the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 27 Mar 2024
Seasonal Snow-Atmosphere Modeling: Let's do it
Abstract. Mountain snowpack forecasting relies on accurate mass and energy input information to the snowpack. For this reason, coupled snow-atmosphere models, which downscale input fields to the snow model using atmospheric physics, have been developed. These coupled models are often limited in the spatial and temporal extent of their use by computational constraints. In addressing this challenge, we introduce HICARsnow, an intermediate-complexity coupled snow-atmosphere model. HICARsnow couples two physics-based models of intermediate complexity to enable basin-scale snow and atmospheric modeling at seasonal time scales. To showcase the efficacy and capability of HICARsnow, we present results from its application to a high-elevation basin in the Swiss Alps. The simulated snow depth is compared throughout the snow season to aerial LiDAR data. The model shows reasonable agreement with observations from peak accumulation through late-season melt-out, representing areas of high snow accumulation due to redistribution processes, as well as melt patterns caused by interactions between radiation and topography. HICARsnow is also found to resolve preferential deposition, with model output suggesting that parameterizations of the process using surface wind fields only may be inappropriate under certain atmospheric conditions. The two-way coupled model also improves surface air temperatures over late-season snow, demonstrating added value for the atmospheric model as well. Differences between observations and model output during the accumulation season indicate a poor representation of redistribution processes away from exposed ridges and steep terrain, and a low-bias in albedo at high elevations during the ablation season. Overall, HICARsnow shows great promise for applications in operational snow forecasting and studying the representation of snow accumulation and ablation processes.
-
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
(27679 KB)
-
Supplement
(225 KB)
-
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(27679 KB) - Metadata XML
-
Supplement
(225 KB) - BibTeX
- EndNote
- Final revised paper
Journal article(s) based on this preprint
Interactive discussion
Status: closed
-
CC1: 'Comment on egusphere-2024-489', Yang Yu, 23 Apr 2024
The paper “Seasonal Snow-Atmosphere Modeling: Let’s do it” developed a model coupled a snowpack model with an atmosphere model. The model is verified through comparing the simulating results with the Li-DAR data. The results about snow preferential deposition, snow redistribution, snow melt and the near surface temperature is discussed. The overall model results show great agrements with the experiments results at hectometer scale in mountainous area. The writing and organization are good. Here are some suggestions:
- In the introduction part, the snow preferential deposition background knowledge is introduction in details, but the introduction about how radiation influence the snow melt is a little less.
- Section 2.1 about the model coupling, a framework figure would help a lot.
- Section 2.2, paragraph 2 (line 114 and line 117), how a reference in 2002 (Doorschot and Lehning, 2007) explains the shortness of the scheme in 2007 (Liston et al., 2007).
- Section 2.2, line 125, “standard halo exchange’ better quote a paper for reader to understand.
- Section 3.1.1, as the snowfall process is highly related to the wind field, a wind speed rose map of the studying area could help explain the results in figure 2.
- The font size in figure 6 is smaller than other figure, unify the font size for all figures helps.
- What is the dark solid line means in the last subfigure in figure 7?
- Line 277 – 301, these paragraphs discuss the influence of the 3D flow field to the cloud microphysics and the near-surface particle-flow interactions, paper “Huang N, Yu Y, Shao Y, et al. Numerical Simulation of Falling‐Snow Deposition Pattern Over 3D‐Hill[J]. Journal of Geophysical Research: Atmospheres, 2024, 129(2): e2023JD039898.” could support these discussions.
- At the first paragraph of section 3.2, author write “A low bias in high-elevation albedo, combined with a slight warm bias, could explain this excessive melt.” to explain the low snow melt at higher elevation in the model results. May be this is because the atmosphere model lacks the physics process called “temperature inversion”. Due to strong radiation cooling based on lower humidity and cold air run-off from a higher mountain region into the valley or basin by a local circulation, the air temperature in the valley or basin becomes lower and lower during night in wintertime, and temperature is higher at high elevation. This process occurred and a temperature inversion layer exited almost every day during wintertime. This physics process could also explain the error in the near surface air temperature modeling. Author could refer to paper “Du, M., et al. "Temperature distribution in the high mountain regions on the Tibetan Plateau-Measurement and simulation." Proc. MODSIM 2007 International Congress on Modelling and Simulation. Modelling and Simulation Society of Australia and New Zealand. 2007.”.
- If the “temperature inversion” do matters, the conclusion about the snowmelt and air temperature parts need rewrite.
Citation: https://doi.org/10.5194/egusphere-2024-489-CC1 -
RC1: 'Comment on egusphere-2024-489', Manuel Tobias Blau, 11 May 2024
The manuscript "Seasonal Snow-Atmosphere Modeling: Let's do it" by Reynolds et al. introduces a new model setup based on the coupling of the HICAR with FSM2trans to resolve snow depth and snow-atmosphere interaction more effectively. For model performance demonstration, the authors applied the model setup to a case in the Swiss Alps and validated the model output with survey data from LiDAR flights. They found improvements in the model performance for surface temperature, blowing snow sublimation, and snow depth after coupling the snow component of FSM2trans (FSM2oshd) and using the ISHMAEL microphysics scheme that discriminates three types of ice states.
The well-written study may contribute to scientific advancement in snow modeling and other applications. Therefore, it may interest a broad range of scientists in various disciplines. However, a few more general points could be addressed and discussed more deeply in the manuscript.
When referring to figures, sometimes "Fig." and sometimes "Figure" are used. A uniform referencing system would be better.
Section 2.1: This section could be supported with a schematic diagram that outlines the model setup.
Section 2.2: Which of the redistribution processes was calculated first?
L 128: How sensitive is the model to different amounts of iterations?
Section 2.4: How sensitive is the model to the amount of snow layers? Does the top layer represent the surface that interacts with the atmosphere?
L. 170: There is a (TODO: HERE)?
L. 205: Which criteria are used for partitioning precipitation into snowfall and rainfall? How is the rain-snow slope line defined?
Section 3.1.1: What contributes most to the advection of snow particles and snowfall pattern during the event (wind aloft, above the ridge, or the updraft on the downwind slopes)?
Section 3.1.2: How much do the two discussed redistribution processes (wind-driven or gravitational redistribution) individually contribute to total redistribution? Can you give an estimate?
L. 333: It would be good if there were a brief statement of model performance in terms of meteorological variables earlier in the manuscript.
L. 355: "5" is "Figure 5"?
L.359: "The following paragraph ..." could be a new paragraph.
L. 363: Is the warm bias due to the too low albedo in high elevations?
How is snow albedo parameterized, and which values are assumed for fresh and accumulated snow?
L. 375: "4" is Figure 4"?
L. 435: Are the differences in the blowing snow sublimation percentage of snowfall due to differences in snowfall or the rate of sublimated snow?
Citation: https://doi.org/10.5194/egusphere-2024-489-RC1 - AC1: 'Reply on RC1', Dylan Reynolds, 30 May 2024
-
RC2: 'Comment on egusphere-2024-489', Anonymous Referee #2, 27 May 2024
This is a review of "Seasonal Snow-Atmosphere Modeling: Let’s do it”. Overall this is generally a quite well written and readable manuscript. The science is really exciting and at the forefront of the field.
My main suggestion is that the authors very clearly articulate the science questions that motivate this study. The introduction does a good job of setting up the research gaps, but I found many of the sections came at me a bit out of the blue. For example, I would really like to see clear science questions that establish the need for Morrison v. ISHMAEL, and specific method which the authors had a priori anticipated working well in this domain.
Some of the model decisions are behind in-prep manuscripts, which makes it a bit hard to follow some of the modelling decisions. For example, why the spatial resolutions were chosen. Indeed, it is pretty obvious to those of us who do this kind of modelling, but I think it would help readability if these model decisions were better articulated, the sensitive of the results to them, and for them to be tied to the underlying research questions. If this is tightened up, it should be a very easy-to-follow manuscript.
The domain has very few (none?) trees. Realizing it is a bit out of scope, I am very curious to know what the authors think of applying these methods to more forested catchments that have complex interactions of canopy-wind in the downslope regions.
w/c = Word choiceL20 Exposed to what?
L21 is it really a discontinuity though? The derivative remains well defined. I think this would be better restated as "sharp gradient" or similar
L23 "modified" w/c. Modified feels like a model input term
L34 space between amount an dunit – 100m -> 100 m. Fix throughout
L34 "at elevations" - "at a height above the surface of" might be more clear to match the language on L41
L57 "modern" -> contemporary?
L58 "troubling" w/c
L65 "These conditions have been followed by the earlier studies" what conditions are these and how can they be followed by an earlier study?
L66 " This is due to" unclear what this is in reference to
L67 horizontal and vertical resolutions?
L67 "One exception to this " what is this?
L67+ There is a lot of "this" throughout the section, which makes it difficult to follow. Generally I can infer what "this" is in reference to, but it would be best to explicitly state it to ease readability.
L72 "Caveat aside" which caveat is this in reference to?
L74 "adopting this strategy" this == intermediate complexity atmospheric models?
L85 what are the scientific questions, specifically?
L90 Later in the MS I found the mix of FSM2trans (e.g. fig 11), HICARsnow, NoahMP, a bit confusing. L 103 notes the name depends on the coupling strategy. I think a small table or a very clear description of all the comparisons would help readability
L93 define `oshd` ?
L96 static lib, are you meaning a `.a` ? This is very specific — do you mean this to preclude using a dynamic lib?
L109 FSM2 has a soil routine. How are frozen soils + the ground heatflux coupled into this? Or is this fully disabled in this configuration
L128 Is there an opportunity to use a dynamic iteration based on an error term versus a fixed iteration count ?
L125 “image” is not clear. I have a vague recollection this is maybe a iSNOBAL term? For the raster? Can you please clarity.
L170 fix TODO: HERE
L171 can you list, even briefly the details? Difficult to tell with an in-prep manuscript (but I do understand)
L235 Maybe it is just the layout, but it seems that the text goes from citing Fig 2 to Fig 6&7. My recommendation would be number the figures in the order they appear. As is, I’m finding it difficult to find the text that references a specific figure
L 275 Figure 7: What are the purple contours in the 2nd row? I don’t see them in the legend. I like this plot though.
L331 How is the surface roughness interaction with HICAR modelled?
L359 “The following paragraph” — suggest you start a new paragraph
L361 how big is “slight”?
L414 Do you think this excessive cooling impact ablation rates? I had originally read this as a FSM2 characteristic, but rereading the text I’m uncertain if it is actually NOAHMP. Please tighten this up a bit
L443 these tradeoffs should be clearly noted in the methodologyCitation: https://doi.org/10.5194/egusphere-2024-489-RC2 - AC2: 'Reply on RC2', Dylan Reynolds, 05 Jun 2024
Interactive discussion
Status: closed
-
CC1: 'Comment on egusphere-2024-489', Yang Yu, 23 Apr 2024
The paper “Seasonal Snow-Atmosphere Modeling: Let’s do it” developed a model coupled a snowpack model with an atmosphere model. The model is verified through comparing the simulating results with the Li-DAR data. The results about snow preferential deposition, snow redistribution, snow melt and the near surface temperature is discussed. The overall model results show great agrements with the experiments results at hectometer scale in mountainous area. The writing and organization are good. Here are some suggestions:
- In the introduction part, the snow preferential deposition background knowledge is introduction in details, but the introduction about how radiation influence the snow melt is a little less.
- Section 2.1 about the model coupling, a framework figure would help a lot.
- Section 2.2, paragraph 2 (line 114 and line 117), how a reference in 2002 (Doorschot and Lehning, 2007) explains the shortness of the scheme in 2007 (Liston et al., 2007).
- Section 2.2, line 125, “standard halo exchange’ better quote a paper for reader to understand.
- Section 3.1.1, as the snowfall process is highly related to the wind field, a wind speed rose map of the studying area could help explain the results in figure 2.
- The font size in figure 6 is smaller than other figure, unify the font size for all figures helps.
- What is the dark solid line means in the last subfigure in figure 7?
- Line 277 – 301, these paragraphs discuss the influence of the 3D flow field to the cloud microphysics and the near-surface particle-flow interactions, paper “Huang N, Yu Y, Shao Y, et al. Numerical Simulation of Falling‐Snow Deposition Pattern Over 3D‐Hill[J]. Journal of Geophysical Research: Atmospheres, 2024, 129(2): e2023JD039898.” could support these discussions.
- At the first paragraph of section 3.2, author write “A low bias in high-elevation albedo, combined with a slight warm bias, could explain this excessive melt.” to explain the low snow melt at higher elevation in the model results. May be this is because the atmosphere model lacks the physics process called “temperature inversion”. Due to strong radiation cooling based on lower humidity and cold air run-off from a higher mountain region into the valley or basin by a local circulation, the air temperature in the valley or basin becomes lower and lower during night in wintertime, and temperature is higher at high elevation. This process occurred and a temperature inversion layer exited almost every day during wintertime. This physics process could also explain the error in the near surface air temperature modeling. Author could refer to paper “Du, M., et al. "Temperature distribution in the high mountain regions on the Tibetan Plateau-Measurement and simulation." Proc. MODSIM 2007 International Congress on Modelling and Simulation. Modelling and Simulation Society of Australia and New Zealand. 2007.”.
- If the “temperature inversion” do matters, the conclusion about the snowmelt and air temperature parts need rewrite.
Citation: https://doi.org/10.5194/egusphere-2024-489-CC1 -
RC1: 'Comment on egusphere-2024-489', Manuel Tobias Blau, 11 May 2024
The manuscript "Seasonal Snow-Atmosphere Modeling: Let's do it" by Reynolds et al. introduces a new model setup based on the coupling of the HICAR with FSM2trans to resolve snow depth and snow-atmosphere interaction more effectively. For model performance demonstration, the authors applied the model setup to a case in the Swiss Alps and validated the model output with survey data from LiDAR flights. They found improvements in the model performance for surface temperature, blowing snow sublimation, and snow depth after coupling the snow component of FSM2trans (FSM2oshd) and using the ISHMAEL microphysics scheme that discriminates three types of ice states.
The well-written study may contribute to scientific advancement in snow modeling and other applications. Therefore, it may interest a broad range of scientists in various disciplines. However, a few more general points could be addressed and discussed more deeply in the manuscript.
When referring to figures, sometimes "Fig." and sometimes "Figure" are used. A uniform referencing system would be better.
Section 2.1: This section could be supported with a schematic diagram that outlines the model setup.
Section 2.2: Which of the redistribution processes was calculated first?
L 128: How sensitive is the model to different amounts of iterations?
Section 2.4: How sensitive is the model to the amount of snow layers? Does the top layer represent the surface that interacts with the atmosphere?
L. 170: There is a (TODO: HERE)?
L. 205: Which criteria are used for partitioning precipitation into snowfall and rainfall? How is the rain-snow slope line defined?
Section 3.1.1: What contributes most to the advection of snow particles and snowfall pattern during the event (wind aloft, above the ridge, or the updraft on the downwind slopes)?
Section 3.1.2: How much do the two discussed redistribution processes (wind-driven or gravitational redistribution) individually contribute to total redistribution? Can you give an estimate?
L. 333: It would be good if there were a brief statement of model performance in terms of meteorological variables earlier in the manuscript.
L. 355: "5" is "Figure 5"?
L.359: "The following paragraph ..." could be a new paragraph.
L. 363: Is the warm bias due to the too low albedo in high elevations?
How is snow albedo parameterized, and which values are assumed for fresh and accumulated snow?
L. 375: "4" is Figure 4"?
L. 435: Are the differences in the blowing snow sublimation percentage of snowfall due to differences in snowfall or the rate of sublimated snow?
Citation: https://doi.org/10.5194/egusphere-2024-489-RC1 - AC1: 'Reply on RC1', Dylan Reynolds, 30 May 2024
-
RC2: 'Comment on egusphere-2024-489', Anonymous Referee #2, 27 May 2024
This is a review of "Seasonal Snow-Atmosphere Modeling: Let’s do it”. Overall this is generally a quite well written and readable manuscript. The science is really exciting and at the forefront of the field.
My main suggestion is that the authors very clearly articulate the science questions that motivate this study. The introduction does a good job of setting up the research gaps, but I found many of the sections came at me a bit out of the blue. For example, I would really like to see clear science questions that establish the need for Morrison v. ISHMAEL, and specific method which the authors had a priori anticipated working well in this domain.
Some of the model decisions are behind in-prep manuscripts, which makes it a bit hard to follow some of the modelling decisions. For example, why the spatial resolutions were chosen. Indeed, it is pretty obvious to those of us who do this kind of modelling, but I think it would help readability if these model decisions were better articulated, the sensitive of the results to them, and for them to be tied to the underlying research questions. If this is tightened up, it should be a very easy-to-follow manuscript.
The domain has very few (none?) trees. Realizing it is a bit out of scope, I am very curious to know what the authors think of applying these methods to more forested catchments that have complex interactions of canopy-wind in the downslope regions.
w/c = Word choiceL20 Exposed to what?
L21 is it really a discontinuity though? The derivative remains well defined. I think this would be better restated as "sharp gradient" or similar
L23 "modified" w/c. Modified feels like a model input term
L34 space between amount an dunit – 100m -> 100 m. Fix throughout
L34 "at elevations" - "at a height above the surface of" might be more clear to match the language on L41
L57 "modern" -> contemporary?
L58 "troubling" w/c
L65 "These conditions have been followed by the earlier studies" what conditions are these and how can they be followed by an earlier study?
L66 " This is due to" unclear what this is in reference to
L67 horizontal and vertical resolutions?
L67 "One exception to this " what is this?
L67+ There is a lot of "this" throughout the section, which makes it difficult to follow. Generally I can infer what "this" is in reference to, but it would be best to explicitly state it to ease readability.
L72 "Caveat aside" which caveat is this in reference to?
L74 "adopting this strategy" this == intermediate complexity atmospheric models?
L85 what are the scientific questions, specifically?
L90 Later in the MS I found the mix of FSM2trans (e.g. fig 11), HICARsnow, NoahMP, a bit confusing. L 103 notes the name depends on the coupling strategy. I think a small table or a very clear description of all the comparisons would help readability
L93 define `oshd` ?
L96 static lib, are you meaning a `.a` ? This is very specific — do you mean this to preclude using a dynamic lib?
L109 FSM2 has a soil routine. How are frozen soils + the ground heatflux coupled into this? Or is this fully disabled in this configuration
L128 Is there an opportunity to use a dynamic iteration based on an error term versus a fixed iteration count ?
L125 “image” is not clear. I have a vague recollection this is maybe a iSNOBAL term? For the raster? Can you please clarity.
L170 fix TODO: HERE
L171 can you list, even briefly the details? Difficult to tell with an in-prep manuscript (but I do understand)
L235 Maybe it is just the layout, but it seems that the text goes from citing Fig 2 to Fig 6&7. My recommendation would be number the figures in the order they appear. As is, I’m finding it difficult to find the text that references a specific figure
L 275 Figure 7: What are the purple contours in the 2nd row? I don’t see them in the legend. I like this plot though.
L331 How is the surface roughness interaction with HICAR modelled?
L359 “The following paragraph” — suggest you start a new paragraph
L361 how big is “slight”?
L414 Do you think this excessive cooling impact ablation rates? I had originally read this as a FSM2 characteristic, but rereading the text I’m uncertain if it is actually NOAHMP. Please tighten this up a bit
L443 these tradeoffs should be clearly noted in the methodologyCitation: https://doi.org/10.5194/egusphere-2024-489-RC2 - AC2: 'Reply on RC2', Dylan Reynolds, 05 Jun 2024
Peer review completion
Journal article(s) based on this preprint
Model code and software
HICARsnow Model Code Dylan Reynolds https://doi.org/10.5281/zenodo.10679464
Video supplement
Preferential Deposition Processes Dylan Reynolds https://doi.org/10.16904/envidat.482
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 392 | 138 | 43 | 573 | 40 | 21 | 22 |
- HTML: 392
- PDF: 138
- XML: 43
- Total: 573
- Supplement: 40
- BibTeX: 21
- EndNote: 22
Viewed (geographical distribution)
| Country | # | Views | % |
|---|---|---|---|
| United States of America | 1 | 242 | 39 |
| Switzerland | 2 | 78 | 12 |
| China | 3 | 65 | 10 |
| Germany | 4 | 34 | 5 |
| Canada | 5 | 30 | 4 |
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
- 242
Louis Quéno
Michael Lehning
Mahdi Jafari
Justine Berg
Tobias Jonas
Michael Haugeneder
Rebecca Mott
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(27679 KB) - Metadata XML
-
Supplement
(225 KB) - BibTeX
- EndNote
- Final revised paper