the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 08 May 2025
Datasets and protocols for including anomalous freshwater from melting ice sheets in climate simulations
Abstract. Anomalous freshwater fluxes from the Greenland and Antarctic ice sheets and ice shelves are impacting the surrounding oceans and we need to be able to account for these effects in climate model simulations over the historical period and beyond. In previous phases of the Coupled Model Intercomparison Project (CMIP), models mostly either assumed that the ice sheets were in mass balance, or that discharge from the ice sheets was constant, but in neither case was the observed increasing discharge properly accounted for. In this paper, we present an updateable dataset of absolute and anomalous freshwater mass fluxes from both ice sheets. These fluxes can be implemented in historical climate simulations as a forcing for models that do not (yet) include interactive ice sheets, or used to evaluate models that do. We also make recommendations for how climatological and anomalous fluxes can be implemented in climate models that may have different approaches to interactions with the ice sheets. These forcings are available for CMIP7 simulations and should lead to more robust and coherent simulation of sea surface temperature, sea ice and regional sea level trends in the recent historical period, and improve the credibility of projections.
-
Notice on discussion status
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
-
Preprint
(2459 KB)
-
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(2459 KB) - Metadata XML
- BibTeX
- EndNote
- Final revised paper
Journal article(s) based on this preprint
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2025-1940', John Dunne, 24 May 2025
- AC1: 'Reply on RC1', Gavin A. Schmidt, 02 Aug 2025
-
RC2: 'Comment on egusphere-2025-1940', Cecilia Bitz, 13 Jun 2025
This paper provides a valuable service by outlining the range of ways that ESMs model the effects of ice sheets and glaciers and by constructing datasets and recommendations to deal with their mass and energy imbalance. There are many positive aspects, like the community effort and presumably consensus behind the work. The list of definitions at the start is very good too - though I suggest adding Surface Mass Balance (SMB) to it. I am grateful to know where these authors "judge" something to be true or offer views. Some of this information would be impossible to know for certain since not all modeling centers answered their survey. The authors provide valuable guidance about simulation assumptions and practices with a summary of choices made by modeling centers with some indication of the consequences. This is helpful for model developers and those analyzing models alike.
Please considered the following specific comments:
1) There are a few occasions when the manuscript needlessly goes into minor issues about conservation. For example, I think the point made about moving snow down hill at line 166 is ridiculous, or needs further explanation. We don't worry about the conversion of PE to heat by raindrops when they hit the ground so why should we worry about the converstion of PE to heat when snow moves downhill. Further if it amounts to 14C it is not a wrinkle that might make a small difference, so something is awry about this paragraph. Another example is at Line 461-3 with regard to icebergs possibly melting a bit due to geothermal heating, etc. While all this may be true, there is little point in itemizing a bunch of stuff that no ESM developer would bother with. It makes the paper longer than it need be.
2) It is good to see some discussion of how transient additions of freshwater will alter the ocean salinity and sea level. Yet, I thought this could be a teachable moment to explain the pros and cons of adding freshwater in volume vs mass conserving ocean models (rather than an oblique reference to there being a difference at line 162). I was intrigued by the discussion that volume conserving models would overestimate sea level changes since most recent ice loss has been from floating ice and models don't do hydrostatic balance of ice correctly. It seems like this should be elevated to a recommendation at the end for modelers to work on.
3) In Fig 1 the Antarctic side shouldn't have surface runoff and much ablation zone at the surface. It is strange to see so much of Greenland below sea level and so little of Antarctica. There are elements that seems reversed between Greenland and Antarctica. Is this really meant to be the final version of this figure?
4) The figure captions could be improved with citations of sources and a bit more explanation of what the reader should glean from them.
An example is Figs 4 & 5 where weighting functions are not defined either in the captions of main text. I would have thought a weighting function would be unitless. The caption should say where such maps are available. A few pages further into reading the manuscript I see the term "Iceber met maps" in the main text but without reference to Figs 4 & 5. Given the units are the same I'm guessing this is what is plotted in Figs 4&5. If so, be sure to reference the figures, use the correct term in the caption, and state in the caption that this is one of your datasets.
Another is Fig 3 which describes data with no citation. Is this one of the products that the authors are providing? If so, I'm concerned about the sudden switch in methods and variability at year 1986 since I expect some users will just prescribe this as forcing to their model without being aware of it and then "discover" an amazing regime change in their model in 1986, similar to the many discoveries in AMIP runs forced with sea ice concentration from HADISST. I recommend providing two seperate products: One of the longer period with the same method throughout and a shorter one since 1986. (Again a few pages further into reading the manuscript I see Fig 3 discussed in the main text, which is good. Please at least say in the caption that this is one of your datasets though.)
5) Figure 6. I don't get why the circled 2 is called an "implicit FW" flux and the term "implicit SMB" is used at line 504. What is meant by "implicit" here and elsewhere (the term is used a lot)? It is supposed to mean implied, but I think maybe it is being used to mean derived or diagnosed from other quantities. It seems to me that no prognostic calculations in a model are implicit, so I can't grasp what implicit is supposed to mean. Also in Figure 6, I would think circled 2 should be labeled SMB rather than FW flux since FW flux is not as specific. The lower panel's cartoon of ice sheet to ocean fluxes is unclear and seems inconsistent with the caption. A cartoon should be an aid not a head scratcher.
6) Lines 111-114 are too hard to follow. Break this up into more sentences and or include a table/equations.
7) Line 138-9 Are you saying that some models just disappear the approx 3300 Gt/yr of accumulation on Antarctica? I thought models at least dropped the mass/volume of water into the ocean, as discussed a few paragraphs later. I've never heard of a credible ESM that didn't conserve freshwater at this level at least.
8) The manuscript jumps between modeling approaches and dataset details a bit in somewhat confusing ways. For example, line 379 is probably about a dataset provided, or possibly it is about a particular model. Either way the text should clarify.
Citation: https://doi.org/10.5194/egusphere-2025-1940-RC2 - AC2: 'Reply on RC2', Gavin A. Schmidt, 02 Aug 2025
-
CEC1: 'Comment on egusphere-2025-1940', Juan Antonio Añel, 20 Jun 2025
Dear authors,
Unfortunately, after checking your manuscript, it has come to our attention that it does not comply with our "Code and Data Policy".
https://www.geoscientific-model-development.net/policies/code_and_data_policy.html
You have archived your code on GitHub. However, GitHub is not a suitable repository for scientific publication. GitHub itself instructs authors to use other long-term archival and publishing alternatives, such as Zenodo. Therefore, the current situation with your manuscript is irregular. Please, publish your code in one of the appropriate repositories and reply to this comment with the relevant information (link and a permanent identifier for it (e.g. DOI)) as soon as possible, as we can not accept manuscripts in Discussions that do not comply with our policy.I must note that if you do not fix this problem, we will have to reject your manuscript for publication in our journal.
Also, you must include a modified 'Code and Data Availability' section in a potentially reviewed manuscript, containing the DOI of the new repositories. Also, in the GitHub site no license is listed for the code. If you do not include a license, the code remains your property and it is not possible to run or distribute it. Therefore, when uploading the model's code to Zenodo, you could want to choose a free software/open-source (FLOSS) license.
Juan A. Añel
Geosci. Model Dev. Executive EditorCitation: https://doi.org/10.5194/egusphere-2025-1940-CEC1 -
CC1: 'Reply on CEC1', Ken Mankoff, 20 Jun 2025
Dear Juan,
I have added a CC0 license to the GitHub repository. I have drafted a release (https://github.com/NASA-GISS/freshwater-forcing-workshop/releases/tag/v0.0.1-GMD-review ) and linked it with Zenodo which will archive and issue a DOI, but it has not yet been pulled by Zenodo or appears there.
Citation: https://doi.org/10.5194/egusphere-2025-1940-CC1 -
CC2: 'Reply on CC1', Ken Mankoff, 20 Jun 2025
DOI for code is now at: https://doi.org/10.5281/zenodo.15707384
Citation: https://doi.org/10.5194/egusphere-2025-1940-CC2 -
CEC2: 'Reply on CC2', Juan Antonio Añel, 21 Jun 2025
Dear authors,
Many thanks for addressging this issue so quickly.
Juan A. Añel
Geosci. Mode Dev. Executive Editor
Citation: https://doi.org/10.5194/egusphere-2025-1940-CEC2
-
CEC2: 'Reply on CC2', Juan Antonio Añel, 21 Jun 2025
-
CC2: 'Reply on CC1', Ken Mankoff, 20 Jun 2025
-
CC1: 'Reply on CEC1', Ken Mankoff, 20 Jun 2025
-
RC3: 'Comment on egusphere-2025-1940', Anonymous Referee #3, 28 Jun 2025
Reviewer comments
“Datasets and protocols for including anomalous freshwater from melting ice sheets in climate simulations” By Schmidt et al. provides valuable transient forcing of ice sheets freshwater with protocol recommendations for modellers implementing the dataset consistently. This work certainly makes an important contribution to both the Earth system modelling and ice sheet modelling communities
Please consider my suggested revisions below:
- Missing figures showing the timeseries of each freshwater product listed in Section 3.3. These products are the key output of the paper. However, the features of some products are not presented and discussed. Figure 3-5 do this for Greenland ice discharge and iceberg maps, although Figure 5 is not cited in the main text. It would be good to include figures and discussions for the remaining products.
- Missing evaluations and discussions of how well this transient freshwater forcing and suggested modelling approaches improves the representation of ocean circulation, salinity, sea level, etc., in model simulations. The modelling approaches in the two runs are described in Section 4 (Pre-industrial control runs and historical simulations); however, evaluations of the methods and products are not provided. I appreciate the authors’ time and effort, it would be good to include at least some preliminary modelled results using freshwater forcing from this work, along with comparisons to other studies.
See specific comments below:
L5 – The abstract does not mention projections. Recommend clarifying that the proposed protocol relates solely to historical simulations but that freshwater forcing is likely to become very much more important in future projection scenarios.
L9 – Can the authors confirm that freshwater forcing will be included in the forcings for the CMIP7 historical experiments? CMIP7 is now very close to releasing this forcing, and it is unclear whether freshwater is actually included.
L21 – Not convinced that computational expenses have hamper the incorporation of ice sheets into ESM. They are by comparison to other elements of the climate system cheap.
Providing historical freshwater forcing is a step forward but it does introduce an issue around projections. What are ESMs supposed to do at the start of the projection period?
L42 – There is no Appendix B.
L60 – Not convinced that these definitions are helpful. Are there not standardized definitions (CMIP or ISMIP6/7) that could be used? Runs the risk of making what is a fairly confused set of terms worse. Definitions could be improved by adding relevant units.
L65 – This is confusing and does not match the definition of discharge immediately above (in which discharge sits between ice sheets and ice shelves, while here discharge is after ice shelves before calving).
L67 – Upstream of grounding line? What about ice floating on subglacial lakes? This is upstream but not grounded.
L81 – Runoff is not purely a mass loss process if it includes rainfall (and melt of seasonal snow).
This section may be helped by linking the definition to Figure 1 more tightly. At present, terms in Figure 1 do not match the definitions (eg basal melt) and vice vera (eg discharge).
L97 onwards – Defining symbols for these various quantities and how they relate to one another would be helpful here. There are a number of things going on in the section (baseline/anomaly, time period of baseline, separation of freshwater into various fluxes such as grounded, shelf melt and shelf calving) that need to be explained more clearly. There is also some danger in referring to ‘ice mass anomalies’ because terms such as this are often used in ice mass budget calculations (freshwater fluxes are related primarily to the loss terms where mass budget relates to the net of gain and loss).
L100 – Needs to be made clear that GRACE does not directly measure freshwater flux.
L104 – Figure 3 is cited before Figure 2.
L105 – This seems to be very detailed and relate to the way in which the forcing has been calculated. Would be better in Section 3 or 4 rather than Introduction.
L118 – Good to state that basal melt of grounded ice not included in the assessment however requires some justification to demonstrate, for instance, that this flux is much smaller than the other freshwater fluxes leaving the ice mass.
L124 – quasi.
L153 – This is too vague – this assertion needs to be supported by references.
L159 – Use of the term ‘Discharge’ here and in next paragraph confusing. Does this refer to the previous definition of ‘discharge’ or discharge of freshwater? If the latter, then it is not clear that the ESM calculations discussed here would explicitly calculate discharge (line 171).
L200 – Define what is meant by ‘regional flow’ here. Will this be done on the basis of sectors/basins as per IMBIE etc?
L203 – based on Rignot et al. (2013) and Mouginot and Rignot (2019).
L210 – Earlier grounded basal melt was explicitly disregarded.
L213 – Clarify differences between refreezing and retention. Retention on what timescales?
L223 – Why choose flux gates 5 km upstream from the grounding line?
L242 – Clarification required here – grounding line retreat in itself is not a source of freshwater (increased freshwater flux and GL retreat are both consequences of increased melt for instance). Repeated confusion around inclusion of grounded ice melt.
L245 – ‘Surface runoff’ Ideally all terms of this type should be defined in Section 1.1.
L254 – Figures 3 and 4 are not cited in this section. The section does not include any discussion about the melt of ice bergs (i.e., Figure 4).
The discussion on p5 about baselines and anomalies is not mentioned at all. It is unclear at this stage what form the freshwater forcing will be take.
L262 – This is confusing Antarctic SMB has little to do with freshwater release from Antarctica and it is not clear what information on freshwater can be gleaned from trends in SMB.
L266+271 – There are a few steps between the observations mentioned in the preceding paragraph and ‘good estimates of Antarctic freshwater fluxes’ that maybe worth explaining.
L273 – Similar issue - sea level rise (ie net mass budget) is not directly linked to freshwater discharge (which is primarily to do with mass loss only).
L289 – Is the ratio by each basin or the entire Antarctica?
L297 – It would be help to have information on units here, as well as temporal resolution etc. Also it is unclear what the relationship is between the various products. For instance, were discharges and runoff used to calculate the freshwater anomalies for Greenland? Similarly, for Antarctica was calving and submarine melt used to find freshwater anomaly? How? A conceptual figure may help here (also note use of products in protocol mentioned below).
How is this related to the iceberg melt maps?
These products are the key output of the paper. It would be useful to have some plots showing key features. Figures 4 and 5 do this for iceberg melt although they are not cited in this section.
It would also be helpful to add some explanation of how each of the products is used in the protocols described in the rest of the paper and whether they are used as anomalies or absolute values. The two runs described in Section 4 are pre-industrial and historical: can the authors indicate which product is used when?
L351 – Tracer is not previously discussed.
L353 – ‘Relaxation time constant’ seems to imply a specific way of incorporating freshwater into the pre-industrial but it is unclear what this method would actually be. More detail required.
L396 onwards – This appears to be a discussion around the fine-scale physics governing local redistribution of freshwater in fjord and around glacier fronts. Not clear how it relates to the implementation of the products in ESMs. Perhaps this text should be moved to an early section?
L421 – It would be good to have more discussion on the differences/consequences of the injection depth (surface vs depth) of freshwater from icebergs and ice shelves melting in the Southern Ocean.
L429 – Not clear why the depth range of 130-230 m is suggested? ‘130 m is the mean depth of current ice shelf fronts in Antarctica.’, but where does the number ‘230 m’ come from?
L434 – Please clarify it is the configuration for the Greenland iceberg model? I don't think it is what was used in Mathiot and Jourdain (2023).
L436 – The translation of iceberg fluxes to spatially variable melt rates is a key component of this paper and is lost here in the middle of a very detailed section. This important material would be better in its own subsection positioned to accompany Figures 4 and 5 perhaps between 3.2 and 3.3. Note that this text does not cite Figure 5. Currently, this text is located in the pre-industrial subsection although it would appear to be more generic than this and would be relevant to all time periods?
L456 – Again, this subsection appears to contain a generic discussion around freshwater and energy. It is not clear why it appears under the pre-industrial.
L481 – Although this subsection is about the historical, the approach described (and shown in Figure 6) would appear to be applicable to projections as well (assuming suitable freshwater forcing was available).
L487 – May be mixing papers on historical and projection scenarios here.
L493 – It is more consistent to use 'Antarctica and Greenland' rather than 'SH and NH', as mountain glaciers are also included in SH and NH. Moreover, SH or NH acronym is not defined.
L508 – It is not clear how this discussion relates to the specific products that have been created. Which of the products are used and which are not? For instance, do the iceberg melt maps play a role?
L589 – Worth pointing out that using this protocol for the historical implies that some extension of the freshwater forcing is required for projection experiments. Otherwise, there would be an abrupt change in forcing at his start of the projection period. An alternative might be to maintain the historical freshwater forcing into the projection period which would avoid a hiatus but become increasing unrealistic as (presumably) freshwater fluxes from the ice sheets increase into the future.
L590 – Would it be possible to use this protocol with the variables supplied by ISMIP6? What additional work would be required? Presumably, as a minimum the iceberg melt patterns would need to be calculated for time in the future.
L667 – This is a bit misleading. The products for Antarctica do not cover the period before 1991.
Figures
Figure 1. Indication of scale required for all panels.
Depiction of runoff across and through the ice sheet on the Antarctica side of the diagram is misleading and should be on the Greenland side. If the diagram is meant to represent the contemporary ice sheets then presence of a substantial ablation zone in Antarctica is also extremely misleading.
Figure 4 – Please clarify how the iceberg melt rates are weighted in the caption and the main context. The same as Figure 5.
Figure 5 – What are the white lines in front of the Ross Ice Shelf in the subpanels [11] and [12]? Are they NaNs, or are they due to the grid mesh? If the latter, why are there no white lines in, for example, the panel [9], [10] and All?
Citation: https://doi.org/10.5194/egusphere-2025-1940-RC3 - AC3: 'Reply on RC3', Gavin A. Schmidt, 02 Aug 2025
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2025-1940', John Dunne, 24 May 2025
The manuscript “Datasets and protocols for including anomalous freshwater from melting ice sheets in climate simulations” by Schmidt et al provides an interdisciplinary assessment of the state of understanding, uncertainties, and many technical issues involved in representing the transient liquid and solid freshwater forcing from the Greenland and Antarctic ice sheets in coupled climate Earth system models as efforts to couple these models with ice sheet models continue to advance. It certainly makes the case that this is a very technically complex problem and progress on the interim solution of applying freshwater anomalies is an important step in improving the representativeness of the ocean circulation and climate responses to anthropogenic climate change in general and the recent observed changes to ice sheet mass balance. My main issues with the present version are:
- The fluxes for Antarctica equivalent to Figure 3 for Greenland are not presented
- The overall value of representing these fluxes to improve representation of sea level rise is not provided… how much of non-steric sea level rise will this provide?
- There are many occasions of handwavy statements of the status quo being insufficient without clear recommendations to relieve the problem. I have pointed these instances out and made suggestions.
See technical suggestions below.
Line 2 – add comma before “and”
Line 5 – add “over the historical period” after “discharge”, replace “accounted for” with “addressed” or “incorporated” or “represented”, and remove “an updateable dataset of”
Line 14 – doesn’t “over the last century” include “and in recent decades”? Suggest removing or changing to “and has accelerated in recent decades”
Line 30 – Another example of the role of freshwater on Southern Ocean circulation
Bronselaer, B., Winton, M., Griffies, S.M., Hurlin, W.J., Rodgers, K.B., Sergienko, O.V., Stouffer, R.J. and Russell, J.L., 2018. Change in future climate due to Antarctic meltwater. Nature, 564(7734), pp.53-58.
line 63 - There does not seem to be consistency between the definition of "Discharge" here (as an ice flux) compared to the terms in Figure 1 (e.g. "Discharge" not identified but "Subglacial Discharge" seemingly identified as a liquid flux)… is “Discharge” equal to “Iceberg Flux” or to to the sum of several terms in Figure 1?
Line 72 – “Ice front” is not provided in Figure 1, but “Frontal retreat” is provided twice.
Line 74 – “Ice shelf” is not provided in Figure 1
Line 114 – “, and again this is with respect” should be “relative”
Line 124 – “uasi” should be “quasi”
Line 129 – I don’t think the sentence, “The implications of adding a new forcing dataset needs to be considered for each of these different configurations and experiments.” Is helpful without further contextualization and should be removed unless these implications are to be detailed.
Line 160 – The GFDL CMIP6 models included explicit icebergs: Adcroft, A., Anderson, W., Balaji, V., Blanton, C., Bushuk, M., Dufour, C.O., Dunne, J.P., Griffies, S.M., Hallberg, R., Harrison, M.J. and Held, I.M., 2019. The GFDL global ocean and sea ice model OM4. 0: Model description and simulation features. Journal of Advances in Modeling Earth Systems, 11(10), pp.3167-3211.
Line 181 – add comma before “it”
Line 185 – “observational changes” should be “observations” to avoid repeating “changes”
Line 295 – Before moving on, it would be helpful to know how these freshwater fluxes compare to those in Swart et al., 2023 for SOFIA which is cited earlier.
Line 310-312 – Is “with respect to the 1850–1900 pre-industrial period” mean that 1850-1900 is the “baseline period”? I think so from line 102, but it is not clear why different wording is being used for these two things if they are indeed the same thing.
326 – remove “the choices available in”
Line 402 – “1.5 orders of magnitude” should be “50 times”
Line 404 – remove “, such as that provided by”
Line 412 – a good reference for “estuarine box model” is
Sun, Q., Whitney, M.M., Bryan, F.O. and Tseng, Y.H., 2017. A box model for representing estuarine physical processes in Earth system models. Ocean Modelling, 112, pp.139-153.
And
Sun, Q., Whitney, M.M., Bryan, F.O. and Tseng, Y.H., 2019. Assessing the skill of the improved treatment of riverine freshwater in the Community Earth System Model (CESM) relative to a new salinity climatology. Journal of Advances in Modeling Earth Systems, 11(5), pp.1189-1206.
Line 434 – add comma before “and”
Line 449 – add comma before “and”
Line 450 – remove comma before “and”
Line 451 – add comma before “or”
Line 515 – add comma before “but”
Line 515-517 – I do not understand the statement “One could remove an equivalent mass of deep water at the continental boundary to match the mass of freshwater coming from the floating source to allow the freshwater fluxes to be accurate, while also matching the sea level rise.” Is the assumption here that models have rigid lids and virtual salt fluxes? This suggestion would not seem to appropriate with models that use a free surface and real freshwater fluxes.
Line 537 - add “spread” before “uniformly”
Line 557 – The statement “Increasing meltwater can act as a negative feedback on ocean temperatures, potentially reducing the relevant climate sensitivities” has two parts that have opposing influences on ocean stratification and surface warming response – the extraction of ocean heat and buoyancy for warming and melting of ice, and the addition of buoyancy from freshwater. I don’t think it helps to combine them as a single statement unless one effect strongly outweighs the other on density depending on the fraction added as liquid.
Lines 576-581 – These sentences should be restructured as explicit guidance rather than as a set of hypotheticals, i.e. “For models that do not include explicit ice sheets, we propose that freshwater forcing be included as part of the anthropogenic suite of forcings”
Line 585 – Indeed, the GFDL models both had interactive dust in CMIP6.
Line 589-608 – I am surprised that no recommendation is made here to at least maintain the same freshwater fluxes from the end of the historical run through the future projections. This would seem preferable to using no freshwater fluxes. It would seem ill-advised for models to try to participate with these historical freshwater fluxes without having a plan for what to do with those fluxes through the transition to projections.
Line 599 – add comma before “and”
Line 613 – I would rephrase “judge the credibility of future simulations” as “judge the credibility of historical and future simulations”
Line 619 – remove “the” before “very”
Line 631 – Need to add something like “we therefore highly encourage ice sheet models to same long term averages”
Line 635 – Need to add the implication of this long term imbalance for the provision of forcing, required length of ESM simulation, or otherwise… “We therefore highly encourage…”
Line 639 – Helpful here would be a recommendation for observational reconstruction references that should be considered as helpful to provide these constraints.
Line 644 – is there a recommendation to be made here?
Line 654-657 – These two sentences are wandering and hand-waving. How about, “While the community makes long term progress on explicit coupled ESM-ISMs, there remains urgent need to make near term progress with interim configurations treating freshwater anomalies as external forcings.”
Citation: https://doi.org/10.5194/egusphere-2025-1940-RC1 - AC1: 'Reply on RC1', Gavin A. Schmidt, 02 Aug 2025
-
RC2: 'Comment on egusphere-2025-1940', Cecilia Bitz, 13 Jun 2025
This paper provides a valuable service by outlining the range of ways that ESMs model the effects of ice sheets and glaciers and by constructing datasets and recommendations to deal with their mass and energy imbalance. There are many positive aspects, like the community effort and presumably consensus behind the work. The list of definitions at the start is very good too - though I suggest adding Surface Mass Balance (SMB) to it. I am grateful to know where these authors "judge" something to be true or offer views. Some of this information would be impossible to know for certain since not all modeling centers answered their survey. The authors provide valuable guidance about simulation assumptions and practices with a summary of choices made by modeling centers with some indication of the consequences. This is helpful for model developers and those analyzing models alike.
Please considered the following specific comments:
1) There are a few occasions when the manuscript needlessly goes into minor issues about conservation. For example, I think the point made about moving snow down hill at line 166 is ridiculous, or needs further explanation. We don't worry about the conversion of PE to heat by raindrops when they hit the ground so why should we worry about the converstion of PE to heat when snow moves downhill. Further if it amounts to 14C it is not a wrinkle that might make a small difference, so something is awry about this paragraph. Another example is at Line 461-3 with regard to icebergs possibly melting a bit due to geothermal heating, etc. While all this may be true, there is little point in itemizing a bunch of stuff that no ESM developer would bother with. It makes the paper longer than it need be.
2) It is good to see some discussion of how transient additions of freshwater will alter the ocean salinity and sea level. Yet, I thought this could be a teachable moment to explain the pros and cons of adding freshwater in volume vs mass conserving ocean models (rather than an oblique reference to there being a difference at line 162). I was intrigued by the discussion that volume conserving models would overestimate sea level changes since most recent ice loss has been from floating ice and models don't do hydrostatic balance of ice correctly. It seems like this should be elevated to a recommendation at the end for modelers to work on.
3) In Fig 1 the Antarctic side shouldn't have surface runoff and much ablation zone at the surface. It is strange to see so much of Greenland below sea level and so little of Antarctica. There are elements that seems reversed between Greenland and Antarctica. Is this really meant to be the final version of this figure?
4) The figure captions could be improved with citations of sources and a bit more explanation of what the reader should glean from them.
An example is Figs 4 & 5 where weighting functions are not defined either in the captions of main text. I would have thought a weighting function would be unitless. The caption should say where such maps are available. A few pages further into reading the manuscript I see the term "Iceber met maps" in the main text but without reference to Figs 4 & 5. Given the units are the same I'm guessing this is what is plotted in Figs 4&5. If so, be sure to reference the figures, use the correct term in the caption, and state in the caption that this is one of your datasets.
Another is Fig 3 which describes data with no citation. Is this one of the products that the authors are providing? If so, I'm concerned about the sudden switch in methods and variability at year 1986 since I expect some users will just prescribe this as forcing to their model without being aware of it and then "discover" an amazing regime change in their model in 1986, similar to the many discoveries in AMIP runs forced with sea ice concentration from HADISST. I recommend providing two seperate products: One of the longer period with the same method throughout and a shorter one since 1986. (Again a few pages further into reading the manuscript I see Fig 3 discussed in the main text, which is good. Please at least say in the caption that this is one of your datasets though.)
5) Figure 6. I don't get why the circled 2 is called an "implicit FW" flux and the term "implicit SMB" is used at line 504. What is meant by "implicit" here and elsewhere (the term is used a lot)? It is supposed to mean implied, but I think maybe it is being used to mean derived or diagnosed from other quantities. It seems to me that no prognostic calculations in a model are implicit, so I can't grasp what implicit is supposed to mean. Also in Figure 6, I would think circled 2 should be labeled SMB rather than FW flux since FW flux is not as specific. The lower panel's cartoon of ice sheet to ocean fluxes is unclear and seems inconsistent with the caption. A cartoon should be an aid not a head scratcher.
6) Lines 111-114 are too hard to follow. Break this up into more sentences and or include a table/equations.
7) Line 138-9 Are you saying that some models just disappear the approx 3300 Gt/yr of accumulation on Antarctica? I thought models at least dropped the mass/volume of water into the ocean, as discussed a few paragraphs later. I've never heard of a credible ESM that didn't conserve freshwater at this level at least.
8) The manuscript jumps between modeling approaches and dataset details a bit in somewhat confusing ways. For example, line 379 is probably about a dataset provided, or possibly it is about a particular model. Either way the text should clarify.
Citation: https://doi.org/10.5194/egusphere-2025-1940-RC2 - AC2: 'Reply on RC2', Gavin A. Schmidt, 02 Aug 2025
-
CEC1: 'Comment on egusphere-2025-1940', Juan Antonio Añel, 20 Jun 2025
Dear authors,
Unfortunately, after checking your manuscript, it has come to our attention that it does not comply with our "Code and Data Policy".
https://www.geoscientific-model-development.net/policies/code_and_data_policy.html
You have archived your code on GitHub. However, GitHub is not a suitable repository for scientific publication. GitHub itself instructs authors to use other long-term archival and publishing alternatives, such as Zenodo. Therefore, the current situation with your manuscript is irregular. Please, publish your code in one of the appropriate repositories and reply to this comment with the relevant information (link and a permanent identifier for it (e.g. DOI)) as soon as possible, as we can not accept manuscripts in Discussions that do not comply with our policy.I must note that if you do not fix this problem, we will have to reject your manuscript for publication in our journal.
Also, you must include a modified 'Code and Data Availability' section in a potentially reviewed manuscript, containing the DOI of the new repositories. Also, in the GitHub site no license is listed for the code. If you do not include a license, the code remains your property and it is not possible to run or distribute it. Therefore, when uploading the model's code to Zenodo, you could want to choose a free software/open-source (FLOSS) license.
Juan A. Añel
Geosci. Model Dev. Executive EditorCitation: https://doi.org/10.5194/egusphere-2025-1940-CEC1 -
CC1: 'Reply on CEC1', Ken Mankoff, 20 Jun 2025
Dear Juan,
I have added a CC0 license to the GitHub repository. I have drafted a release (https://github.com/NASA-GISS/freshwater-forcing-workshop/releases/tag/v0.0.1-GMD-review ) and linked it with Zenodo which will archive and issue a DOI, but it has not yet been pulled by Zenodo or appears there.
Citation: https://doi.org/10.5194/egusphere-2025-1940-CC1 -
CC2: 'Reply on CC1', Ken Mankoff, 20 Jun 2025
DOI for code is now at: https://doi.org/10.5281/zenodo.15707384
Citation: https://doi.org/10.5194/egusphere-2025-1940-CC2 -
CEC2: 'Reply on CC2', Juan Antonio Añel, 21 Jun 2025
Dear authors,
Many thanks for addressging this issue so quickly.
Juan A. Añel
Geosci. Mode Dev. Executive Editor
Citation: https://doi.org/10.5194/egusphere-2025-1940-CEC2
-
CEC2: 'Reply on CC2', Juan Antonio Añel, 21 Jun 2025
-
CC2: 'Reply on CC1', Ken Mankoff, 20 Jun 2025
-
CC1: 'Reply on CEC1', Ken Mankoff, 20 Jun 2025
-
RC3: 'Comment on egusphere-2025-1940', Anonymous Referee #3, 28 Jun 2025
Reviewer comments
“Datasets and protocols for including anomalous freshwater from melting ice sheets in climate simulations” By Schmidt et al. provides valuable transient forcing of ice sheets freshwater with protocol recommendations for modellers implementing the dataset consistently. This work certainly makes an important contribution to both the Earth system modelling and ice sheet modelling communities
Please consider my suggested revisions below:
- Missing figures showing the timeseries of each freshwater product listed in Section 3.3. These products are the key output of the paper. However, the features of some products are not presented and discussed. Figure 3-5 do this for Greenland ice discharge and iceberg maps, although Figure 5 is not cited in the main text. It would be good to include figures and discussions for the remaining products.
- Missing evaluations and discussions of how well this transient freshwater forcing and suggested modelling approaches improves the representation of ocean circulation, salinity, sea level, etc., in model simulations. The modelling approaches in the two runs are described in Section 4 (Pre-industrial control runs and historical simulations); however, evaluations of the methods and products are not provided. I appreciate the authors’ time and effort, it would be good to include at least some preliminary modelled results using freshwater forcing from this work, along with comparisons to other studies.
See specific comments below:
L5 – The abstract does not mention projections. Recommend clarifying that the proposed protocol relates solely to historical simulations but that freshwater forcing is likely to become very much more important in future projection scenarios.
L9 – Can the authors confirm that freshwater forcing will be included in the forcings for the CMIP7 historical experiments? CMIP7 is now very close to releasing this forcing, and it is unclear whether freshwater is actually included.
L21 – Not convinced that computational expenses have hamper the incorporation of ice sheets into ESM. They are by comparison to other elements of the climate system cheap.
Providing historical freshwater forcing is a step forward but it does introduce an issue around projections. What are ESMs supposed to do at the start of the projection period?
L42 – There is no Appendix B.
L60 – Not convinced that these definitions are helpful. Are there not standardized definitions (CMIP or ISMIP6/7) that could be used? Runs the risk of making what is a fairly confused set of terms worse. Definitions could be improved by adding relevant units.
L65 – This is confusing and does not match the definition of discharge immediately above (in which discharge sits between ice sheets and ice shelves, while here discharge is after ice shelves before calving).
L67 – Upstream of grounding line? What about ice floating on subglacial lakes? This is upstream but not grounded.
L81 – Runoff is not purely a mass loss process if it includes rainfall (and melt of seasonal snow).
This section may be helped by linking the definition to Figure 1 more tightly. At present, terms in Figure 1 do not match the definitions (eg basal melt) and vice vera (eg discharge).
L97 onwards – Defining symbols for these various quantities and how they relate to one another would be helpful here. There are a number of things going on in the section (baseline/anomaly, time period of baseline, separation of freshwater into various fluxes such as grounded, shelf melt and shelf calving) that need to be explained more clearly. There is also some danger in referring to ‘ice mass anomalies’ because terms such as this are often used in ice mass budget calculations (freshwater fluxes are related primarily to the loss terms where mass budget relates to the net of gain and loss).
L100 – Needs to be made clear that GRACE does not directly measure freshwater flux.
L104 – Figure 3 is cited before Figure 2.
L105 – This seems to be very detailed and relate to the way in which the forcing has been calculated. Would be better in Section 3 or 4 rather than Introduction.
L118 – Good to state that basal melt of grounded ice not included in the assessment however requires some justification to demonstrate, for instance, that this flux is much smaller than the other freshwater fluxes leaving the ice mass.
L124 – quasi.
L153 – This is too vague – this assertion needs to be supported by references.
L159 – Use of the term ‘Discharge’ here and in next paragraph confusing. Does this refer to the previous definition of ‘discharge’ or discharge of freshwater? If the latter, then it is not clear that the ESM calculations discussed here would explicitly calculate discharge (line 171).
L200 – Define what is meant by ‘regional flow’ here. Will this be done on the basis of sectors/basins as per IMBIE etc?
L203 – based on Rignot et al. (2013) and Mouginot and Rignot (2019).
L210 – Earlier grounded basal melt was explicitly disregarded.
L213 – Clarify differences between refreezing and retention. Retention on what timescales?
L223 – Why choose flux gates 5 km upstream from the grounding line?
L242 – Clarification required here – grounding line retreat in itself is not a source of freshwater (increased freshwater flux and GL retreat are both consequences of increased melt for instance). Repeated confusion around inclusion of grounded ice melt.
L245 – ‘Surface runoff’ Ideally all terms of this type should be defined in Section 1.1.
L254 – Figures 3 and 4 are not cited in this section. The section does not include any discussion about the melt of ice bergs (i.e., Figure 4).
The discussion on p5 about baselines and anomalies is not mentioned at all. It is unclear at this stage what form the freshwater forcing will be take.
L262 – This is confusing Antarctic SMB has little to do with freshwater release from Antarctica and it is not clear what information on freshwater can be gleaned from trends in SMB.
L266+271 – There are a few steps between the observations mentioned in the preceding paragraph and ‘good estimates of Antarctic freshwater fluxes’ that maybe worth explaining.
L273 – Similar issue - sea level rise (ie net mass budget) is not directly linked to freshwater discharge (which is primarily to do with mass loss only).
L289 – Is the ratio by each basin or the entire Antarctica?
L297 – It would be help to have information on units here, as well as temporal resolution etc. Also it is unclear what the relationship is between the various products. For instance, were discharges and runoff used to calculate the freshwater anomalies for Greenland? Similarly, for Antarctica was calving and submarine melt used to find freshwater anomaly? How? A conceptual figure may help here (also note use of products in protocol mentioned below).
How is this related to the iceberg melt maps?
These products are the key output of the paper. It would be useful to have some plots showing key features. Figures 4 and 5 do this for iceberg melt although they are not cited in this section.
It would also be helpful to add some explanation of how each of the products is used in the protocols described in the rest of the paper and whether they are used as anomalies or absolute values. The two runs described in Section 4 are pre-industrial and historical: can the authors indicate which product is used when?
L351 – Tracer is not previously discussed.
L353 – ‘Relaxation time constant’ seems to imply a specific way of incorporating freshwater into the pre-industrial but it is unclear what this method would actually be. More detail required.
L396 onwards – This appears to be a discussion around the fine-scale physics governing local redistribution of freshwater in fjord and around glacier fronts. Not clear how it relates to the implementation of the products in ESMs. Perhaps this text should be moved to an early section?
L421 – It would be good to have more discussion on the differences/consequences of the injection depth (surface vs depth) of freshwater from icebergs and ice shelves melting in the Southern Ocean.
L429 – Not clear why the depth range of 130-230 m is suggested? ‘130 m is the mean depth of current ice shelf fronts in Antarctica.’, but where does the number ‘230 m’ come from?
L434 – Please clarify it is the configuration for the Greenland iceberg model? I don't think it is what was used in Mathiot and Jourdain (2023).
L436 – The translation of iceberg fluxes to spatially variable melt rates is a key component of this paper and is lost here in the middle of a very detailed section. This important material would be better in its own subsection positioned to accompany Figures 4 and 5 perhaps between 3.2 and 3.3. Note that this text does not cite Figure 5. Currently, this text is located in the pre-industrial subsection although it would appear to be more generic than this and would be relevant to all time periods?
L456 – Again, this subsection appears to contain a generic discussion around freshwater and energy. It is not clear why it appears under the pre-industrial.
L481 – Although this subsection is about the historical, the approach described (and shown in Figure 6) would appear to be applicable to projections as well (assuming suitable freshwater forcing was available).
L487 – May be mixing papers on historical and projection scenarios here.
L493 – It is more consistent to use 'Antarctica and Greenland' rather than 'SH and NH', as mountain glaciers are also included in SH and NH. Moreover, SH or NH acronym is not defined.
L508 – It is not clear how this discussion relates to the specific products that have been created. Which of the products are used and which are not? For instance, do the iceberg melt maps play a role?
L589 – Worth pointing out that using this protocol for the historical implies that some extension of the freshwater forcing is required for projection experiments. Otherwise, there would be an abrupt change in forcing at his start of the projection period. An alternative might be to maintain the historical freshwater forcing into the projection period which would avoid a hiatus but become increasing unrealistic as (presumably) freshwater fluxes from the ice sheets increase into the future.
L590 – Would it be possible to use this protocol with the variables supplied by ISMIP6? What additional work would be required? Presumably, as a minimum the iceberg melt patterns would need to be calculated for time in the future.
L667 – This is a bit misleading. The products for Antarctica do not cover the period before 1991.
Figures
Figure 1. Indication of scale required for all panels.
Depiction of runoff across and through the ice sheet on the Antarctica side of the diagram is misleading and should be on the Greenland side. If the diagram is meant to represent the contemporary ice sheets then presence of a substantial ablation zone in Antarctica is also extremely misleading.
Figure 4 – Please clarify how the iceberg melt rates are weighted in the caption and the main context. The same as Figure 5.
Figure 5 – What are the white lines in front of the Ross Ice Shelf in the subpanels [11] and [12]? Are they NaNs, or are they due to the grid mesh? If the latter, why are there no white lines in, for example, the panel [9], [10] and All?
Citation: https://doi.org/10.5194/egusphere-2025-1940-RC3 - AC3: 'Reply on RC3', Gavin A. Schmidt, 02 Aug 2025
Peer review completion
Journal article(s) based on this preprint
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 2,604 | 215 | 35 | 2,854 | 34 | 51 |
- HTML: 2,604
- PDF: 215
- XML: 35
- Total: 2,854
- BibTeX: 34
- EndNote: 51
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
Gavin A. Schmidt
Kenneth D. Mankoff
Jonathan L. Bamber
Dustin Carroll
David M. Chandler
Violaine Coulon
Benjamin J. Davison
Matthew H. England
Paul R. Holland
Nicolas C. Jourdain
Qian Li
Juliana M. Marson
Pierre Mathiot
Clive R. McMahon
Twila A. Moon
Ruth Mottram
Sophie Nowicki
Anne Olivé Abelló
Andrew G. Pauling
Thomas Rackow
Damien Ringeisen
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(2459 KB) - Metadata XML
The manuscript “Datasets and protocols for including anomalous freshwater from melting ice sheets in climate simulations” by Schmidt et al provides an interdisciplinary assessment of the state of understanding, uncertainties, and many technical issues involved in representing the transient liquid and solid freshwater forcing from the Greenland and Antarctic ice sheets in coupled climate Earth system models as efforts to couple these models with ice sheet models continue to advance. It certainly makes the case that this is a very technically complex problem and progress on the interim solution of applying freshwater anomalies is an important step in improving the representativeness of the ocean circulation and climate responses to anthropogenic climate change in general and the recent observed changes to ice sheet mass balance. My main issues with the present version are:
See technical suggestions below.
Line 2 – add comma before “and”
Line 5 – add “over the historical period” after “discharge”, replace “accounted for” with “addressed” or “incorporated” or “represented”, and remove “an updateable dataset of”
Line 14 – doesn’t “over the last century” include “and in recent decades”? Suggest removing or changing to “and has accelerated in recent decades”
Line 30 – Another example of the role of freshwater on Southern Ocean circulation
Bronselaer, B., Winton, M., Griffies, S.M., Hurlin, W.J., Rodgers, K.B., Sergienko, O.V., Stouffer, R.J. and Russell, J.L., 2018. Change in future climate due to Antarctic meltwater. Nature, 564(7734), pp.53-58.
line 63 - There does not seem to be consistency between the definition of "Discharge" here (as an ice flux) compared to the terms in Figure 1 (e.g. "Discharge" not identified but "Subglacial Discharge" seemingly identified as a liquid flux)… is “Discharge” equal to “Iceberg Flux” or to to the sum of several terms in Figure 1?
Line 72 – “Ice front” is not provided in Figure 1, but “Frontal retreat” is provided twice.
Line 74 – “Ice shelf” is not provided in Figure 1
Line 114 – “, and again this is with respect” should be “relative”
Line 124 – “uasi” should be “quasi”
Line 129 – I don’t think the sentence, “The implications of adding a new forcing dataset needs to be considered for each of these different configurations and experiments.” Is helpful without further contextualization and should be removed unless these implications are to be detailed.
Line 160 – The GFDL CMIP6 models included explicit icebergs: Adcroft, A., Anderson, W., Balaji, V., Blanton, C., Bushuk, M., Dufour, C.O., Dunne, J.P., Griffies, S.M., Hallberg, R., Harrison, M.J. and Held, I.M., 2019. The GFDL global ocean and sea ice model OM4. 0: Model description and simulation features. Journal of Advances in Modeling Earth Systems, 11(10), pp.3167-3211.
Line 181 – add comma before “it”
Line 185 – “observational changes” should be “observations” to avoid repeating “changes”
Line 295 – Before moving on, it would be helpful to know how these freshwater fluxes compare to those in Swart et al., 2023 for SOFIA which is cited earlier.
Line 310-312 – Is “with respect to the 1850–1900 pre-industrial period” mean that 1850-1900 is the “baseline period”? I think so from line 102, but it is not clear why different wording is being used for these two things if they are indeed the same thing.
326 – remove “the choices available in”
Line 402 – “1.5 orders of magnitude” should be “50 times”
Line 404 – remove “, such as that provided by”
Line 412 – a good reference for “estuarine box model” is
Sun, Q., Whitney, M.M., Bryan, F.O. and Tseng, Y.H., 2017. A box model for representing estuarine physical processes in Earth system models. Ocean Modelling, 112, pp.139-153.
And
Sun, Q., Whitney, M.M., Bryan, F.O. and Tseng, Y.H., 2019. Assessing the skill of the improved treatment of riverine freshwater in the Community Earth System Model (CESM) relative to a new salinity climatology. Journal of Advances in Modeling Earth Systems, 11(5), pp.1189-1206.
Line 434 – add comma before “and”
Line 449 – add comma before “and”
Line 450 – remove comma before “and”
Line 451 – add comma before “or”
Line 515 – add comma before “but”
Line 515-517 – I do not understand the statement “One could remove an equivalent mass of deep water at the continental boundary to match the mass of freshwater coming from the floating source to allow the freshwater fluxes to be accurate, while also matching the sea level rise.” Is the assumption here that models have rigid lids and virtual salt fluxes? This suggestion would not seem to appropriate with models that use a free surface and real freshwater fluxes.
Line 537 - add “spread” before “uniformly”
Line 557 – The statement “Increasing meltwater can act as a negative feedback on ocean temperatures, potentially reducing the relevant climate sensitivities” has two parts that have opposing influences on ocean stratification and surface warming response – the extraction of ocean heat and buoyancy for warming and melting of ice, and the addition of buoyancy from freshwater. I don’t think it helps to combine them as a single statement unless one effect strongly outweighs the other on density depending on the fraction added as liquid.
Lines 576-581 – These sentences should be restructured as explicit guidance rather than as a set of hypotheticals, i.e. “For models that do not include explicit ice sheets, we propose that freshwater forcing be included as part of the anthropogenic suite of forcings”
Line 585 – Indeed, the GFDL models both had interactive dust in CMIP6.
Line 589-608 – I am surprised that no recommendation is made here to at least maintain the same freshwater fluxes from the end of the historical run through the future projections. This would seem preferable to using no freshwater fluxes. It would seem ill-advised for models to try to participate with these historical freshwater fluxes without having a plan for what to do with those fluxes through the transition to projections.
Line 599 – add comma before “and”
Line 613 – I would rephrase “judge the credibility of future simulations” as “judge the credibility of historical and future simulations”
Line 619 – remove “the” before “very”
Line 631 – Need to add something like “we therefore highly encourage ice sheet models to same long term averages”
Line 635 – Need to add the implication of this long term imbalance for the provision of forcing, required length of ESM simulation, or otherwise… “We therefore highly encourage…”
Line 639 – Helpful here would be a recommendation for observational reconstruction references that should be considered as helpful to provide these constraints.
Line 644 – is there a recommendation to be made here?
Line 654-657 – These two sentences are wandering and hand-waving. How about, “While the community makes long term progress on explicit coupled ESM-ISMs, there remains urgent need to make near term progress with interim configurations treating freshwater anomalies as external forcings.”