the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 21 Jun 2024
Modelling firn density at Dye-2 and KAN_U, two sites in the percolation zone of the Greenland ice sheet
Abstract. Modelling firn density is crucial for understanding the current mass balance of the Greenland ice sheet (GrIS) and predicting its future. As snowmelt increases in a warming climate on the GrIS, accurate information about firn density and its variability over time and space becomes increasingly important in the percolation zone. Previous research indicates that none of the existing models accurately simulate firn properties at sites with varying snowmelt rates due to the limited knowledge of liquid water percolation and the complexity of firn densification. Here, we enhance the representation of the firn densification model based on the Community Firn Model (CFM) by (i) using the stressed-based dry-firn densification scheme and recently published parametrizations for firn characteristics, (ii) modifying the expression of irreducible water in the Darcy-flow scheme, and (iii) allowing the time step to be adaptive. The improved model is employed at two climatologically distinct sites, Dye-2 and KAN_U, two sites in the southwest percolation zone of the GrIS, to evaluate its performance. The modelled firn depth-density profiles at the two study sites generally agree well with the in situ measurements obtained from 16 firn cores drilled between 2012 and 2019. At Dye-2, with comparatively high accumulation and low snowmelt rates, the model simulates thin ice lenses and/or ice layers in the top 10 m of the firn column. The modelled firn density aligns with the average level of observations, with the relative bias in density ranging from 0.36 % to 6 %. At the KAN_U site, characterized by relatively low accumulation and high snowmelt rates, the model captures high-density layers (~917 kg · m-3) caused by the refreezing of liquid water. The observed ice slabs are partly reproduced, and the relative bias in density between simulations and observations at all 8 cores is within ±5 %.
- Preprint
(9442 KB) - Metadata XML
- BibTeX
- EndNote
Status: closed
-
RC1: 'Comment on egusphere-2024-1726', Baptiste Vandecrux, 25 Jul 2024
Review of Modelling Firn Density at Dye-2 and KAN_U, Two Sites in the Percolation Zone of the Greenland Ice Sheet
by Dr. Xueyu Zhang et al.This study uses the Community Firn Model (CFM) and RACMO2.3p2 to simulate the evolution at two sites on the Greenland ice sheet. The CFM is improved with a new, Darcy-like infiltration scheme. Despite the (much appreciated) comparison of the model results to previous firn model intercomparison exercises and the care given to the writing and illustration, the study currently misses some crucial elements that would make it suitable for publication in Cryosphere:
The two selected sites, separated by ~70 km, are insufficient to evaluate a firn model that has the potential to be applied to the entire Greenland ice sheet. These sites are also inappropriate to evaluate dry compaction schemes (which is an entire section of the manuscript).
The evaluation of the infiltration scheme is done with the density profiles and firn air content which only give, on a few dates, an indirect idea of whether the model refreezes water at the appropriate depth. But there are other datasets that allow a more direct (and better resolved in time) evaluation of meltwater infiltration at DYE-2 and KAN_U, such as Time-Domain Reflectometry (https://doi.org/10.1029/2020GL089211, https://doi.org/10.1029/2021JF006295), firn temperature measurements (https://doi.org/10.18739/A2BN9X444, https://doi.org/10.1017/aog.2016.2, https://doi.org/10.22008/FK2/OVGQZ9, https://doi.org/10.22008/FK2/IW73UU) and up-GPR (https://doi.org/10.5194/tc-12-1851-2018, unfortunately no data available, but timing and infiltration's maximal depth can be inferred from the figures). Of course, not all of these observational datasets need to be used. Some comparisons, like for observed and modelled maximal depth of infiltration, can even be given in 1-2 sentences.
The manuscript currently gives a good presentation of the model performance at those two sites, but I did not see conclusions that could be directly used for other models or recommendations for new measurements to be conducted. In its current form, the relevance of the presented conclusions (only relevant for your model at two sites) is insufficient for publication in the Cryosphere. The sensitivity analysis could nevertheless give interesting insights into the choice of densification and infiltration schemes. Such insights should be properly highlighted in the abstract and conclusion.
The manuscript currently describes the entire model structure, making it difficult to identify the new additions made to the CFM. I suggest relying more on the CFM model description article to make the model description more concise (especially when a module is the standard one in CFM). Then more room would be left to explain why there was a need for a new infiltration scheme, what the new scheme is (this is currently described in the manuscript), and how it differs from the existing schemes in the CFM.
The simulation of firn is relevant to the broader cryospheric community because 1) it is necessary to convert remotely sensed height changes into mass changes, and 2) it is necessary to determine how much meltwater is sent to runoff within each grid cell of a gridded surface mass balance model (such as regional climate models). But the study does not present/evaluate the evolution of height change at these two sites (like for instance is done in https://doi.org/10.1029/2020GL088864 or https://doi.org/10.5194/tc-17-789-2023, see also surface height measurements in https://doi.org/10.22008/FK2/IW73UU), nor evaluate the amount of water that goes to runoff calculated by the model (see the range of modelled runoff in the RetMIP study, and the estimation of runoff at KAN_U in 2012 by https://doi.org/10.1038/nclimate2899). If the manuscript could give insights on those two valuable parameters and guidellines on how to properly simulate them, then two study sites could be sufficient.
In addition to these major comments, I listed specific questions or suggestions below.
I consider these shortcomings major, and a revised manuscript (with more test sites, a better evaluation of infiltration, and a broadened relevance through the focus on insights usable in other models) could be considered a completely new manuscript. Therefore, I am suggesting the rejection of the manuscript in its current form and encourage the resubmission of an improved manuscript with all the necessary elements.
Sincerely,
Baptiste Vandecrux---
L.118 "observed" I believe these are modelled values.
Section 2.3. When comparing firn cores to simulations, I don't understand why you need to gap-fill the observations. Couldn't comparison statistics be calculated only for depths where density observations are available? In particular, filling the top section of a density profile with an average value is inappropriate because it will also depend on the quality of the average value being used.
L.176-178 At which temporal resolution is the model run? I understood that there is the same amount of snow accumulating (being added at the surface) at each time step. Is that correct? If it is, then that is very far behind the current level of other snow and firn models (including the one usually coupled with RACMO) that now all include time-dependent accumulation.
L.183 "within each layer, firn properties are assumed to remain constant" How does the model then simulate transient events like the appearance and burial of an ice layer?
L.230 Patankar, 1989 seems to be for "engineering equipment." Please provide a reference more suitable for snow. Or just cite the CFM model description paper.
L.237-241 No need to give so much background about conductivity studies since it is not a key parameter for your study. And the second sentence is not true since Calonne et al. (2019) is well accepted and covers a good range of density and temperature.
Section 3.4: This densification scheme is usually referred to as the "overburden pressure" approach or more simply as the "densification scheme from Vionnet et al. (2012)." Please do not call it the "CROCUS model," which refers to the entire CROCUS snow model. But again, if your study doesn't rely on this precise scheme, it could be boiled down to one sentence.
L.265 Maybe I misunderstood something, but the irreducible water content is what snow can hold through capillary suction. So saying that water will always percolate due to gravitation is not true: there is no percolation until the layer's irreducible water content is filled.
L.276-277 I do not understand this last sentence. Why is the irreducible water amount different from the water holding capacity? The layer's water content is either below, at, or above the irreducible water content, determining how much water is available for percolation. I don't see where this product comes from.
L.279-287 Much clearer explanation than the previous paragraph.
Eq 11. I find the formulation difficult to understand. Consider using Vw(t, z) with t - dt as the value of the previous time step. It is also very confusing to formulate the liquid water content (LWC) of a layer based on the input at the surface and the LWC of all overlying layers. I would guess that this calculation is done iteratively from the top to bottom. So please make explicit what is done at each iteration: What is the LWC of a layer and water flux to the next layer as a function of the water flux from the layer above and the layer characteristics.
Eq. 13: Here you present Shimizu (1970), but line 272 you mention Calonne et al. (2019). Why not use the most recent work?
Eq. 14: Here you present van Genuchten (1980), but line 273 you mention Hirashima et al. (2010). Please mention "van Genuchten (1980) as implemented in Hirashima et al. (2010)" in line 273.L.339 In the equation at the start of the line, why is runoff mentioned but not refreezing?
Section 5.1: The comparison of your model with the previous model evaluations is a necessary step of the model evaluation. Such exercises should help identify key characteristics of your model and help understand its performance. In its current form, this long section repeates findings already presented in our RetMIP paper and in Verjans et al. (2019). This makes new insights about your model, and about firn modelling in general, more difficult to find. I strongly suggest rewriting this section to focus on 1) your model has very good metrics compared to previously published models, 2) can your model confirm or contradict high-level conclusions from Vandecrux et al. (2020) and Verjans et al. (2019), and 3) why do you think your model gives better performance than the best models in those two studies. No need to present models that were already shown as suboptimal in previous studies. Please mention that the surface forcing used by your model and those previously published model outputs are also different and may explain the difference in performance.
Section 5.2: This section could either be considered a sensitivity test or a preliminary assessment that would lead to the choice of the Vionnet et al. (2012) compaction scheme. In both cases, it should be introduced in the method section. Figure 8 could be moved to the supplementary or even removed because the main result (the scheme from Vionnet et al. (2012) works best) is already visible from Table 8 and the text. In the presentation of this exercise, in the method, the other models (HL, KM, LIG, SIM) should be concisely presented. Later in the presentation of the result or in a separate discussion section, please provide your explanation of why the densification scheme from Vionnet et al. (2012) outperforms the other schemes. I am also wondering whether this exercise is important at DYE-2 and KAN_U, where meltwater refreezing is controlling the density profile. The density profiles in Figure 8 do not show a great spread and the densification scheme may give similar results if their internal parameters are adjusted within their respective uncertainty range. This shows the lack of a proper assessment of densification at a site where the density profile is more dependent on the densification scheme. Please mention whether your results confirm or contradict conclusions from https://doi.org/10.1017/jog.2016.114.
Section 5.3: I do not understand the difference between the meltwater infiltration scheme described in Section 3.5 and what is called here "Darcy flow scheme." I thought this study introduced Darcy flow into the CFM (e.g., your eq 12 in section 3.5.2). Just like for the densification schemes, the infiltration schemes tested here should be properly described in the methods (e.g., which irreducible water content is used for the bucket scheme). Please try to explain the differences between the different schemes (choice of parameter values within each scheme, conceptual description of infiltration, etc.). I am quite surprised that the density profiles in 2017 (or 2013 for KAN_U) are so close for infiltration schemes radically different (in particular Darcy vs. bucket). Did they go through the same spin-up procedure? At DYE-2, they produce the same peak at 4 m and the same step at 10-11 m.
Section 5.4:
- "negative snowfall amounts occur within certain time steps in the RACMO2.3p2 model, which are ignored in the time-driven method." In addition to wind erosion, these periods could also correspond to periods of sublimation. These are true processes that are handled by most existing snow and firn models. They should not be ignored because they can become important in certain regions/years.
- Please give the time step eventually used for the accumulation-driven method. For the time-driven run, you mention that "Stevens et al. (2020) ran the CFM model with monthly time steps," but you don't state which time step you used for your model run. Please make it clear.
- If the purpose of the manuscript is to present an updated infiltration scheme that is implemented within the CFM, then this new infiltration scheme should be used in the time step sensitivity analysis.To ensure the reproducibility of your study, it is highly recommended to make your model code available, and since it feeds into the Community Firn Model, it could take te form of a branch or a pull request of the CFM with the new infiltration module.
Citation: https://doi.org/10.5194/egusphere-2024-1726-RC1 -
RC2: 'Comment on egusphere-2024-1726', Charles Amory, 27 Sep 2024
This article presents the developments made to the Community Firn Model, notably in terms of water percolation and densification processes, and its application at two nearby sites in the percolation zone of the Greenland ice sheet. The model is evaluated against observed density profiles (and derived firn air contents), and its performance is compared to other firn models, although with a different atmospheric forcing and initialization procedure.
Despite a thorough methodology and high-quality illustrations that generally support the conclusions, I regret that the paper (although elegantly written) only tells part of the (promising) story, which could be expanded without too much additional effort. The main issue is that the model evaluation does not sufficiently demonstrate its improved capacity to represent Greenland’s firn characteristics. The field data used for evaluation appear quite limited compared to the broader set of observations available in the literature (see, for instance, the evaluation done in Vandecrux et al., 2020), which would be necessary to fully assess the benefits of the described implementations. In its current form, the study only addresses density profiles observed at two sites typical of Greenland’s percolation zone. However, the model could be more comprehensively assessed by applying the same methodology, incorporating additional observations, and including site(s) with different climatic conditions which are representative of the diversity encountered across the Greenland ice sheet and which we expect such a firn model to capture. Specifically, the additional analysis should include at least:
- An evaluation of the temperature: Since temperature influences firn cold content (and thus refreezing) and densification rates, it is essential to verify that the model simulates accurate density profiles for the right reasons.
- An evaluation of density and temperature profiles at a dry snow site: This would allow for a more appropriate assessment of the model's ability to simulate densification processes in the absence of liquid water. Summit, where such data is available (Vandecrux et al., 2020), seems like a suitable candidate.
Applying the same methodology and rigor to additional sites and relevant observations would significantly improve the manuscript's scope and significance and help it fully meet the requirements for publication in The Cryosphere. I look forward to seeing the improved version !
Good luck,
Charles Amory
Vandecrux, B., Mottram, R., Langen, P. L., Fausto, R. S., Olesen, M., Stevens, C. M., Verjans, V., Leeson, A., Ligtenberg, S., Kuipers Munneke, P., Marchenko, S., van Pelt, W., Meyer, C. R., Simonsen, S. B., Heilig, A., Samimi, S., Marshall, S., Machguth, H., MacFerrin, M., Niwano, M., Miller, O., Voss, C. I., and Box, J. E.: The firn meltwater Retention Model Intercomparison Project (RetMIP): evaluation of nine firn models at four weather station sites on the Greenland ice sheet, The Cryosphere, 14, 3785–3810, https://doi.org/10.5194/tc-14-3785-2020, 2020.
Citation: https://doi.org/10.5194/egusphere-2024-1726-RC2
Status: closed
-
RC1: 'Comment on egusphere-2024-1726', Baptiste Vandecrux, 25 Jul 2024
Review of Modelling Firn Density at Dye-2 and KAN_U, Two Sites in the Percolation Zone of the Greenland Ice Sheet
by Dr. Xueyu Zhang et al.This study uses the Community Firn Model (CFM) and RACMO2.3p2 to simulate the evolution at two sites on the Greenland ice sheet. The CFM is improved with a new, Darcy-like infiltration scheme. Despite the (much appreciated) comparison of the model results to previous firn model intercomparison exercises and the care given to the writing and illustration, the study currently misses some crucial elements that would make it suitable for publication in Cryosphere:
The two selected sites, separated by ~70 km, are insufficient to evaluate a firn model that has the potential to be applied to the entire Greenland ice sheet. These sites are also inappropriate to evaluate dry compaction schemes (which is an entire section of the manuscript).
The evaluation of the infiltration scheme is done with the density profiles and firn air content which only give, on a few dates, an indirect idea of whether the model refreezes water at the appropriate depth. But there are other datasets that allow a more direct (and better resolved in time) evaluation of meltwater infiltration at DYE-2 and KAN_U, such as Time-Domain Reflectometry (https://doi.org/10.1029/2020GL089211, https://doi.org/10.1029/2021JF006295), firn temperature measurements (https://doi.org/10.18739/A2BN9X444, https://doi.org/10.1017/aog.2016.2, https://doi.org/10.22008/FK2/OVGQZ9, https://doi.org/10.22008/FK2/IW73UU) and up-GPR (https://doi.org/10.5194/tc-12-1851-2018, unfortunately no data available, but timing and infiltration's maximal depth can be inferred from the figures). Of course, not all of these observational datasets need to be used. Some comparisons, like for observed and modelled maximal depth of infiltration, can even be given in 1-2 sentences.
The manuscript currently gives a good presentation of the model performance at those two sites, but I did not see conclusions that could be directly used for other models or recommendations for new measurements to be conducted. In its current form, the relevance of the presented conclusions (only relevant for your model at two sites) is insufficient for publication in the Cryosphere. The sensitivity analysis could nevertheless give interesting insights into the choice of densification and infiltration schemes. Such insights should be properly highlighted in the abstract and conclusion.
The manuscript currently describes the entire model structure, making it difficult to identify the new additions made to the CFM. I suggest relying more on the CFM model description article to make the model description more concise (especially when a module is the standard one in CFM). Then more room would be left to explain why there was a need for a new infiltration scheme, what the new scheme is (this is currently described in the manuscript), and how it differs from the existing schemes in the CFM.
The simulation of firn is relevant to the broader cryospheric community because 1) it is necessary to convert remotely sensed height changes into mass changes, and 2) it is necessary to determine how much meltwater is sent to runoff within each grid cell of a gridded surface mass balance model (such as regional climate models). But the study does not present/evaluate the evolution of height change at these two sites (like for instance is done in https://doi.org/10.1029/2020GL088864 or https://doi.org/10.5194/tc-17-789-2023, see also surface height measurements in https://doi.org/10.22008/FK2/IW73UU), nor evaluate the amount of water that goes to runoff calculated by the model (see the range of modelled runoff in the RetMIP study, and the estimation of runoff at KAN_U in 2012 by https://doi.org/10.1038/nclimate2899). If the manuscript could give insights on those two valuable parameters and guidellines on how to properly simulate them, then two study sites could be sufficient.
In addition to these major comments, I listed specific questions or suggestions below.
I consider these shortcomings major, and a revised manuscript (with more test sites, a better evaluation of infiltration, and a broadened relevance through the focus on insights usable in other models) could be considered a completely new manuscript. Therefore, I am suggesting the rejection of the manuscript in its current form and encourage the resubmission of an improved manuscript with all the necessary elements.
Sincerely,
Baptiste Vandecrux---
L.118 "observed" I believe these are modelled values.
Section 2.3. When comparing firn cores to simulations, I don't understand why you need to gap-fill the observations. Couldn't comparison statistics be calculated only for depths where density observations are available? In particular, filling the top section of a density profile with an average value is inappropriate because it will also depend on the quality of the average value being used.
L.176-178 At which temporal resolution is the model run? I understood that there is the same amount of snow accumulating (being added at the surface) at each time step. Is that correct? If it is, then that is very far behind the current level of other snow and firn models (including the one usually coupled with RACMO) that now all include time-dependent accumulation.
L.183 "within each layer, firn properties are assumed to remain constant" How does the model then simulate transient events like the appearance and burial of an ice layer?
L.230 Patankar, 1989 seems to be for "engineering equipment." Please provide a reference more suitable for snow. Or just cite the CFM model description paper.
L.237-241 No need to give so much background about conductivity studies since it is not a key parameter for your study. And the second sentence is not true since Calonne et al. (2019) is well accepted and covers a good range of density and temperature.
Section 3.4: This densification scheme is usually referred to as the "overburden pressure" approach or more simply as the "densification scheme from Vionnet et al. (2012)." Please do not call it the "CROCUS model," which refers to the entire CROCUS snow model. But again, if your study doesn't rely on this precise scheme, it could be boiled down to one sentence.
L.265 Maybe I misunderstood something, but the irreducible water content is what snow can hold through capillary suction. So saying that water will always percolate due to gravitation is not true: there is no percolation until the layer's irreducible water content is filled.
L.276-277 I do not understand this last sentence. Why is the irreducible water amount different from the water holding capacity? The layer's water content is either below, at, or above the irreducible water content, determining how much water is available for percolation. I don't see where this product comes from.
L.279-287 Much clearer explanation than the previous paragraph.
Eq 11. I find the formulation difficult to understand. Consider using Vw(t, z) with t - dt as the value of the previous time step. It is also very confusing to formulate the liquid water content (LWC) of a layer based on the input at the surface and the LWC of all overlying layers. I would guess that this calculation is done iteratively from the top to bottom. So please make explicit what is done at each iteration: What is the LWC of a layer and water flux to the next layer as a function of the water flux from the layer above and the layer characteristics.
Eq. 13: Here you present Shimizu (1970), but line 272 you mention Calonne et al. (2019). Why not use the most recent work?
Eq. 14: Here you present van Genuchten (1980), but line 273 you mention Hirashima et al. (2010). Please mention "van Genuchten (1980) as implemented in Hirashima et al. (2010)" in line 273.L.339 In the equation at the start of the line, why is runoff mentioned but not refreezing?
Section 5.1: The comparison of your model with the previous model evaluations is a necessary step of the model evaluation. Such exercises should help identify key characteristics of your model and help understand its performance. In its current form, this long section repeates findings already presented in our RetMIP paper and in Verjans et al. (2019). This makes new insights about your model, and about firn modelling in general, more difficult to find. I strongly suggest rewriting this section to focus on 1) your model has very good metrics compared to previously published models, 2) can your model confirm or contradict high-level conclusions from Vandecrux et al. (2020) and Verjans et al. (2019), and 3) why do you think your model gives better performance than the best models in those two studies. No need to present models that were already shown as suboptimal in previous studies. Please mention that the surface forcing used by your model and those previously published model outputs are also different and may explain the difference in performance.
Section 5.2: This section could either be considered a sensitivity test or a preliminary assessment that would lead to the choice of the Vionnet et al. (2012) compaction scheme. In both cases, it should be introduced in the method section. Figure 8 could be moved to the supplementary or even removed because the main result (the scheme from Vionnet et al. (2012) works best) is already visible from Table 8 and the text. In the presentation of this exercise, in the method, the other models (HL, KM, LIG, SIM) should be concisely presented. Later in the presentation of the result or in a separate discussion section, please provide your explanation of why the densification scheme from Vionnet et al. (2012) outperforms the other schemes. I am also wondering whether this exercise is important at DYE-2 and KAN_U, where meltwater refreezing is controlling the density profile. The density profiles in Figure 8 do not show a great spread and the densification scheme may give similar results if their internal parameters are adjusted within their respective uncertainty range. This shows the lack of a proper assessment of densification at a site where the density profile is more dependent on the densification scheme. Please mention whether your results confirm or contradict conclusions from https://doi.org/10.1017/jog.2016.114.
Section 5.3: I do not understand the difference between the meltwater infiltration scheme described in Section 3.5 and what is called here "Darcy flow scheme." I thought this study introduced Darcy flow into the CFM (e.g., your eq 12 in section 3.5.2). Just like for the densification schemes, the infiltration schemes tested here should be properly described in the methods (e.g., which irreducible water content is used for the bucket scheme). Please try to explain the differences between the different schemes (choice of parameter values within each scheme, conceptual description of infiltration, etc.). I am quite surprised that the density profiles in 2017 (or 2013 for KAN_U) are so close for infiltration schemes radically different (in particular Darcy vs. bucket). Did they go through the same spin-up procedure? At DYE-2, they produce the same peak at 4 m and the same step at 10-11 m.
Section 5.4:
- "negative snowfall amounts occur within certain time steps in the RACMO2.3p2 model, which are ignored in the time-driven method." In addition to wind erosion, these periods could also correspond to periods of sublimation. These are true processes that are handled by most existing snow and firn models. They should not be ignored because they can become important in certain regions/years.
- Please give the time step eventually used for the accumulation-driven method. For the time-driven run, you mention that "Stevens et al. (2020) ran the CFM model with monthly time steps," but you don't state which time step you used for your model run. Please make it clear.
- If the purpose of the manuscript is to present an updated infiltration scheme that is implemented within the CFM, then this new infiltration scheme should be used in the time step sensitivity analysis.To ensure the reproducibility of your study, it is highly recommended to make your model code available, and since it feeds into the Community Firn Model, it could take te form of a branch or a pull request of the CFM with the new infiltration module.
Citation: https://doi.org/10.5194/egusphere-2024-1726-RC1 -
RC2: 'Comment on egusphere-2024-1726', Charles Amory, 27 Sep 2024
This article presents the developments made to the Community Firn Model, notably in terms of water percolation and densification processes, and its application at two nearby sites in the percolation zone of the Greenland ice sheet. The model is evaluated against observed density profiles (and derived firn air contents), and its performance is compared to other firn models, although with a different atmospheric forcing and initialization procedure.
Despite a thorough methodology and high-quality illustrations that generally support the conclusions, I regret that the paper (although elegantly written) only tells part of the (promising) story, which could be expanded without too much additional effort. The main issue is that the model evaluation does not sufficiently demonstrate its improved capacity to represent Greenland’s firn characteristics. The field data used for evaluation appear quite limited compared to the broader set of observations available in the literature (see, for instance, the evaluation done in Vandecrux et al., 2020), which would be necessary to fully assess the benefits of the described implementations. In its current form, the study only addresses density profiles observed at two sites typical of Greenland’s percolation zone. However, the model could be more comprehensively assessed by applying the same methodology, incorporating additional observations, and including site(s) with different climatic conditions which are representative of the diversity encountered across the Greenland ice sheet and which we expect such a firn model to capture. Specifically, the additional analysis should include at least:
- An evaluation of the temperature: Since temperature influences firn cold content (and thus refreezing) and densification rates, it is essential to verify that the model simulates accurate density profiles for the right reasons.
- An evaluation of density and temperature profiles at a dry snow site: This would allow for a more appropriate assessment of the model's ability to simulate densification processes in the absence of liquid water. Summit, where such data is available (Vandecrux et al., 2020), seems like a suitable candidate.
Applying the same methodology and rigor to additional sites and relevant observations would significantly improve the manuscript's scope and significance and help it fully meet the requirements for publication in The Cryosphere. I look forward to seeing the improved version !
Good luck,
Charles Amory
Vandecrux, B., Mottram, R., Langen, P. L., Fausto, R. S., Olesen, M., Stevens, C. M., Verjans, V., Leeson, A., Ligtenberg, S., Kuipers Munneke, P., Marchenko, S., van Pelt, W., Meyer, C. R., Simonsen, S. B., Heilig, A., Samimi, S., Marshall, S., Machguth, H., MacFerrin, M., Niwano, M., Miller, O., Voss, C. I., and Box, J. E.: The firn meltwater Retention Model Intercomparison Project (RetMIP): evaluation of nine firn models at four weather station sites on the Greenland ice sheet, The Cryosphere, 14, 3785–3810, https://doi.org/10.5194/tc-14-3785-2020, 2020.
Citation: https://doi.org/10.5194/egusphere-2024-1726-RC2
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 210 | 71 | 23 | 304 | 12 | 15 |
- HTML: 210
- PDF: 71
- XML: 23
- Total: 304
- BibTeX: 12
- EndNote: 15
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1