Preprints
https://doi.org/10.5194/egusphere-2024-2546
© Author(s) 2024. This work is distributed under
the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
https://doi.org/10.5194/egusphere-2024-2546
© Author(s) 2024. This work is distributed under
the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
11 Oct 2024
| 11 Oct 2024
A Flexible Snow Model (FSM 2.1.0) including a forest canopy
Abstract. Multiple options for representing physical processes in forest canopies are added to a model with multiple options for representing physical processes in snow on the ground. The canopy processes represented are shortwave and longwave radiative transfer, turbulent transfers of heat and moisture, and interception, sublimation, unloading and melt of snow in the canopy. There are options for Beer's Law or two-stream approximation canopy radiative transfer, linear or non-linear canopy snow interception efficiency, and time/melt-dependent or temperature/wind-dependent canopy snow unloading. Canopy mass and energy balance equations can be solved with one or two model layers. Model behaviour on stand scales is compared with observations of above and below canopy shortwave and longwave radiation, below canopy wind speed, snow mass on the ground and subjective estimates of canopy snow load. Large-scale simulations of snow cover extent, snow mass and albedo for the Northern Hemisphere are compared with observations and land-only simulations by state-of-the-art Earth System Models. Without accounting for uncertainty in forest structure metrics and parameter values, the ranges of multi-physics ensemble simulations are not as wide as seen in intercomparisons of existing models.
How to cite. Essery, R., Mazzotti, G., Barr, S., Jonas, T., Quaife, T., and Rutter, N.: A Flexible Snow Model (FSM 2.1.0) including a forest canopy, EGUsphere [preprint], https://doi.org/10.5194/egusphere-2024-2546, 2024.
Publisher's note: Copernicus Publications remains neutral with regard to jurisdictional claims made in the text, published maps, institutional affiliations, or any other geographical representation in this preprint. The responsibility to include appropriate place names lies with the authors.
Download & links
Download & links
- Preprint
(6881 KB) - Metadata XML
- BibTeX
- EndNote
Status: closed
Comment types: AC – author | RC – referee | CC – community | EC – editor | CEC – chief editor
| : Report abuse
-
RC1: 'Comment on egusphere-2024-2546', Anonymous Referee #1, 13 Nov 2024
In the paper "A Flexible Snow Model (FSM 2.1.0) including a forest canopy", Essery et al. present the Flexible Snow Model FSM2, an extended version of the Fractional Snow Model (FSM) in which canopy snow processes are described. The manuscript includes a thorough model description of shortwave radiative transfer, longwave radiative fluxes and turbulent heat exchanges between the snow and the (single or double layered) canopy, as well as the canopy mass balance. The authors show simulation results with respect to the modeling options for radiative transfer, interception efficiency and canopy snow unloading. Modeling results are then compared with field observations in Scandinavia and Switzerland and with large scale remote sensing simulations over the Northern Hemisphere.
This is a high-level model description paper that I enjoyed reading. The model is clearly presented, the figures are clean and this paper is a relevant contribution to the snow forest modeling community. The manuscript is already very good and I have no major concerns. However, I do suggest a few changes that I think could improve the overall quality of this work. I recommend that the authors consider these before the publication of this paper in its final form.
General comments:
- Introduction. In my opinion, the introduction is very solid. It is well written and guides the reader to the section describing the model. As mentioned by the authors, FSM has been widely used in previous work by many people in the snow modeling community from different countries and has contributed to the development of scientists in academia (l. 19-20). However, I think the introduction would benefit from an additional paragraph mentioning and describing the context in which researchers have used FSM in previous studies. This would highlight some shortcomings in the description of canopy snow processes in the model and underline the importance of developing a robust and flexible modelling scheme for it.
- Discussion. As there are several options that have to be decided by the user, I think the authors should provide better guidance to the community on how to use their model. Based on the simulations performed at three different sites (two in Scandinavia, and one in Switzerland), based on the literature supporting the different parameterisations described in the paper, and based on their understanding of the physical processes represented in the model, I suggest that the authors provide insights regarding the options to use for specific climates or modeling purposes.
- Outlook. It would be interesting to have a few words about the ongoing and future developments of FSM2. I think the current limitations of the model should also be exposed.
Specific and technical comments.
- l. 1: I would explicitly name the model “FSM” instead of referring to a generic “model” in the first sentence of the manuscript.
- l. 11: I feel that a sentence could be added at the end of the abstract mentioning how this new model development can help the scientific community to improve their ability to model snow in forested environments.
- l. 66: Given that several models use the leaf area index (LAI) as a parameter for forest structure, and that the LAI is commonly (and rather easily) measured in the field, it is worth adding a few words on the difference between the effective vegetation index, as used in FSM, and the LAI.
- l. 94: Please add one or two sentences to briefly summarize the method from Erbs et al. (1982).
- Some default parameters are mentioned in the text (l. 129, l. 256, l. 262, l. 277, l. 349) without any explanation of the choice of the specific value for these parameters. Were these values determined from a sensitivity analysis? Are they based on previous modeling or experimental work or arbitrarily chosen? Please indicate how each of these parameters was defined.
- l. 171: What is the meaning of "canopy gaps"? Looking at equation 14, I understand that the authors refer to the space in a canopy layer without leaves or branches that allows the transmission of diffuse shortwave and longwave radiation. As "canopy gaps" generally refer to a sub-environment where energy and mass fluxes are altered compared to a "full canopy" environment in the snow-forest scientific literature, this could be misleading. Please consider replacing it with another term.
- l. 301. Please specify the parameter from which the iteration starts.
- l. 353. Given the questionable physical basis of this 2/3 exponent in equation 80, I am curious about the sensitivity of the model to this parameter. Have you tried running simulations with exponents other than 2/3?
- l. 365. This suggests that snow in the canopy evolves in the same way as snow on the ground. Please state this clearly in the manuscript.
- l. 397 to 405: Please consider moving this to a subsection of the method in which the study sites would be described.
- l. 441 to 443: Please specify which parameters have been adjusted and which ones have been let as default
- l. 446-447. I imagine that a subjective discrimination between 9 levels of canopy interception must be very difficult to do. Showing photos of these 9 interception levels would help the reader to see the nuance between each stage of canopy load. I suggest adding this in a supplementary material document.
- Table 8: Some plots are somehow a bit messy with all the black lines (especially 8b). Consider showing the ensemble median as a black line with a min-max envelope.
- Table 1: This is perhaps a personal preference, but as a reader I would find it easier to grasp differences in modelling results between each simulation with a figure rather than a table. This could be done by plotting two variables against each other on an x-y plot and a color map for the third variable. As there are only 16 points to plot, you could then write a short name for each simulation on the graph without oversaturating it.
- I find that Figure 9b is very interesting. I would like to see how this partitioning would change with the different simulation schemes (Table 1). I am curious about what drives the melt energy for sites with different canopy densities. Consider including this in the supplementary material.
- l. 521. Can you explain why this particular model behaves differently to the others?
- Figure 10. Please enlarge this figure.
- l. 554-555. Can you elaborate on how these modeling choices regarding canopy snow unloading and canopy snow-rainfall interactions may be more appropriate for some climates and less appropriate for others?
- l. 571-572: Please, revise this sentence. There seems to be a word missing.
Citation: https://doi.org/10.5194/egusphere-2024-2546-RC1 -
AC1: 'Response to comments on egusphere-2024-2546', Richard L.H. Essery, 15 Feb 2025
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2024/egusphere-2024-2546/egusphere-2024-2546-AC1-supplement.pdf
-
RC2: 'Comment on egusphere-2024-2546', Isabelle Gouttevin, 16 Dec 2024
This article by Essery et al. is a thorough description of the FSM2 multi-physics canopy model intended for snow-covered areas and to represent the interactions between the atmosphere, the canopy and the underlying or intercepted snow. Besides the description of the different physically-based options, the paper includes comparisons between some parameterisations, as well as site-based and global evaluations or comparisons to state-of-the art surface/snow models, considering the spread induced by the multi-physics.This paper is, in my opinion, a very significant piece of work, due to (i) the state-of-the art quality of the model concieved and the excellent quality of its description (ii) the value of the multi-physics approach for a wealth of applications like uncertainty characterizations or ensemble assimilations, and (iii) the comparison it provides of the impact, onto the sub-canopy snowpack, of different parameterizations.I furthermore congratulate the authors for their highly pedagogic style, which is very valuable in model description publications. In the contexte of this multi-physics model, the pedadogy spreads over a wealth of modelling choices made in the history of snow/canopy model developments, and the present papers helps to understand the rich landscape of these models and get to the root of some paramerizations - an effort much appreciated and imo usefull to all modellers.Still I have a few remarks that may improve the manuscript a little bit, so here they are :- Dense vs sparse canopies are sometimes mentioned, with impacts on the model parameters (e.g. L 257 ; L 285-287). It is not very clear by what is meant with these words (occurence of canopy gaps or homogeneous, sparse canopies, or both ?) and maybe, should be made more explicit. How is it done in large-scale simulations with explicit forest fractions ? (L513-514)- L 95 / section 2.2 or 2.2.4 : I missed some background, here or in the introduction, on the Beer's Law vs Two-stream approximation options. What would motivate one over the other in general, and in forest-specific litterature ?- L 129 / eq 15, the choice for the default values of the snow-free and snow-covered dense canopy albedo parameters should be justified- L 139 : wouldn't there be a typo here, and wouldn't a_Lambda rather be alpha_Lambda ? Otherwise I don't see that a_Lambda has been used earlier nor is given any analytical formulation ; in that case, what's the reason for not using omega all the way long instead ; how to estimate this parameter ?- Fig 3 and 5 : Has such a comparison between the two kinds of radiative transfer models for canopy ever been produced, and with similar outcomes ? I find these very interesting results.- L 312 : a comma after "fluxes" may clarify the sentence-L 345 : It seems to me that VAI has not been defined earlier in the text ; it should be defined line 66.- L 372 : It seems to me that Sf has not been introduced ; it could probably be done L 370.- L 374 : If I am correct, Eq 84 holds for each layer with its respective VAI to drive Sc ; I feel it would be worth mentioning it to make it easier to understand the effects on interception when going to the 2-layer version.- L 460-461 : it would be worth mentioning that SnowMIP2 simulations involved models with a very crude, parametric representation of canopies, which probably explains a larger scatter in these simulations than in the FSM2-ensemble.- L 483-484 : I remember getting on average higher subcanopy longwave radiations upon the use of a 2-layer canopy model in SNOWPACK, despite indeed lower daily values (see Table 3 in Gouttevin et al. 2015). There are several differences between the 2-layer model you developped and the 2-layer canopy model of SNOWPACK but I'm curious if you have an explanation for this ?- Figure 11 : it would be nice to mention an explanation for the snow mass overestimations by most models, if there exists one.Citation: https://doi.org/
10.5194/egusphere-2024-2546-RC2 -
AC1: 'Response to comments on egusphere-2024-2546', Richard L.H. Essery, 15 Feb 2025
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2024/egusphere-2024-2546/egusphere-2024-2546-AC1-supplement.pdf
-
AC1: 'Response to comments on egusphere-2024-2546', Richard L.H. Essery, 15 Feb 2025
-
AC1: 'Response to comments on egusphere-2024-2546', Richard L.H. Essery, 15 Feb 2025
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2024/egusphere-2024-2546/egusphere-2024-2546-AC1-supplement.pdf
Status: closed
Comment types: AC – author | RC – referee | CC – community | EC – editor | CEC – chief editor
| : Report abuse
-
RC1: 'Comment on egusphere-2024-2546', Anonymous Referee #1, 13 Nov 2024
In the paper "A Flexible Snow Model (FSM 2.1.0) including a forest canopy", Essery et al. present the Flexible Snow Model FSM2, an extended version of the Fractional Snow Model (FSM) in which canopy snow processes are described. The manuscript includes a thorough model description of shortwave radiative transfer, longwave radiative fluxes and turbulent heat exchanges between the snow and the (single or double layered) canopy, as well as the canopy mass balance. The authors show simulation results with respect to the modeling options for radiative transfer, interception efficiency and canopy snow unloading. Modeling results are then compared with field observations in Scandinavia and Switzerland and with large scale remote sensing simulations over the Northern Hemisphere.
This is a high-level model description paper that I enjoyed reading. The model is clearly presented, the figures are clean and this paper is a relevant contribution to the snow forest modeling community. The manuscript is already very good and I have no major concerns. However, I do suggest a few changes that I think could improve the overall quality of this work. I recommend that the authors consider these before the publication of this paper in its final form.
General comments:
- Introduction. In my opinion, the introduction is very solid. It is well written and guides the reader to the section describing the model. As mentioned by the authors, FSM has been widely used in previous work by many people in the snow modeling community from different countries and has contributed to the development of scientists in academia (l. 19-20). However, I think the introduction would benefit from an additional paragraph mentioning and describing the context in which researchers have used FSM in previous studies. This would highlight some shortcomings in the description of canopy snow processes in the model and underline the importance of developing a robust and flexible modelling scheme for it.
- Discussion. As there are several options that have to be decided by the user, I think the authors should provide better guidance to the community on how to use their model. Based on the simulations performed at three different sites (two in Scandinavia, and one in Switzerland), based on the literature supporting the different parameterisations described in the paper, and based on their understanding of the physical processes represented in the model, I suggest that the authors provide insights regarding the options to use for specific climates or modeling purposes.
- Outlook. It would be interesting to have a few words about the ongoing and future developments of FSM2. I think the current limitations of the model should also be exposed.
Specific and technical comments.
- l. 1: I would explicitly name the model “FSM” instead of referring to a generic “model” in the first sentence of the manuscript.
- l. 11: I feel that a sentence could be added at the end of the abstract mentioning how this new model development can help the scientific community to improve their ability to model snow in forested environments.
- l. 66: Given that several models use the leaf area index (LAI) as a parameter for forest structure, and that the LAI is commonly (and rather easily) measured in the field, it is worth adding a few words on the difference between the effective vegetation index, as used in FSM, and the LAI.
- l. 94: Please add one or two sentences to briefly summarize the method from Erbs et al. (1982).
- Some default parameters are mentioned in the text (l. 129, l. 256, l. 262, l. 277, l. 349) without any explanation of the choice of the specific value for these parameters. Were these values determined from a sensitivity analysis? Are they based on previous modeling or experimental work or arbitrarily chosen? Please indicate how each of these parameters was defined.
- l. 171: What is the meaning of "canopy gaps"? Looking at equation 14, I understand that the authors refer to the space in a canopy layer without leaves or branches that allows the transmission of diffuse shortwave and longwave radiation. As "canopy gaps" generally refer to a sub-environment where energy and mass fluxes are altered compared to a "full canopy" environment in the snow-forest scientific literature, this could be misleading. Please consider replacing it with another term.
- l. 301. Please specify the parameter from which the iteration starts.
- l. 353. Given the questionable physical basis of this 2/3 exponent in equation 80, I am curious about the sensitivity of the model to this parameter. Have you tried running simulations with exponents other than 2/3?
- l. 365. This suggests that snow in the canopy evolves in the same way as snow on the ground. Please state this clearly in the manuscript.
- l. 397 to 405: Please consider moving this to a subsection of the method in which the study sites would be described.
- l. 441 to 443: Please specify which parameters have been adjusted and which ones have been let as default
- l. 446-447. I imagine that a subjective discrimination between 9 levels of canopy interception must be very difficult to do. Showing photos of these 9 interception levels would help the reader to see the nuance between each stage of canopy load. I suggest adding this in a supplementary material document.
- Table 8: Some plots are somehow a bit messy with all the black lines (especially 8b). Consider showing the ensemble median as a black line with a min-max envelope.
- Table 1: This is perhaps a personal preference, but as a reader I would find it easier to grasp differences in modelling results between each simulation with a figure rather than a table. This could be done by plotting two variables against each other on an x-y plot and a color map for the third variable. As there are only 16 points to plot, you could then write a short name for each simulation on the graph without oversaturating it.
- I find that Figure 9b is very interesting. I would like to see how this partitioning would change with the different simulation schemes (Table 1). I am curious about what drives the melt energy for sites with different canopy densities. Consider including this in the supplementary material.
- l. 521. Can you explain why this particular model behaves differently to the others?
- Figure 10. Please enlarge this figure.
- l. 554-555. Can you elaborate on how these modeling choices regarding canopy snow unloading and canopy snow-rainfall interactions may be more appropriate for some climates and less appropriate for others?
- l. 571-572: Please, revise this sentence. There seems to be a word missing.
Citation: https://doi.org/10.5194/egusphere-2024-2546-RC1 -
AC1: 'Response to comments on egusphere-2024-2546', Richard L.H. Essery, 15 Feb 2025
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2024/egusphere-2024-2546/egusphere-2024-2546-AC1-supplement.pdf
-
RC2: 'Comment on egusphere-2024-2546', Isabelle Gouttevin, 16 Dec 2024
This article by Essery et al. is a thorough description of the FSM2 multi-physics canopy model intended for snow-covered areas and to represent the interactions between the atmosphere, the canopy and the underlying or intercepted snow. Besides the description of the different physically-based options, the paper includes comparisons between some parameterisations, as well as site-based and global evaluations or comparisons to state-of-the art surface/snow models, considering the spread induced by the multi-physics.This paper is, in my opinion, a very significant piece of work, due to (i) the state-of-the art quality of the model concieved and the excellent quality of its description (ii) the value of the multi-physics approach for a wealth of applications like uncertainty characterizations or ensemble assimilations, and (iii) the comparison it provides of the impact, onto the sub-canopy snowpack, of different parameterizations.I furthermore congratulate the authors for their highly pedagogic style, which is very valuable in model description publications. In the contexte of this multi-physics model, the pedadogy spreads over a wealth of modelling choices made in the history of snow/canopy model developments, and the present papers helps to understand the rich landscape of these models and get to the root of some paramerizations - an effort much appreciated and imo usefull to all modellers.Still I have a few remarks that may improve the manuscript a little bit, so here they are :- Dense vs sparse canopies are sometimes mentioned, with impacts on the model parameters (e.g. L 257 ; L 285-287). It is not very clear by what is meant with these words (occurence of canopy gaps or homogeneous, sparse canopies, or both ?) and maybe, should be made more explicit. How is it done in large-scale simulations with explicit forest fractions ? (L513-514)- L 95 / section 2.2 or 2.2.4 : I missed some background, here or in the introduction, on the Beer's Law vs Two-stream approximation options. What would motivate one over the other in general, and in forest-specific litterature ?- L 129 / eq 15, the choice for the default values of the snow-free and snow-covered dense canopy albedo parameters should be justified- L 139 : wouldn't there be a typo here, and wouldn't a_Lambda rather be alpha_Lambda ? Otherwise I don't see that a_Lambda has been used earlier nor is given any analytical formulation ; in that case, what's the reason for not using omega all the way long instead ; how to estimate this parameter ?- Fig 3 and 5 : Has such a comparison between the two kinds of radiative transfer models for canopy ever been produced, and with similar outcomes ? I find these very interesting results.- L 312 : a comma after "fluxes" may clarify the sentence-L 345 : It seems to me that VAI has not been defined earlier in the text ; it should be defined line 66.- L 372 : It seems to me that Sf has not been introduced ; it could probably be done L 370.- L 374 : If I am correct, Eq 84 holds for each layer with its respective VAI to drive Sc ; I feel it would be worth mentioning it to make it easier to understand the effects on interception when going to the 2-layer version.- L 460-461 : it would be worth mentioning that SnowMIP2 simulations involved models with a very crude, parametric representation of canopies, which probably explains a larger scatter in these simulations than in the FSM2-ensemble.- L 483-484 : I remember getting on average higher subcanopy longwave radiations upon the use of a 2-layer canopy model in SNOWPACK, despite indeed lower daily values (see Table 3 in Gouttevin et al. 2015). There are several differences between the 2-layer model you developped and the 2-layer canopy model of SNOWPACK but I'm curious if you have an explanation for this ?- Figure 11 : it would be nice to mention an explanation for the snow mass overestimations by most models, if there exists one.Citation: https://doi.org/
10.5194/egusphere-2024-2546-RC2 -
AC1: 'Response to comments on egusphere-2024-2546', Richard L.H. Essery, 15 Feb 2025
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2024/egusphere-2024-2546/egusphere-2024-2546-AC1-supplement.pdf
-
AC1: 'Response to comments on egusphere-2024-2546', Richard L.H. Essery, 15 Feb 2025
-
AC1: 'Response to comments on egusphere-2024-2546', Richard L.H. Essery, 15 Feb 2025
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2024/egusphere-2024-2546/egusphere-2024-2546-AC1-supplement.pdf
Model code and software
FSM 2.1.0 Richard Essery, Giulia Mazzotti, Sarah Barr, Tobias Jonas, Tristan Quaife, and Nick Rutter https://doi.org/10.5281/zenodo.13308507
Viewed
Total article views: 674 (including HTML, PDF, and XML)
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 372 | 116 | 186 | 674 | 13 | 16 |
- HTML: 372
- PDF: 116
- XML: 186
- Total: 674
- BibTeX: 13
- EndNote: 16
Cumulative views and downloads
(calculated since 11 Oct 2024)
Viewed (geographical distribution)
Total article views: 659 (including HTML, PDF, and XML)
Thereof 659 with geography defined
and 0 with unknown origin.
| Country | # | Views | % |
|---|---|---|---|
| United States of America | 1 | 180 | 27 |
| Canada | 2 | 37 | 5 |
| Switzerland | 3 | 35 | 5 |
| France | 4 | 32 | 4 |
| United Kingdom | 5 | 32 | 4 |
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
- 180
1
180
Latest update: 24 Mar 2025
School of GeoSciences, University of Edinburgh, Edinburgh, UK
Giulia Mazzotti
WSL Institute for Snow and Avalanche Research SLF, Davos Dorf, Switzerland
Sarah Barr
WSL Institute for Snow and Avalanche Research SLF, Davos Dorf, Switzerland
Department of Earth and Environmental Sciences, University of Manchester, Manchester, UK
Tobias Jonas
WSL Institute for Snow and Avalanche Research SLF, Davos Dorf, Switzerland
Tristan Quaife
National Centre for Earth Observation, University of Reading, Reading, UK
Nick Rutter
Department of Geography and Environmental Sciences, Northumbria University, Newcastle upon Tyne, UK
Short summary
How forests influence accumulation and melt of snow on the ground is of long-standing interest, but uncertainty remains in how best to model forest snow processes. We developed the Flexible Snow Model version 2 to quantify these uncertainties. In a first model demonstration, how unloading of intercepted snow from the forest canopy is represented is responsible for the largest uncertainty. Global mapping of forest distribution is also likely to be a large source of uncertainty in existing models.
How forests influence accumulation and melt of snow on the ground is of long-standing interest,...