the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 15 Jul 2024
Improving large-scale snow albedo modelling using a climatology of light-absorbing particle deposition
Abstract. Light-absorbing particles (LAPs) deposited at the snow surface significantly reduce its albedo and strongly affect the snow melt dynamics. The explicit simulation of these effects with advanced snow radiative transfer models is generally associated with a large computational cost. Consequently, many albedo schemes used in snowpack models still rely on empirical parameterizations that do not account for the spatial variability of LAP deposition. In this study, a new strategy of intermediate complexity that includes the effects of spatially variable LAP deposition on snow albedo is tested with the snowpack model Crocus. It relies on an optimization of the parameter that controls the evolution of snow albedo in the visible range. Optimized values for multi-year snow albedo simulations with Crocus were generated at ten reference experimental sites spanning a large variety of climates across the world. A regression was then established between these optimal values and climatological deposition of LAP on snow at the location of the experimental sites extracted from a global climatology developed in this study. This regression was finally combined with the global climatology to obtain an LAP-informed and spatially variable parameter for the Crocus albedo parameterization. The revised parameter improved snow albedo simulations on average by 10 % with the largest improvements found in the Arctic (more than 25 %). This methodology can be applied to other land surface models using the global climatology of LAP deposition on snow developed for this study.
Competing interests: Some authors are members of the editorial board of The Cryosphere.
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.-
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
(8042 KB)
-
Supplement
(1494 KB)
-
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(8042 KB) - Metadata XML
-
Supplement
(1494 KB) - BibTeX
- EndNote
- Final revised paper
Journal article(s) based on this preprint
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2024-1795', Anonymous Referee #1, 18 Aug 2024
Light-absorbing particles (LAPs, e.g., BC and dust) deposited at the snow surface significantly reduce snow albedo and strongly affect the snow dynamics. This study analyzed the relationship between LAP concentration and the optimized parameter that controls the evolution of snow albedo in Crocus model. However, there are limited sites in the study and cannot be used to upscale the impacts from site to globe. The spatial match between field measurements and gridded snow cover and LAP data is not well evaluated. The global simulations are not well evaluated and compared. The authors used climatological yearly-average data to analyze the relationship and neglected the seasonal variability and interannual variability of LAP concentration. Please see below for my specific comments.
Major concerns:
- Line 45: The authors stated that some snowpack models in LSMs often rely on empirical parameterizations. However, some widely-used LSMs, e.g., CLM and ELM, use a mechanism-based SNICAR model to simulate snow albedo with the consideration of LAP impacts. Indeed, they need the input of LAP deposition fluxes. Please state the limitation of such models.
- Line 54: The authors stated that the parameterizations use fixed time constants and were optimized using observation data. Looks like this study also used limited field data to optimize the models, which cannot ensure the accuracy at the untested grids. Please clearly explain it.
- Table 1: Please provide more information on the empirical equations. How did the authors set the values in the equations.
- Equation 1: the ratio of incoming radiation at the three bands can vary with sky conditions. The fixed ratio in this equation can induce some uncertainties. Please evaluate it and discuss it.
- In my view, snow age should be related to snow grain size. Please explain the relationship between d_opt and snow aging coefficient. I am not sure whether calling gamma as snow aging coefficient is reasonable or not. How did the authors model snow age?
- Section 3.2.1: Please briefly introduce the accuracy of two products: LAP and snow cover.
- The authors neglected the impacts of brown carbon. Please at least discuss the potential uncertainty.
- The authors matched the field measurements (snow albedo) and coarse simulated gridded data. This process may cause large uncertainties. Please evaluate the site spatial representativeness before do such spatial matching.
- As I know, the interannual variability of LAP deposition is very large. However, the authors used the climatological data in the study. Please evaluate whether such use is reasonable.
- Considering just limited sites and poor spatial representativeness, upscaling from site to globe is not reliable. I don’t think Figure 12 is accuracy enough. I suggest the authors remove the results related to global mapping. At least, the authors need to evaluate the global simulations with and without optimizing the parameters.
- How does the model simulate the temporal evolution of LAP impacts? As I know, the LAP concentration in snow varies with time largely. Please show whether the optimized parameters improve the seasonal variations and interannual variability.
Minor concerns:
- Line 20-23: Topography can also affect snow albedo.
- Line 54: ‘…’ _> etc.
- Line 82: What is this variable?
- Figure 2: Please show the snow types in the figure.
- Line 171: How did the authors calculate the daily albedo?
- Please check Appendix A: Parameterizations used in Crocus.
- Equation 3-4: How about R square?
- Figure 5: Is the site difference related to snow type?
Citation: https://doi.org/10.5194/egusphere-2024-1795-RC1 - AC1: 'Reply on RC1', Vincent Vionnet, 18 Oct 2024
-
RC2: 'Comment on egusphere-2024-1795', Anonymous Referee #2, 25 Aug 2024
The review of “Improving large-scale snow albedo modeling using a climatology of light-absorbing particle deposition” by Gaillard et al.
This research project aims to represent the effect of Light-Absorbing Particles (LAP) on snow albedo within the Crocus model as snow aging parameters, and to establish a global climatology of LAP deposition on snow, with the intention of applying these results to land surface models. The effort to achieve a global climatology of this phenomenon, by accounting for the altitude dependence of snow aging and making it regionally specific, is commendable. However, there has been insufficient discussion regarding the methodology for estimating the parameters related to snow aging and LAP climatology when scaling this approach globally. Consequently, it cannot be stated that the reliability of the climatology related to LAP deposition, which is the final outcome, is fully assured. In particular, the following points must be thoroughly discussed:
(major comments)
1. As the author demonstrates, the parameters associated with snow aging show significant differences between the accumulation and melting periods (Fig. 3). However, this study only considered the regional dependency of snow aging. It would be more beneficial for the authors to evaluate snow aging separately for the accumulation and melting periods and then discuss the relationship between snow aging and LAP deposition on the snow. In other words, the climatology of LAP deposition will be influenced not only by altitude but also by the timing of the accumulation and melting periods.
2. (L173) If the optical thickness of the snow cover is insufficient, the snow albedo will be influenced by the albedo (reflectance) of the ground surface beneath the snow cover. Consequently, γ (gamma) will also reflect the influence of the ground surface. In forested areas, the effect of vegetation must be considered as well. Therefore, γ is not only affected by snow aging but may also be strongly influenced by local factors. It is necessary to provide a comprehensive explanation of this point when determining the relationship between D (climatological deposition rate of LAPs) and γ.
3. The climatology of LAP deposition will be a valuable dataset for global land surface models. However, the results presented in Fig. 12 have not been fully validated, and their accuracy, including associated uncertainties, remains unclear. Additionally, because the validation sites are limited, there are likely to be high uncertainties in the LAP data for regions such as South Asia, particularly in mountain glaciers. The authors should release the dataset only after thorough validation.
(minor commnets)
1. It is understood that Crocus is a multi-layer model used to simulate snow conditions across various layers. However, only two snow grain size parameters (d_opt and d') are employed to calculate the albedo in the visible and near-infrared regions (Table 1). In the visible range, ice exhibits weak light absorption, allowing light to penetrate more deeply. Therefore, to accurately calculate the albedo in the visible range, it is crucial to consider the vertical distribution of snow grain sizes. Please add an explanation of which snowpack layer d_opt parameter represents.
2. Please explain how the solar zenith angle dependence of snow albedo and the effects of the atmosphere, especially clouds, on snow albedo can be expressed in terms of Eq. 1.
3. For reference, please also show the results for SSA=10 m^2kg^-1 (~granular snow) in Fig. 1.
Citation: https://doi.org/10.5194/egusphere-2024-1795-RC2 - AC2: 'Reply on RC2', Vincent Vionnet, 18 Oct 2024
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2024-1795', Anonymous Referee #1, 18 Aug 2024
Light-absorbing particles (LAPs, e.g., BC and dust) deposited at the snow surface significantly reduce snow albedo and strongly affect the snow dynamics. This study analyzed the relationship between LAP concentration and the optimized parameter that controls the evolution of snow albedo in Crocus model. However, there are limited sites in the study and cannot be used to upscale the impacts from site to globe. The spatial match between field measurements and gridded snow cover and LAP data is not well evaluated. The global simulations are not well evaluated and compared. The authors used climatological yearly-average data to analyze the relationship and neglected the seasonal variability and interannual variability of LAP concentration. Please see below for my specific comments.
Major concerns:
- Line 45: The authors stated that some snowpack models in LSMs often rely on empirical parameterizations. However, some widely-used LSMs, e.g., CLM and ELM, use a mechanism-based SNICAR model to simulate snow albedo with the consideration of LAP impacts. Indeed, they need the input of LAP deposition fluxes. Please state the limitation of such models.
- Line 54: The authors stated that the parameterizations use fixed time constants and were optimized using observation data. Looks like this study also used limited field data to optimize the models, which cannot ensure the accuracy at the untested grids. Please clearly explain it.
- Table 1: Please provide more information on the empirical equations. How did the authors set the values in the equations.
- Equation 1: the ratio of incoming radiation at the three bands can vary with sky conditions. The fixed ratio in this equation can induce some uncertainties. Please evaluate it and discuss it.
- In my view, snow age should be related to snow grain size. Please explain the relationship between d_opt and snow aging coefficient. I am not sure whether calling gamma as snow aging coefficient is reasonable or not. How did the authors model snow age?
- Section 3.2.1: Please briefly introduce the accuracy of two products: LAP and snow cover.
- The authors neglected the impacts of brown carbon. Please at least discuss the potential uncertainty.
- The authors matched the field measurements (snow albedo) and coarse simulated gridded data. This process may cause large uncertainties. Please evaluate the site spatial representativeness before do such spatial matching.
- As I know, the interannual variability of LAP deposition is very large. However, the authors used the climatological data in the study. Please evaluate whether such use is reasonable.
- Considering just limited sites and poor spatial representativeness, upscaling from site to globe is not reliable. I don’t think Figure 12 is accuracy enough. I suggest the authors remove the results related to global mapping. At least, the authors need to evaluate the global simulations with and without optimizing the parameters.
- How does the model simulate the temporal evolution of LAP impacts? As I know, the LAP concentration in snow varies with time largely. Please show whether the optimized parameters improve the seasonal variations and interannual variability.
Minor concerns:
- Line 20-23: Topography can also affect snow albedo.
- Line 54: ‘…’ _> etc.
- Line 82: What is this variable?
- Figure 2: Please show the snow types in the figure.
- Line 171: How did the authors calculate the daily albedo?
- Please check Appendix A: Parameterizations used in Crocus.
- Equation 3-4: How about R square?
- Figure 5: Is the site difference related to snow type?
Citation: https://doi.org/10.5194/egusphere-2024-1795-RC1 - AC1: 'Reply on RC1', Vincent Vionnet, 18 Oct 2024
-
RC2: 'Comment on egusphere-2024-1795', Anonymous Referee #2, 25 Aug 2024
The review of “Improving large-scale snow albedo modeling using a climatology of light-absorbing particle deposition” by Gaillard et al.
This research project aims to represent the effect of Light-Absorbing Particles (LAP) on snow albedo within the Crocus model as snow aging parameters, and to establish a global climatology of LAP deposition on snow, with the intention of applying these results to land surface models. The effort to achieve a global climatology of this phenomenon, by accounting for the altitude dependence of snow aging and making it regionally specific, is commendable. However, there has been insufficient discussion regarding the methodology for estimating the parameters related to snow aging and LAP climatology when scaling this approach globally. Consequently, it cannot be stated that the reliability of the climatology related to LAP deposition, which is the final outcome, is fully assured. In particular, the following points must be thoroughly discussed:
(major comments)
1. As the author demonstrates, the parameters associated with snow aging show significant differences between the accumulation and melting periods (Fig. 3). However, this study only considered the regional dependency of snow aging. It would be more beneficial for the authors to evaluate snow aging separately for the accumulation and melting periods and then discuss the relationship between snow aging and LAP deposition on the snow. In other words, the climatology of LAP deposition will be influenced not only by altitude but also by the timing of the accumulation and melting periods.
2. (L173) If the optical thickness of the snow cover is insufficient, the snow albedo will be influenced by the albedo (reflectance) of the ground surface beneath the snow cover. Consequently, γ (gamma) will also reflect the influence of the ground surface. In forested areas, the effect of vegetation must be considered as well. Therefore, γ is not only affected by snow aging but may also be strongly influenced by local factors. It is necessary to provide a comprehensive explanation of this point when determining the relationship between D (climatological deposition rate of LAPs) and γ.
3. The climatology of LAP deposition will be a valuable dataset for global land surface models. However, the results presented in Fig. 12 have not been fully validated, and their accuracy, including associated uncertainties, remains unclear. Additionally, because the validation sites are limited, there are likely to be high uncertainties in the LAP data for regions such as South Asia, particularly in mountain glaciers. The authors should release the dataset only after thorough validation.
(minor commnets)
1. It is understood that Crocus is a multi-layer model used to simulate snow conditions across various layers. However, only two snow grain size parameters (d_opt and d') are employed to calculate the albedo in the visible and near-infrared regions (Table 1). In the visible range, ice exhibits weak light absorption, allowing light to penetrate more deeply. Therefore, to accurately calculate the albedo in the visible range, it is crucial to consider the vertical distribution of snow grain sizes. Please add an explanation of which snowpack layer d_opt parameter represents.
2. Please explain how the solar zenith angle dependence of snow albedo and the effects of the atmosphere, especially clouds, on snow albedo can be expressed in terms of Eq. 1.
3. For reference, please also show the results for SSA=10 m^2kg^-1 (~granular snow) in Fig. 1.
Citation: https://doi.org/10.5194/egusphere-2024-1795-RC2 - AC2: 'Reply on RC2', Vincent Vionnet, 18 Oct 2024
Peer review completion
Journal article(s) based on this preprint
Data sets
Improved snow ageing parameter for large-scale albedo modelling with Crocus Manon Gaillard, Vincent Vionnet, Matthieu Lafaysse, Marie Dumont, and Paul Ginoux https://zenodo.org/doi/10.5281/zenodo.11554925
Global climatology of light-absorbing particle deposition on snow Manon Gaillard, Vincent Vionnet, Matthieu Lafaysse, Marie Dumont, and Paul Ginoux https://zenodo.org/doi/10.5281/zenodo.11554782
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 333 | 132 | 53 | 518 | 35 | 19 | 20 |
- HTML: 333
- PDF: 132
- XML: 53
- Total: 518
- Supplement: 35
- BibTeX: 19
- EndNote: 20
Viewed (geographical distribution)
| Country | # | Views | % |
|---|---|---|---|
| United States of America | 1 | 233 | 43 |
| France | 2 | 39 | 7 |
| China | 3 | 33 | 6 |
| Canada | 4 | 27 | 5 |
| Germany | 5 | 21 | 3 |
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
- 233
Manon Gaillard
Matthieu Lafaysse
Marie Dumont
Paul Ginoux
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(8042 KB) - Metadata XML
-
Supplement
(1494 KB) - BibTeX
- EndNote
- Final revised paper