the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 22 Jan 2024
Coupling a large-scale glacier and hydrological model (OGGM v1.5.3 and CWatM V1.08) – Towards an improved representation of mountain water resources in global assessments
Abstract. Glaciers are present in many large river basins, and due to climate change, they are undergoing considerable changes in terms of area, volume, runoff and seasonality. Although the spatial extent of glaciers is very limited in most large river basins, their role in hydrology can be substantial because glaciers store large amounts of water at varying time scales. Large-scale hydrological models are an important tool to assess climate change impacts on water resources in large river basins worldwide. Nevertheless, glaciers remain poorly represented in large-scale hydrological models. Here we present a coupling between the large-scale glacier model OGGM (v1.5.3) and the large-scale hydrological model CWatM (V1.08). We evaluated the improved glacier representation in the coupled model against the baseline hydrological model for selected river basins at 5 arcmin resolution and globally at 30 arcmin resolution, focusing on discharge projections under climate change scenarios. We find that increases in future discharge are attenuated, whereas decreases are exacerbated when glaciers are represented explicitly in the large-scale hydrological model simulations. This is explained by a projected decrease in glacier runoff in almost all basins. Calibration can compensate for lacking glacier representation in large-scale hydrological models in the past. Nevertheless, only an improved glacier representation can prevent underestimating future discharge changes, even far downstream at the outlets of large glacierized river basins. Therefore, incorporating a glacier representation into large-scale hydrological models is important for climate change impact studies, particularly when focusing on summer months or extreme years.
-
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
(29205 KB)
-
Supplement
(21701 KB)
-
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(29205 KB) - Metadata XML
-
Supplement
(21701 KB) - BibTeX
- EndNote
- Final revised paper
Journal article(s) based on this preprint
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2023-2562', Bettina Schaefli, 03 Mar 2024
This paper on coupling of a large scale glacier model and hydrological model is very well written and presented; it represents an interesting contribution for large scale modelling as well as for other model coupling studies. It discusses in good detail the challenges related to such a loose one-way coupling (adapting the output of one model to be usable as input for the other model and dividing the model domain such that each model represents one domain). The actual results in terms of prediction future water resources for the selected basins are perhaps a bit less interesting because they are probably not unexpected and because there are still too many uncertainties regarding the actual water input and because of the crucial lack of details regarding the role of groundwater. These limitations could perhaps be a bit more prominent in the abstract. But this does not reduce the value of the paper.
Detailed comments:
- The paper uses the terms runoff, surface runoff, discharge and streamflow, sometimes with unclear distinction of what is what; this mixing up of terminology is omnipresent in glacio-hydrological modelling and the actual meaning is often clear from the context but only for specialists; however, for gridded, large-scale models, it might be useful to specify if the model actually represents surface runoff (i.e. overland flow) and differentiate between simulated runoff and streamflow (I understand that the model has a transfer function to go from runoff to streamflow, see line 91, “routing);
- Human water use is mentioned at several instances; would perhaps be important to discuss somewhere (if possible) the order of magnitude of the impact of human water on streamflow (in the considered catchments) compared to glacier -sourced streamflow
- Perhaps it would be good to introduced a precise term for the water originating from glaciers, such as glacier-sourced rather than glacier melt? To avoid any mixing with “ice melt”?
- Perhaps it would be good to discuss somewhere what options the model has to “create” water if ice melt is missing (if we have negative mass balances, this represents a sources of water that the model is not able to simulate): either it drastically reduces ET or it requires a precipitation correction (more efficient)
- What do you recommend for modelling settings where there is no mass balance for model calibration, should the glacier be represented nevertheless or is the risk high that without calibration, the errors are too high to be useful for future projections? Or do your results not shed light on this?
Even more detailed comment:
- Is it mentioned somewhere if the selected catchments are past peak water or not? Was this a criterion?
- Table 1: does the CWatM not use the DEM is input? Melt representation: seasonality of what, of the degree-day factor?
- Figure 2: the Rhone catchment is missing? Can you add the names to the catchments?
- Can you say something about what the 12 calibrated parameters of CWaTM represent? What penalty function did you use for the snow cover error, what is excessive snow accumulation?
- Perhaps seasonal flow duration curves would also be useful to show how well the model does?
- Line 286: where do we see that the calibrated parameters reduce ET to compensate for missing runoff
- Line 302: can you add a comment on why future discharge is likely overestimated
- Line 441: how do you know if precipitation is underestimated? Do you have some external estimate of ET ?
- Thereafter: how is the precipitation factor applied, simple multiplicative factor on all time steps?
- Line 484: differences outside the glacier melt season: I do not fully understand : s, now on glaciers is not modelled, not during nor after the glacier melt season, what is meant here?
- It could perhaps be a bit more prominent how often the model domain is updated (glacier area), every year?
- An automatic check of commas could further improve the readability of the paper
Citation: https://doi.org/10.5194/egusphere-2023-2562-RC1 -
RC2: 'Comment on egusphere-2023-2562', Lander Van Tricht, 22 Mar 2024
The paper described the integration of a large-scale open-source glacier model, OGGM, with a (large-scale) hydrological model (CWatM). Both the glacier and hydrology communities anticipated such direct coupling, its impacts, and its effects on statements about water availability. Consequently, the study by Hanus et al. is deemed a highly valuable contribution, being certainly worth publication in GMD.
The study couples two existing models, not directly, but by linking the hydrological model with the results of the glacier model, a one-way sequential coupling. These models were not drastically modified. Therefore, the novelty of the study lies more in discussing the results and the impact of incorporating glaciers as an additional water resource rather than in the development of a new model, although there are some new developments such as daily mass balance modelling, which I think is very interesting at itself as well.
The paper is well written, with clear descriptions of methods and formulations, making it easily accessible for others to apply and further develop the proposed method. Additionally, the main ideas are well presented. In summary, the study deserves publication, with some small technical revisions and clarifications which I mention below.
Specific comments:
- Title: The last three words "in global assessments" could be removed. The focus of the paper is not on global-scale applications only. But I leave this to the authors to decide.
- Line 2: What do you mean by seasonality? Is this "runoff seasonality"?
- Line 5: Poorly or actually not even at all I think
- Line 6: "We evaluate" -> present tense?
- Line 7: "Selected basins" is a bit vague. How many selected river basins? I think this can be stated here already. Later on I see the number if four? You can state "four" or mention their names.
- Line 8: Which climate change scenarios? You could potentially add this information at this stage. For example, low and high emission climate change scenarios.
- Line 28: You could maybe give an example or two, for some of the basins that have passed peak water and some that are still heading towards the peak.
- Line 29-34: It's a bit unclear why you state these references here. Especially 29/31 focus on HMA but the study does not only.
- Line 29 – 67: I have the feeling that this part could be written a bit more concisely. Some information is repeated and could be removed, making the introduction also a bit shorter.
- Line 71: Was -> is
- Line 83: Could you mention here the number of selected basins?
- Line 112: Used -> used
- Line 116-119: Past or present tense?
- Line 142-145: Does this mean that precipitation is different at similar elevations in OGGM and CWatM?
- Table 1: Does this mean that precipitation does not increase with elevation in OGGM?
- Line 177-183: Does this mean that you do not take into account the elevation of the glacierized areas? In other words, a large glacier which has its highest elevation in grid cell i loses the same fractional area as the grid cell j with a lower elevation? This simplification could be stated more clearly.
- Line 208-209: Can be removed to make the section somewhat shorter and less repetitive
- Line 210-214: These 5 lines can be written a bit shorter I have the feeling
- Figure 2: Can you add the different basin names on this map? The distinction between Rhone and Rhine should also be clearer. Mention somewhere the glacier volume in the basins?
- Line 233-235: I am not familiar with these metrics. Can you add a bit more explanation on these metrics?
- Line 242-247: I do not understand the "five additional parameter sets". How does this differ from the global parameter set? Do you select different subsets of discharge data for calibration?
- Line 248-249: Can you estimate the effect hereof? Would your results have been different when calibrating again? You show later in the paper that calibration is important (as it can partially compensate for the role of glaciers).
- Line 252: Why 50?
- Figure 3: Is the mean shown here? There seems to be a lot of variation. Insert -> inset?
- Line 277: pronounced -> pronounced
- Line 277: remove the space after "basins".
- Line 279: Very cool to see!
- Line 283-285: I’m wondering which process than matter in summer to provide discharge if glaciers are not represented?
- Line 315: Studies -> Could you add more references here?
- Line 317: Contribute -> Contribute most?
- Line 328-329: Repetition. In the previous sentence -> 1.5% is negligible. This can be removed (for me).
- Line 325-334: This section could be rewritten somewhat less extensively. For example,
- Line 344: The performance increased -> Improved?
- Line 350: Idea to show or more mention which processes compensate for the missing glaciers?
- Line 350: Idea to show how the importance or improvement evolves as a function of glacier fraction?
- Line 379-386: What about the precipitation projections?
- Line 388-390: Can you say something about why in these basins there is no net decrease? Is glacier runoff by the end of the century still larger than at present due to the large glaciers in these basins still reacting to climate change?
- Line 401: Which processes are considered regarding human water use? Can you give some more details about this as it appears to be very important
- Line 411-413: Interesting statement. Using your results, could it be possible to estimate the maximum/minimum contribution of glaciers? This could be something interesting to mention already earlier in the paper
- Line 454-458: This is actually quite important and should be mentioned earlier in the manuscript
- Line 460-466; The differences are large. I think this should be clearer in the manuscript while also putting more forward the importance of better constraining the precipitation factor.
- Line 490: What about the glacier runoff routing? Is glacier runoff instantaneously supplied to CWatM or can there be a delay depending on glacier size? Is refreezing included in OGGM? If not, this could also be stated in this paragraph.
- Line 515: Further -> Other?
Citation: https://doi.org/10.5194/egusphere-2023-2562-RC2 - AC1: 'Comment on egusphere-2023-2562', Sarah Hanus, 12 Apr 2024
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2023-2562', Bettina Schaefli, 03 Mar 2024
This paper on coupling of a large scale glacier model and hydrological model is very well written and presented; it represents an interesting contribution for large scale modelling as well as for other model coupling studies. It discusses in good detail the challenges related to such a loose one-way coupling (adapting the output of one model to be usable as input for the other model and dividing the model domain such that each model represents one domain). The actual results in terms of prediction future water resources for the selected basins are perhaps a bit less interesting because they are probably not unexpected and because there are still too many uncertainties regarding the actual water input and because of the crucial lack of details regarding the role of groundwater. These limitations could perhaps be a bit more prominent in the abstract. But this does not reduce the value of the paper.
Detailed comments:
- The paper uses the terms runoff, surface runoff, discharge and streamflow, sometimes with unclear distinction of what is what; this mixing up of terminology is omnipresent in glacio-hydrological modelling and the actual meaning is often clear from the context but only for specialists; however, for gridded, large-scale models, it might be useful to specify if the model actually represents surface runoff (i.e. overland flow) and differentiate between simulated runoff and streamflow (I understand that the model has a transfer function to go from runoff to streamflow, see line 91, “routing);
- Human water use is mentioned at several instances; would perhaps be important to discuss somewhere (if possible) the order of magnitude of the impact of human water on streamflow (in the considered catchments) compared to glacier -sourced streamflow
- Perhaps it would be good to introduced a precise term for the water originating from glaciers, such as glacier-sourced rather than glacier melt? To avoid any mixing with “ice melt”?
- Perhaps it would be good to discuss somewhere what options the model has to “create” water if ice melt is missing (if we have negative mass balances, this represents a sources of water that the model is not able to simulate): either it drastically reduces ET or it requires a precipitation correction (more efficient)
- What do you recommend for modelling settings where there is no mass balance for model calibration, should the glacier be represented nevertheless or is the risk high that without calibration, the errors are too high to be useful for future projections? Or do your results not shed light on this?
Even more detailed comment:
- Is it mentioned somewhere if the selected catchments are past peak water or not? Was this a criterion?
- Table 1: does the CWatM not use the DEM is input? Melt representation: seasonality of what, of the degree-day factor?
- Figure 2: the Rhone catchment is missing? Can you add the names to the catchments?
- Can you say something about what the 12 calibrated parameters of CWaTM represent? What penalty function did you use for the snow cover error, what is excessive snow accumulation?
- Perhaps seasonal flow duration curves would also be useful to show how well the model does?
- Line 286: where do we see that the calibrated parameters reduce ET to compensate for missing runoff
- Line 302: can you add a comment on why future discharge is likely overestimated
- Line 441: how do you know if precipitation is underestimated? Do you have some external estimate of ET ?
- Thereafter: how is the precipitation factor applied, simple multiplicative factor on all time steps?
- Line 484: differences outside the glacier melt season: I do not fully understand : s, now on glaciers is not modelled, not during nor after the glacier melt season, what is meant here?
- It could perhaps be a bit more prominent how often the model domain is updated (glacier area), every year?
- An automatic check of commas could further improve the readability of the paper
Citation: https://doi.org/10.5194/egusphere-2023-2562-RC1 -
RC2: 'Comment on egusphere-2023-2562', Lander Van Tricht, 22 Mar 2024
The paper described the integration of a large-scale open-source glacier model, OGGM, with a (large-scale) hydrological model (CWatM). Both the glacier and hydrology communities anticipated such direct coupling, its impacts, and its effects on statements about water availability. Consequently, the study by Hanus et al. is deemed a highly valuable contribution, being certainly worth publication in GMD.
The study couples two existing models, not directly, but by linking the hydrological model with the results of the glacier model, a one-way sequential coupling. These models were not drastically modified. Therefore, the novelty of the study lies more in discussing the results and the impact of incorporating glaciers as an additional water resource rather than in the development of a new model, although there are some new developments such as daily mass balance modelling, which I think is very interesting at itself as well.
The paper is well written, with clear descriptions of methods and formulations, making it easily accessible for others to apply and further develop the proposed method. Additionally, the main ideas are well presented. In summary, the study deserves publication, with some small technical revisions and clarifications which I mention below.
Specific comments:
- Title: The last three words "in global assessments" could be removed. The focus of the paper is not on global-scale applications only. But I leave this to the authors to decide.
- Line 2: What do you mean by seasonality? Is this "runoff seasonality"?
- Line 5: Poorly or actually not even at all I think
- Line 6: "We evaluate" -> present tense?
- Line 7: "Selected basins" is a bit vague. How many selected river basins? I think this can be stated here already. Later on I see the number if four? You can state "four" or mention their names.
- Line 8: Which climate change scenarios? You could potentially add this information at this stage. For example, low and high emission climate change scenarios.
- Line 28: You could maybe give an example or two, for some of the basins that have passed peak water and some that are still heading towards the peak.
- Line 29-34: It's a bit unclear why you state these references here. Especially 29/31 focus on HMA but the study does not only.
- Line 29 – 67: I have the feeling that this part could be written a bit more concisely. Some information is repeated and could be removed, making the introduction also a bit shorter.
- Line 71: Was -> is
- Line 83: Could you mention here the number of selected basins?
- Line 112: Used -> used
- Line 116-119: Past or present tense?
- Line 142-145: Does this mean that precipitation is different at similar elevations in OGGM and CWatM?
- Table 1: Does this mean that precipitation does not increase with elevation in OGGM?
- Line 177-183: Does this mean that you do not take into account the elevation of the glacierized areas? In other words, a large glacier which has its highest elevation in grid cell i loses the same fractional area as the grid cell j with a lower elevation? This simplification could be stated more clearly.
- Line 208-209: Can be removed to make the section somewhat shorter and less repetitive
- Line 210-214: These 5 lines can be written a bit shorter I have the feeling
- Figure 2: Can you add the different basin names on this map? The distinction between Rhone and Rhine should also be clearer. Mention somewhere the glacier volume in the basins?
- Line 233-235: I am not familiar with these metrics. Can you add a bit more explanation on these metrics?
- Line 242-247: I do not understand the "five additional parameter sets". How does this differ from the global parameter set? Do you select different subsets of discharge data for calibration?
- Line 248-249: Can you estimate the effect hereof? Would your results have been different when calibrating again? You show later in the paper that calibration is important (as it can partially compensate for the role of glaciers).
- Line 252: Why 50?
- Figure 3: Is the mean shown here? There seems to be a lot of variation. Insert -> inset?
- Line 277: pronounced -> pronounced
- Line 277: remove the space after "basins".
- Line 279: Very cool to see!
- Line 283-285: I’m wondering which process than matter in summer to provide discharge if glaciers are not represented?
- Line 315: Studies -> Could you add more references here?
- Line 317: Contribute -> Contribute most?
- Line 328-329: Repetition. In the previous sentence -> 1.5% is negligible. This can be removed (for me).
- Line 325-334: This section could be rewritten somewhat less extensively. For example,
- Line 344: The performance increased -> Improved?
- Line 350: Idea to show or more mention which processes compensate for the missing glaciers?
- Line 350: Idea to show how the importance or improvement evolves as a function of glacier fraction?
- Line 379-386: What about the precipitation projections?
- Line 388-390: Can you say something about why in these basins there is no net decrease? Is glacier runoff by the end of the century still larger than at present due to the large glaciers in these basins still reacting to climate change?
- Line 401: Which processes are considered regarding human water use? Can you give some more details about this as it appears to be very important
- Line 411-413: Interesting statement. Using your results, could it be possible to estimate the maximum/minimum contribution of glaciers? This could be something interesting to mention already earlier in the paper
- Line 454-458: This is actually quite important and should be mentioned earlier in the manuscript
- Line 460-466; The differences are large. I think this should be clearer in the manuscript while also putting more forward the importance of better constraining the precipitation factor.
- Line 490: What about the glacier runoff routing? Is glacier runoff instantaneously supplied to CWatM or can there be a delay depending on glacier size? Is refreezing included in OGGM? If not, this could also be stated in this paragraph.
- Line 515: Further -> Other?
Citation: https://doi.org/10.5194/egusphere-2023-2562-RC2 - AC1: 'Comment on egusphere-2023-2562', Sarah Hanus, 12 Apr 2024
Peer review completion
Journal article(s) based on this preprint
Data sets
Coupling a large-scale glacier and hydrological model (OGGM v1.5.3 and CWatM V1.08) - Data Set Sarah Hanus, Lilian Schuster, Peter Burek, Fabien Maussion, Yoshihide Wada, and Daniel Viviroli https://doi.org/10.5281/zenodo.10046823
Model code and software
sarah-hanus/pipeline_oggm_cwatm: v1.0.0 Sarah Hanus https://doi.org/10.5281/zenodo.10048089
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 586 | 287 | 26 | 899 | 63 | 23 | 19 |
- HTML: 586
- PDF: 287
- XML: 26
- Total: 899
- Supplement: 63
- BibTeX: 23
- EndNote: 19
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
Lilian Schuster
Peter Burek
Fabien Maussion
Yoshihide Wada
Daniel Viviroli
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(29205 KB) - Metadata XML
-
Supplement
(21701 KB) - BibTeX
- EndNote
- Final revised paper