, Jens Debernard, and Keguang Wang

Status: final response (author comments only)

RC1: 'Comment on egusphere-2026-1525', Christopher Cardinale & Alek Petty (co-review team), 18 Jun 2026

Review of “Characterizing melt ponds on sea ice in the Norwegian Earth System Model (NorESM2)” by Wang et al.

Summary and general comment
This study evaluates how well the Norwegian Earth System Model, NorESM2, simulates the evolution of melt ponds over Arctic sea ice by comparing NorESM2 model output with MODIS and MERIS satellite observations from 2002 to 2011. The observations show that melt ponds form earlier and are generally more abundant on first-year ice than on multi-year ice. NorESM2 captures the very broad cycle, but underestimates melt pond coverage during the early melt season and tends to simulate the seasonal peak in melt pond area about one month later than observed. The authors attribute much of this bias to insufficient downward solar radiation in the model, which delays melt onset and pond formation, highlighting the need for improved melt pond representation to better simulate Arctic sea ice and climate processes.
Overall, this is a useful topic and the paper is potentially a good contribution to this journal, but the analysis and discussion need to be tightened substantially. It’s evaluating only one climate model which obviously limits its scope, so it’s important to instead try and provide sufficient detail and care in the analysis to make it interesting to a wider audience. I think the May/June bias is probably the most robust result, while the more detailed intercomparison of raw melt pond fraction values is much harder to interpret and I think should be condensed (and ideally improved). The manuscript would be stronger if it focused more on the broad seasonal cycle and general timing biases, rather than presenting detailed numbers as though this were a robust quantitative intercomparison.
The treatment of the observational products was also quite basic, with e.g. limited consideration of significant sampling biases. The sensitivity experiment is also too limited. It mainly demonstrates that increased downward solar radiation leads to earlier melt pond development in general, rather than fully resolving the May bias or isolating the cause of the model-observation differences. The role of clouds/DSR in driving the bias is plausible, but not demonstrated robustly and is thus quite speculative. Other causes could also contribute, including melt physics and broader atmospheric differences. The conclusion statements need to be softened accordingly or the analysis improved considerably.
Specific comments
The introduction jumps around from introductory physics to available observations and then back again. This needs to be reorganized to improve flow. I suggest focusing first on the science problems, then the models and data/observations needed to investigate and constrain them.
We appreciate this was only recently submitted and is only a preprint, but there is a very relevant paper that should be considered and cited (Smith et al., 2026), including an analysis of melt ponds across multiple CMIP6 models including two versions of NorESM (MM and LM). Melt pond area in the MM is pretty similar to the LM and only slightly larger in July/August, despite being colder. This could be cited in the models section as well, where NorESM-MM is introduced.
The observational products are quite different, which makes it hard to ascertain the bias. More importantly, the results need to discuss the observational sampling gaps much more carefully, including their potential impact on the conclusions. As is, the treatment of the obs data is quite weak.
I would avoid quoting so many raw numbers in the model vs obs comparisons. Instead, focus more qualitatively on the broad seasonal cycle and how the simulations compare with the observations. Given the crude treatment of the observational sampling and data gaps/averaging, the current level of detailed numerical intercomparison is not justified.
Similarly, phrases like “clearly overestimated” should be removed or softened. The analysis is not robust enough to support that kind of language in our view. There is too much detailed comparison of results that are not especially reliable given the observational limitations.
The sensitivity experiment mainly demonstrates earlier melt pond development during June rather than resolving the May bias. As it stands, the sensitivity study does not add much beyond the basic point that increased DSR leads to earlier melt pond formation. It needs at least one additional test to be more convincing. The DSR sensitivity study is also extremely basic. Other causes could also contribute to the differences. For example, melt physics is not explored, and atmospheric differences beyond DSR could also be important.
r1i1p1f1, r2i1p1f1, and r3i1p1f1 are used here but we noticed that r1i1p4f1 also provides melt pond output. Is there a reason to not use perturbed physics members?
The justification for using MODIS-Low is quite weak. This needs to be explained more clearly.
There should also be more justification for using the period 2002-2011. There are other observational datasets, e.g. later MPD1 (2017 to 2023) and MPD2 (2017 to 2025) that could have easily been used to extend this forward in time.
We recommend computing the NorESM MPA over the full domain (pole included) and adding to the current Figure 4 or stating how adding the pole hole in the calculations impacts MPA and PAF
The authors do not provide an exhaustive list of causes of the MPF bias. DSR may be the most plausible cause, but more context is needed.
The conclusions are too strongly stated. I would stick to explaining what the sensitivity study does and does not demonstrate.
The northern sea ice age distribution is quite different, so I am not sure an ice-age comparison is especially logical here.
Technical comments
This is not the best written manuscript and needs careful editing. There are a lot of spelling and grammar issues throughout. I have not listed all of these individually.
All the melt pond acronyms are a little confusing. Please simplify where possible and make sure all acronyms are defined clearly. Define acronyms in the figure captions please.
I would simplify the discussion of Figure 5. The current discussion does not seem to add very much.
L51: The statement “AI techniques” is too vague, as this likely just applies to the processing of remote sensing data.
L60: The discussion of AI is not clear and I think it should be dropped. As noted later, the MPF dataset from MODIS uses ANNs.
L125: It is not really a “column version” of CICE; it is the column physics model of CICE, ICEPACK. CICE essentially adds a dynamic core around ICEPACK. This only needs a slight rewording for clarity.
L137: This statement feels too strong given the limited validation. I would suggest softening it.
L143: How are these variables actually used in the analysis?
L160: Why the specific emphasis on ERA-Interim here? The rationale is not clear.
MPF calculation: the calculator of MPF and RMPF is done on monthly timescales here, but it would be better to do this first at daily timescales before averaging monthly (we assume you have access to the full daily output).
L205: The notation here seems unnecessary for simply saying that a mean is calculated over time and across grid cells. More importantly, how are the sampling differences handled, especially given the significant quantity of missing data in the observations? We recommend just stating that the multi-year seasonal mean is computed over all months from May to September and all years from 2002 to 2011. Also, maybe a good idea to remove September due to a lack of observational coverage, especially since the MODIS dataset only includes half of September. You should also consider removing it for any seasonal mean plot for a more relevant model-observation comparison.
L461: This is not an exhaustive list of causes. Presenting it as such creates problems for the discussion that follows and is a major weakness of the current framing. Figure 7 of Smith et al., (2026) also suggests an issue with the parameterization promoting melt pond concentrations that are a strong function of sea ice concentration, that does not seem to be supported by the observations.
L530: Does it really show that? I would stick to explaining what the sensitivity study does and does not demonstrate.

References:
Smith, M., C. Cardinale, A. Petty, H Niehaus, (2026) Regional Biases in Arctic Melt Ponds Between CMIP6 Models and Satellite Observations, ESS Open Archive [in review at JGR Oceans], https://doi.org/10.22541/essoar.15001935

Citation: https://doi.org/10.5194/egusphere-2026-1525-RC1
RC2:
'Comment on egusphere-2026-1525', Jean Sterlin, 10 Aug 2026
Main comments

This study aims to evaluate the representation of melt ponds in NorESM2 using CMIP-style simulations over the historical period. Melt ponds in NorESM2 are compared against two satellite-based observational products of the melt pond area fraction. The melt ponds in NorESM2 are parameterised by the so-called level-ice melt pond scheme of Hunke et al.(2013). This scheme assumes that melt ponds develop preferentially over levelled ice. The authors found that while NorESM is able to capture the spatial distribution of the monthly mean melt pond fraction, it underestimates the melt pond fraction in spring and simulates the seasonal maximum melt pond fraction one month later than in the observations. These results are likely robust. Then, the authors investigated melt ponds in different sectors of the Arctic Ocean and in combination with two sea ice age categories (first-year and multi-year ice). Although this part of the study is interesting and relevant, I’m not sure about the methodology due to the definitions of the sectors and sea ice age categories. Indeed, Fig 8 shows that Arctic sea ice in summer is dominated by one age category. Next, the sensitivity of the modelled melt ponds to downwelling shortwave radiation was investigated using Icepack, the column physics module of CICE. This module is used to further examine the seasonal biases in the modelled melt pond fractions with a more constrained setup than the fully coupled historical experiment. I’m not convinced that the shortwave radiation in NorESM is the only factor explaining the seasonal biases in melt pond fraction over the historical period.

To my knowledge, this study is the first in-depth examination of melt ponds parameterised by the level-ice scheme in an ESM. Overall, having NorESM2 alongside Icepack as a vertical column model is a nice opportunity to examine the biases in fully coupled simulations and provide solutions for the modelling community. These provide all the good ingredients for a very interesting study.

The figures are generally clear. The text is relatively easy to read but the scientific English needs to be improved before publication (see my minor comments). The manuscript focuses significantly on the differences between two observational melt pond products. This comes at the cost of other more important discussions on model biases, comparison to in situ observations and other melt pond modelling studies, etc. In my opinion, the examination of the observational datasets could be sourced from the scientific literature using citations. This will make more room for exploring the main scientific question. Concerning the latter, it needs to be reframed to examine questions relevant to the community. Here are some ideas I have collected but it’s up to you to decide what’s relevant:

Do you find a correlation between the September minimum in sea ice extent with the melt pond fraction in spring (Schroder et al. 2014)?

Is there any significant trends in the melt pond fraction over the historical period and how does it compare with observations (e.g., Feng et al. 2022) or models (Zhang et al. 2018 and my paper sterlin et al. 2021)

Do you agree with the conclusion of Webster et al. (2022) that the level-ice melt pond scheme (and possibly others schemes in ESM) overestimates the summer melt pond fractions in present days?

Many methodological details are unclear. The melt pond scheme and relevant processes are not well presented. In particular, I’m curious about how the refreezing of melt ponds is implemented. In CICE, frzpnd is ’CESM’ or ‘hlid’? This matters because the melt pond area fraction measured by satellites may have another definition than in CICE, and this can impact your results. I’m also curious about the infiltration of snow by melt water since the Icepack documentation says it can delay the onset of melt ponds. Flushing of melt ponds can also influence the modelled melt ponds. In fact, the authors should examine any source or sink of melt pond water, as it may explain the seasonal biases reported in the paper. I also advise examining the simulated melt pond volume and depth. In the model, snow physics could be investigated.

Continuing on the methodology, downwelling shortwave radiation is one of many external drivers that control the evolution of melt ponds. For this reason, I’m not convinced that the results in Section 4.5 obtained with Icepack are robust. Of course, shortwave radiation is important for melt ponds. But as long as the snow cover has not substantially melted, the surface albedo over sea ice remains high, and most of the incoming solar radiation is reflected. On the other hand, we can imagine a warm air mass advected over sea ice. This mass would subsequently increase the turbulent sensible heat flux over sea ice, trigger snowmelt and the onset of melt ponds. I’m taking the example of the surface air temperature because of the warm temperature bias in the Central Arctic present in ERA5 Tian et al. (2024). However, all the terms in the surface energy balance should be examined carefully before concluding.

Because NorESM2 contributes to the CMIP exercises, it may participate in the new OMIP experiments forced by JRA55-do. If that’s the case, the authors could run Icepack simulations forced by JRA55-do and ERA5 to get a sense of the effect of the uncertainties in the atmospheric forcing. It will help bridge the gap between the fully coupled historical simulations and the Icepack runs.

The work by Diamond et al. (2021, 2023) on melt pond schemes in HadGEM is relevant to this study in regard to the purpose of NorESM2 and the style of experiments. But, I think a strict comparison would require an ensemble of control runs. You could still discuss some of their results.

Minor comments

line 5:

It may be worth giving the name of the melt pond scheme. I would specify all the key methodological elements to get the gist of the model study.

line 6:

If you take a step back, the readers going through the abstract won’t be able to tell the differences between all these variables. Without context, there is no real difference between the melt pond fraction and the pond area fraction. In addition, the term relative is ambiguous because fractions are always relative to something.

As a reader, I would love to see variables that are comparable with other studies. Given the objective of your study, I’m thinking of the paper by Zhang et al. 2018 on melt ponds in MIZMAS. But I’m also thinking of the CMIP variable definitions: simpconc_tavg-u-hxy-si, simpeffconc_tavg-u-hxy-si, simprefrozen_tavg-u-hxy-simp and simpthick_tavg-u-hxy-simp (see CMIP7 data request or Notz’s paper for SIMIP). For the latter, this brings me two questions:

Are you presenting results for the total melt pond area fraction including refrozen ice lids or the effective melt pond area fraction exposed to the atmosphere? This difference matters when comparing observations.

Why not discuss the melt pond depths and volume per unit (ice) area? Examining the melt pond sinks and sources of melt water may help you to find out why NorESM2 is underestimating melt ponds in spring.

line 8:

Consider removing “its” next to four sub-regions. Your definition of the sub-region is arbitrary.

line 9:

based; it’s not really ice types but more sea ice age categories

line 10-12:

I would try to rephrase this sentence to focus on the methodology. It should include and connect with Icepack, to describe how you came up with your set of experiments.

lines 12-18:

It’s interesting but it’s a lot of lines for an abstract and for results arguably out of the scope of the study. Let me explain myself. The title says you are doing a characterization of the melt ponds in NorESM2. Here, 20ish percent of the abstract is used to describe the differences between MODIS and MERIS observations.

line 19-20

Check the grammar.

line 24:

I don’t understand this sentence

line 26

This reads funny. It can be understood as MPs in NorESM impact the Arctic climate, which is totally wrong. I think you need to reformulate this last sentence. There is also an issue with MPs play.

line 29:

“in exchanging” I’m not sure about the grammar

line 33:

Sea ice area and extent are somewhat related; so are thickness and volume. Moreover, we don’t know if those quantities are integrated, averaged over the whole Arctic sea ice, or if they are local. You need to be more accurate. There are other trends in Arctic sea ice conditions (e.g., sea ice velocity).

line 34:

You don’t really explain the causality between decreased sea ice extent and sea ice age. In fact, one could think (wrongly) that the decrease in sea ice extent is linked to FYI losses.

In “decrease of -13.4 % per decade” combines two negatives, I’m not sure that it’s correct.

I would avoid citing websites directly in the text.

line 35:

What do you mean by continuous?

Arctic sea ice does not decrease, its area does. Arctic sea ice is shrinking, weakening, thinning, declining, etc.

line 36:

I’m less familiar with projected Arctic sea ice free conditions, but aren’t we already in the 2020s?

line 37:

Check the citation. Is it the right format? I’ve seen Notz’s paper cited as “(SIMIP Community 2020)“.

line 39:

Topál. Review the references as well.

line 40:

“The decrease of Arctic sea ice reduces ..” is weird to read.

solar absorption

line 42:

ice–albedo feedback

line 44:

Although lead is grammatically correct, I tend to be careful when using this verb because it can also be a noun when referring to ice leads. I took a break when reading melt pond leads (but it’s a me problem).

line 47:

So far, I’m not fully at ease with the usage of citations. Here, in particular, none of the cited papers examine the role of MPs for the ecosystem. Citations should support the argumentation. They should be relevant to what’s discussed.

line 49:

What do you mean by sea ice conditions?

line 50:

it’s either “and” or “or”, but cannot be both.

The phrase is not very informative. In addition, there have been efforts to create models to study melt ponds outside of ESMs. I’m thinking of the continuum model from Flocco and Feltham (2017) but there are others out there. Moreover, melt ponds have also been examined in-situ during field campaigns (e.g., Webster et al. 2022).

line 52:

restricted? By whom?

line 54:

“relies on ground measurements” why is it a problem?

line 55:

This sentence is hard to read.

line 57:

You should provide some numbers or scales to support your discussion.

line 60:

Rosel et al. (2012) employed an artificial neural network and MODIS data to build their melt pond dataset. To me, their dataset falls in the broad AI techniques category as well as satellite products. I’m having a hard time to follow the line of thoughts.

line 63:

The introduction of melt ponds is too late in the manuscript. They have already been discussed in the previous paragraphs.

It seems that the pool of open water is a borrowed expression, but it is a bit ambiguous to me. Open water could refer to the ocean. In reality, melt ponds are pools of melt water. Furthermore, melt ponds can be concealed in the snow cover when forming or even trapped under an ice lid. As a result, melt ponds are not always “open”. Concerning the seasonality, I don’t think it’s quite correct. What is late spring? Why no melt ponds in autumns?

line 64:

and

line 66:

What difference do you make between surface roughness and ice morphology? I’m not sure about the citations usage here.

line 67:

Can you explain what you mean by geometric correlation? Anomalies of what ? SST? SAT? Warm anomalies are not geometrical objects.

line 68:

Seasonal evolution?

line 73:

limited relief?

line 75:

“the pond evolves on ice”? What do you mean?

line 79:

I don’t get what you mean because the albedo of MP over MYI should be lower than over FYI since the former are deeper than the later in general. I’m not saying what you say is wrong but it’s not clear.

line 81:

Maybe, split the sentence according to the topic: albedo is different from biogeochemistry.

line 82:

This is not well formulated. The preceding sentence refers to observed MPs. Here, you switch to models. Are you saying that melt ponds in the real world cause sea ice losses in models?

line 87:

Webster is more about observations than climate and forecast models. Not sure that Flocco, Holland and Hunke talk about forecast systems either. I think you need to revisit the citations. They are not well employed.

What do you mean by implicit?

One model can have multiple configurations. Here you seem to be comparing models with configurations without any distinction. By the way, the sentence is similar to “of the 125 CMIP6 model configurations, 14 use an explicit melt pond scheme (11%) and the remaining 111 use implicit treatments “ from Diamond et al. (2023). Are you sure of the reference to Aparício? In any case, it’s better to reformulate this sentence.

line 94:

I’m not sure that CO₂ should be in italic.

line 95:

It’s better to focus on the main conclusions of the studies rather than what they did.

line 99:

You are not well presenting your scientific question and objectives. There are many underlying questions associated with your model setup and how it compares with other ESMs. These are not discussed. For example, ICEPACK just pops up out of nowhere in the next part of the paragraph.

One tip: it would be nice if you can explicitly state what’s the novelty of your study.

line 102:

Usually, sections are introduced in ascending order

line 107:

This suggests that the Norwegian Climate Center is in charge of the development of CESM.

line 108:

It’s not how acronyms should be introduced.

line 119:

atmosphere as a climate component or a model component?

line 114:

CICE comes shipped with three melt pond schemes, I believe. Here, in the text, it seems that the level-ice melt pond scheme is the only param available in CICE.

There should be a paragraph explaining why NorESM2 selected this scheme, how it works, and its specificities compared to the other parameterizations. To my knowledge, melt ponds can develop on deformed ice. I’ve always been curious about why the level-ice scheme does not allow melt ponds to develop on deformed ice. This scheme is an interesting choice that needs further discussion.

line 116:

I don’t think 5 categories is a standard.

line 122:

Is 3 members enough?

It’s not clear to me which CMIP protocol you are following. Sometimes, people loosely refer to CMIP when working for MIPs activities (e.g., OMIP, SIMIP, or even different MIP scenarios). Is it for CMIP6 or CMIP7? You should be clear about your experiments and model setup. The same applies to the method section for Icepack. The difference in coupling methodology should be explained. I think that Icepack is forced by ERA5 whereas NorESM2 is run with the ocean and sea ice coupled to the atmosphere. I had to fish that information outside of the methods section. This information should appear in the method section.

line 126

Icepack is not the column version of CICE.

line 128:

avoid putting hyperlinks directly in the text. They tend to break with time.

line 131:

I don’t understand this sentence. What do you mean by the same configuration? You must make some assumptions to be able to run Icepack without CICE. How do you deal with the advection of sea ice? Lateral boundary conditions? How is your Icepack run initialized? Is the levelled-ice changing with time?

line 137:

Lee et al. produce an intercomparison study of melt pond products. It’s not a study about calibration and validation. In fact, you may want to follow the recommendation of Lee for choosing your observational datasets.

line 141:

Haven’t Istomina et al. (2025) updated their dataset to cover 2002-2023? Is it available? And if so, can you use it along with the dataset of Lee et al. (2020)? I haven’t searched for a new version of Rosel’s dataset. But Feng et al. (2022) have produced a dataset based on MOD09A1 over a longer period than Rosel. The manuscript also cites a paper by Ding et al. (2020). In brief, it would be nice to have more recent and up to date observational datasets.

line 152:

There are two possible subjects in this sentence.

line 159

It’s worth taking the time to discuss the uncertainties in ERA5 in light of other atmospheric reanalyses. How does ERA5 compare with other atmospheric reanalyses in the Arctic?

I’m not sure whether or not NorESM2 will be participating in the OMIP experiments for CMIP7 but if that’s the case, you might want to try to run Icepack with JRA55-do in addition to ERA5. I personally did a comparison between JRA-55 and DFS5.2 (roughly ERA based) datasets and found substantial differences in the surface air temperature in the Arctic. This will enable you to get a sense of the uncertainties in the atmospheric forcing methodology, which are significant in polar regions.

line 168:

Acronyms should be put in parenthesis: Clouds and the Earth’s Radiant Energy System (CERES). Check the grammar in the sentence.

line 172:

Why is the degree character in italic? Can you say 1° or one-degree instead of 1° x 1° ?

line 185:

per unit sea ice area?

line 188:

direct product: you mean, as a multiplication?

line 192:

we don’t know from the math expression what is summed. Is it over the ice thickness categories?

line 194

I suggest borrowing the melt pond terms and diagnostics from Zhang et al. (2018): they did a good job and it will help the readability of the paper. There is no real need for Equations 1 to 8. See how Zhang’s paper is structured. Otherwise, you could select the CMIP definitions.

Figure 1:

I’m not a big fan of the names of the sub-regions. First, Laptev is a sea, Canadian is a region, Pacific is an ocean. Second, Laptev also includes the Kara Sea. Canadian also includes Greenland. Pacific is very broad.

The main issue is that the names make the text ambiguous. For example, the Pacific encompasses the Beaufort Gyre, landfast ice and sea ice formation sites in the East Siberian Sea, and even the Sea of Okhotsk. All of these regions are regulated by different climate processes. As a result, the reference to the Pacific in the text makes it hard to capture which sea ice region or climate is discussed.

I could say the same about the Canadian sector: The Lincoln Sea, Baffin Bay and Davis Strait, the Canadian Arctic Archipelago are different sea ice regions. Is Canadian even grammatically correct?

This problem also impacts abstract (line 8) since a reader pressed with time will unlikely have seen Fig 1. For that reader, the Laptev sector is akin to the Laptev Sea.

I also suspect that the choice of regions makes it hard to see obvious differences or agreement between FYI and MYI in section 3.3. Can you explain the rationale behind these sub-domains? You could take inspiration from the domain defintion of Niehaus et al. (2025). This will help to compare your results with theirs.

Last question: Is Hudson Bay part of the scope of your analysis? If this region is covered by the obs, I would try to include it.

line 180:

I don’t really understand what are MODIS-low and -high. Does MODIS-high correspond to the mean melt pond fraction from the beginning to mid-September? Adding zeros or changing the averaging time period is a bit problematic. To me, it’s like inventing data. Can’t you simply take September MODIS data out of your analysis?

line 197:

Note that the documentation of the sea ice age product mentions a 15% ice concentration cutoff. 15% is also a commonly accepted threshold for calculating the sea ice extent.

Results

Overall, this section is very descriptive and it’s a bit hard to get the salient points from the text alone.

line 222: “The observations clearly showed”.

The phrase should be in present tense since you are discussing things occurring on a regular basis (monthly means).The past tense suggests that melt pond conditions have changed since 2002-2011, and that MODIS and MERIS datasets are now outdated.

You should use a past tense if referring to a specific date of the observational dataset. However, I don’t think it applies to the paragraph. There are other issues with the tense in the following paragraphs and sections.

line 229: “Canadian bays/straits”.

Canadian Arctic Archipelago ? Canada is quite rich in bays and straits. There are sea ice forecasts for the Maritimes provinces on the Atlantic facade of Canada, for example.

line 223

Southern latitudes is a bit odd: it could refer to the Southern Hemisphere or even Antarctic sea ice. I would use high (northern) latitudes to refer to the Arctic. You could say lower latitudes. Besides, there has been only one sea ice edge in the Arctic Ocean, to my knowledge. So the southern sea ice edge is confusing.

line 224

Could it be a bias in the satellite observations? Melt ponds can be trapped under an ice lid covered in snow in autumn. They could remain undetected by the satellites.

line 235:

"over most parts of" or "throughout most of”

Figure 3.

Should read “Same as Fig. 2, but for RMPF.”

line 237:

I know that you are referring to the subregions of Figure 1 but Pacific and Laptev regions can be misleading. It’s not very accurate.

line 238:

I got lost after the “but”. This is hard to read.

line 239:

“north Pole” and “north pole” should be “North Pole”.

The “MPF magnitudes” stands for “the melt pond fraction magnitude”, which is a bit hard to read.

line 244:

Can the “Greenland Sea” replace “east of Greenland of the Atlantic region” ? The “marginal seas of the Pacific region” is a very broad term. And some may be upset by “Canadian region”. Also, it’s not clear to me if the Canadian sector implies the Davis Strait and Baffin Bay in the manuscript.

line 245:

You should discuss the representation of the sea ice concentration in more detail. Indeed, the analysis of the modelled RMPF against the observations implicitly validates the sea ice concentration in NorESM. But, is it well represented in the model and comparable between the observations?

line 246:

I don’t think there is such a constraint on the melt pond fraction per unit sea ice area (RMPF) and per unit area (MPF) are different diagnostics: one can be higher or lower than the other without any constraints as long as the ice concentration is non-zero.

line 258:

A lot of the paragraphs start with Figure X shows. You should focus on main ideas and arguments. The figures are there to support the text, not the other way around.

This will make the text flow better.

line 259:

There should be spaces between the multiplication character (e.g., 2.0 x 106 km2). Some authors write “2.0 million km2”.

line 265:

Can the observed sea ice concentration explain some of the differences between the two observational products?

line 269:

“obviously” suggests that this result is expected and trivial.

I have the same question about the sea ice concentration (total sea ice area) as with MODIS and MERIS MP fractions: Can the sea ice concentration explain the difference in the total melt pond area?

line 271:

For a range, you should use an en dash (–). It will prevent interpreting the difference 0.01-0.04 as -0.03.

line 276:

“The seasonal evolutions … showed a similar seasonal cycle” is weird to read.

Figure 2: Once the refreezing of melt ponds has been introduced, can you specify if the modelled MPFs include or exclude ice lids? Ideally, you should present the melt pond area fraction exposed to the atmosphere.

line 281:

Again, all these paragraphs are very descriptive. I’m honestly getting overwhelmed by all these results. In the meantime, you did a good job with most of the figures. They are easy to examine without the paragraphs. So, you can leave part of the analysis to the reader: if they want to measure something, they can look at the figures.

Your goal with the text is to help the reader summarize the key points and explain to them what they are seeing. For example, I’d like to know why the onset of melt ponds is delayed and underestimated in NorESM2, why melt ponds in the Canadian sector are better represented, how does NorESM2 compare with other model studies (especially Hunke et al. 2013; see also work by Zhang et al. 2018 because you have a similar scientific objective), what can we draw from other observational campaigns (see Webster et al., 2022 amongst other), etc. In addition, you should discuss melt pond processes and drivers further. Are there any melt pond processes missing in NorESM2 that could explain your results? What about the effective melt pond fraction exposed to the atmosphere as opposed to the total melt pond fraction including ice lids?

line 289:

Melt ponds can survive in autumn once capped under an ice lid covered in snow (see work by Flocco et al., 2015). Is it something you experience with NorESM2?

Figure 6:

Is it possible to divide the figure in two panels, one histogram linked to FYI and another one linked to MYI? Same for Figure 7

line 291:

“more larger” is not grammatically correct. Note the symmetry in the sentence: Sea ice in NorESM2 … covers a larger area … due to a larger area in NorESM2. The next sentence is quite poor as well. You should explain somewhere that PAF is representative of the pan-Arctic mean melt pond fraction per unit sea ice area, if I understand the diagnostic correctly.

line 395:

(0) ? I honestly spent less time reviewing Section 3.3. Can you change it to focus more on the key results and propose a discussion? Try to write a scientific story. For example, look at lines 408–411, there is an endless list of results. But what does it mean?

line 414:

Discussing uncertainties in the modelled PAF is awkward because it should be fully determinable and accurate if we leave floating point errors aside. Here, the wording suggests that you are not fully sure of your model diagnostics. On the other hand, there are uncertainties in how the melt pond scheme is designed, which processes are included and how, etc. This information should appear somewhere.

line 417:

Usually, larger should refer to something (larger than what?). Otherwise, you could say large, the largest, or above than normal (if there is such a thing as normal). When referring to the standard deviation, I think it should be low or high instead of small or large.

Section 4:

It’s not really a discussion but more additional result descriptions.

Section 4.2

Note that the sea ice age is discussed after being used in the result section when reporting on the melt pond statistics over different ice types. Is it correct?

At this stage, I haven’t figured out where sea ice age comes from in your simulations. Is it modelled by NorESM2 or do you take it from NSIDC?

Have you considered other ice age categories (e.g. 1–2 years)?

I think the idea is that simulated melt ponds on thick old deformed ice should be smaller in extent but deeper than over thin level ice. You’ve produced results with sea ice age. But, the link with the ice thickness could also be discussed. And because of the type of melt pond scheme used for this study, it would be nice to have a discussion on the representation of level ice in NorESM.

line 419:

Does it mean 9 subregions? 3 subregions for each domain?

line 420:

I don’t understand this sentence: is it both the MYI and FYI or only one of them?

line 421:

Do you take the mean of the standard deviation? Isn’t it more preferable to compute the STD for each month?

line 423:

What does “accounting for 5% of the ensemble mean” mean?

“These results confirmed”: there is a problem of tense.

The acronym NSIDC is ambiguous in this context because it’s a center offering multiple sea ice products (including sea ice concentration). To me, it’s not clear from the text whether there was no ice in CAA or if the sea ice age product failed in CAA in spite of the presence of sea ice. I would change the acronym to be more specific.

Figure 8

Can you recall the time span of the means in the legend.

Figure 8 makes me a bit worried about the significance of melt pond statistics over FYI in summer. Is it possible to add the 15% ice concentration line to the figure to have an idea of the position of the ice edge? My concern is that the observed FYI is too much in the marginal ice zone, and the associated data should be considered with care. Moreover, the documentation of NSIDC–0749 states that “the age estimates are restricted to open ocean areas only”. It looks like the product algorithm needs a certain amount of open water to work effectively. Hence, you might need to examine the ice pack with more care.

In Figure 8, the distinction between FYI and MYI removes a lot of information from the sea ice age product. I understand the choice made by the authors, but I find it a bit annoying when looking at the large dark areas linked to MYI on the maps. I also suspect that the choice of age category prevents finding clear results in Section 3.3. To me, it would be useful to have sub-categories for MYI.

Overall, I welcome the idea of working with the sea ice age to validate certain melt pond aspects. But there is a need to describe the limitations of the methodology further.

line 449:

have challenges to

line 440-456:

The sentences in the paragraphs are quite repetitive. You should make it more concise.

line 461:

The itemization is broken over multiple sentences. Check the placements of commas and periods. I find it easier to break long sentences in smaller ones to help the readers.

I think it’s a bit unfair to start discussing the uncertainties in the observations before model biases. But it’s my opinion as a modeller.

line 466:

As shown.

Figure 9:

If you use SYN1deg in the manuscript, you should use the same acronym in Figure 9. But please consider that CERES is the name of a satellite sensor, which is consistent with MODIS and MERIS.

Why no downwelling shortwave radiation over the open ocean?

line 469:

There is something wrong with the format of the units in the paper

line 468:

isn’t it quite a lot already? I tried to find some maps of DSR in the Arctic and found this paper: https://acp.copernicus.org/articles/26/3321/2026/ (See their Figure 5). In SYN, the DSR is ~250 W/m2 over Arctic sea ice in May, increasing to ~300–350 W/m2 in June. The relative difference between NorESM and SYN is 16 to 32%.

line 471:

In the sentence, are you referring to an underestimation by the model or ERA5? In any case, can you explain why the difference in DSR between the model and ERA5 is smaller than with SYN?

line 473:

I don’t understand the argument. How do you find a total cloud fraction overestimation of 0.40 for NorESM when in the previous sentence you only compare ranges from CERES and SYN1d?

line 475:

ICEPACK is not the column version of CICE.

How do you initialize the sea ice state? In particular, your melt pond scheme requires you to know the amount of level ice. What’s your methodology? Could you add that to the methods section.

line 478:

The hyperlinks tend to go 404 after some time. I haven’t checked the citation but it may be sufficient.

line 483:

“remained at this depth” is not well formulated because it should refer to a value, not a physical variable. Snow depth does not remain at depth.

line 476-484:

The choice of tense is good! On the evolution of the snow depth, I was expecting important changes but after examination of Fig 10a, it seems that the changes in snow depth were relatively small before the onset of snow melt. Can you reformulate the paragraph to help the reader get the key points?

line 490:

of of ; IMB structures

line 492:

Are you sure? On Figure 10b, the snow pack begins to melt on 1st of June while the observed snow thickness starts to decrease on the 12th of June. I agree that snow melt is faster in the model in observations.

line 496:

An abrupt; to be caused by / likely caused by

line 501:

If not knowing about Fig 10a, this sentence comes out of nowhere. You should say that it corresponds to the simulated melt ponds by Icepack at the location of the IMB.

line 402:

it’s the influence (singular). You could say:The sensitivity of met ponds to DSR is …

line 509:

I don’t understand the last sentence. Do you mean in your model or as reported in observations?

line 545:

My two cents is that melt pond parameterisations in sea ice models are getting old if we consider recent observation campaigns (e.g., MOSAIC). Based on your research, what are your recommendations for melt ponds developments in ESM? Which processes are missing in models?

line 599

DOI is not DOing: I clicked on the doi link and got DOI Not Found

References

- CMIP7 data request: https://wcrp-cmip.org/cmip7-data-request-v1-2-2-3/

- Diamond et al. (2023): https://journals.ametsoc.org/view/journals/clim/37/1/JCLI-D-22-0902.1.xml

- Feng et al. (2022): https://www.sciencedirect.com/science/article/abs/pii/S0921818122001990

- Flocco et al., 2015: https://agupubs.onlinelibrary.wiley.com/doi/full/10.1002/2014JC010140)

- Hunke et al.(2013): https://doi.org/10.1016/j.ocemod.2012.11.008

- Istomina et al. (2025): https://doi.org/10.5194/tc-19-83-2025

- Lee et al. (2020): https://www.sciencedirect.com/science/article/abs/pii/S0034425720302893)

- Niehaus et al. (2025): https://tc.copernicus.org/articles/19/3915/2025/

- Notz’s paper on SIMIP: http://dx.doi.org/10.5194/gmd-9-3427-2016

- Schroder et al. 2014: https://www.nature.com/articles/nclimate2203

- Tian et al. (2024): https://www.nature.com/articles/s43247-024-01276-z

- Webster et al. (2022): https://online.ucpress.edu/elementa/article/10/1/000072/169460/Spatiotemporal-evolution-of-melt-ponds-on-Arctic

- Zhang et al. 2018: https://agupubs.onlinelibrary.wiley.com/doi/full/10.1029/2018JC014298

- sterlin et al. (2021): https://www.sciencedirect.com/science/article/abs/pii/S1463500321001256

- zheng et al. (2026): https://acp.copernicus.org/articles/26/3321/2026/
Citation: https://doi.org/10.5194/egusphere-2026-1525-RC2

Caixin Wang, Jens Debernard, and Keguang Wang

Viewed

Total article views: 359 (including HTML, PDF, and XML)

HTML	PDF	XML	Total	BibTeX	EndNote
245	83	31	359	39	25

HTML: 245
PDF: 83
XML: 31
Total: 359
BibTeX: 39
EndNote: 25

Views and downloads (calculated since 30 Apr 2026)

Month	HTML	PDF	XML	Total
Apr 2026	38	7	3	48
May 2026	151	48	16	215
Jun 2026	24	9	4	37
Jul 2026	19	11	5	35
Aug 2026	13	8	3	24

Cumulative views and downloads (calculated since 30 Apr 2026)

Month	HTML	PDF	XML	Total
Apr 2026	38	7	3	48
May 2026	151	48	16	215
Jun 2026	24	9	4	37
Jul 2026	19	11	5	35
Aug 2026	13	8	3	24

Viewed (geographical distribution)

Total article views: 336 (including HTML, PDF, and XML) Thereof 336 with geography defined and 0 with unknown origin.

Country	#	Views	%

Latest update: 20 Aug 2026

Short summary

NorESM2 generally reproduces the spatial distribution and seasonal evolution of MPs, although MPs forms later and develops faster compared to observations. Additionally, NorESM2 demonstrated a systematic underestimation of MPs in May and June, which was attributed to the low DSR in NorESM2 and the potential overestimation of MPF in MODIS and MERIS due to the coarse-resolution of the optical sensors. The absence of MPs in the early melt season has profound implications for the Arctic system.


Total:	0
HTML:	0
PDF:	0
XML:	0