the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 25 Jun 2024
Simulation of snow albedo and solar irradiance profile with the two-stream radiative transfer in snow (TARTES) v2.0 model
Abstract. The Two-streAm Radiative TransfEr in Snow (TARTES) model computes the spectral albedo and the profiles of spectral absorption, irradiance and actinic fluxes for a multi-layer plane-parallel snowpack. Each snow layer is characterized by its specific surface area, density, and impurities content, in addition to shape parameters. In the landscape of snow optical numerical models, TARTES distinguishes itself by taking into account different shapes of the particles through two shape parameters, namely the absorption enhancement parameter B and the asymmetry factor g. This is of primary importance as recent studies working at the microstructure level have demonstrated that snow does not behave as a collection of equivalent ice spheres, a representation widely used in other models. Instead, B and g take specific values that do not correspond to any simple geometrical shape, which leads to the concept of "optical shape of snow". Apart from this specificity, TARTES combines well established radiative transfer principles to compute the scattering and absorption coefficients of pure or polluted snow, and the δ-Eddington two-stream approximation to solve the multi-layer radiative transfer equation. The model is implemented in Python, but conducting TARTES simulations is also possible without any programming through the SnowTARTES web application, making it very accessible to non-experts and for teaching purposes. Here, after describing the theoretical and technical details of the model, we illustrate its main capabilities and present some comparisons with other common snow radiative transfer models (AART, DISORT-Mie, SNICAR-ADv3) as a validation procedure. Overall the agreement on the spectral albedo, when in compatible conditions (i.e. with spheres), is usually within 0.02, and is better in the visible and near-infrared compared to longer wavelengths of the solar domain.
-
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
(948 KB)
-
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(948 KB) - Metadata XML
- BibTeX
- EndNote
- Final revised paper
Journal article(s) based on this preprint
Interactive discussion
Status: closed
-
RC1: 'Review Comment on egusphere-2024-1176', Anonymous Referee #1, 23 Jul 2024
The authors provide a very comprehensive description of TARTES model, which serves as a very good technical documentation for the model. They also compared the TARTES model results with a few other widely-used snow radiative transfer models (AART, DISORT-Mie, SNICAR-ADv3), which generally show reasonable agreement (within 0.02 for snow albedo). Overall, the manuscript is well written. I have a few minor comments for the authors to address.
Minor comments:
- It is not clear what the specific improvements in TARTES version 2 are compared to version 1, which needs to be clarified.
- It would be good to also summarize/discuss some underlying assumptions that may limit the TARTES model applications or the cautions users need to take when applying the model.
- Lines 53-78: These widely-used snow albedo/radiative transfer models have been reviewed and described in details in He and Flanner (2020; https://doi.org/10.1007/978-3-030-38696-2_3), which would be a resource to mention and refer the interested readers to.
- Equation (4): please introduce the delta function and F0 variable.
- It would be good to have a table listing all the input and output variables (with units) for TARTES.
- Equations (36-37): please give the mathematical expressions of “A” and “B”.
- Line 244: “… we incorporate the exponential terms in the vector X”. This sentence is not very clear to me.
- Equations (53-55): It seems that B is not included in these equations, which however should be, right?
- Line 300: For the negative albedo case, did the authors also reset other relevant quantities (e.g., internal light absorption energy, actinic flux, etc.)? If so, reset to what values?
- Line 320: Some studies (e.g., Peltoniemi, 2007: https://doi.org/10.1016/j.jqsrt.2007.05.009; He et al., 2017: https://doi.org/10.1002/2017GL072916) have quantified the impact of treating snow grains as densely packed medium on snow albedo, which is worth discussing briefly here.
- Section 3.2: It is not very clear what the key differences in model physics and/or parameters between the Python and Fortran versions, which needs to be clarified.
- Line 506: It should note that this conclusion here is applied to semi-infinite snowpack tested in this study.
- It would be good to add a short paragraph to discuss future plans for TARTES model improvements/developments.
Citation: https://doi.org/10.5194/egusphere-2024-1176-RC1 - AC1: 'Reply on RC1', Ghislain Picard, 23 Aug 2024
-
RC2: 'Comment on egusphere-2024-1176', Mark Flanner, 26 Jul 2024
This manuscript provides a comprehensive technical description of the TARTES v2.0 radiative transfer model for snowpack. TARTES is widely used by the community and is embedded in the CROCUS snow thermodynamic model. A unique and valuable feature of TARTES is the representation of ice particle asphericity via two parameters that can vary continuously and are not tied to any particular shapes, thus enabling the representation of the "optical shape" of snow via a continuum. The manuscript is well-written and appropriately includes both technical descriptions and comparisons against other snowpack radiative transfer models. The introduction is well-referenced and provides useful background to the topic. I have only minor comments and am happy to recommend publication of the manuscript in GMD.
Minor comments:line 30: "understand" -> "understanding"
lines 34-36: "SSA ... advantageously replaces the grain size as it can be rigorously defined and calculated for any porous medium" - While I agree about the advantages of SSA, "calculating" the surface area of complex shapes and porous media is often non-trivial. To me, the main advantage of SSA is simply that it is a well-defined physical quantity, whereas the meaning of effective radius can be unclear for complex shapes and porous media.
line 107: "an" -> "a"
line 123: The latter part of the sentence also refers to diffuse illumination, so perhaps change the first party to "... illuminated by a beam source and diffuse light..."
line 163: "into" -> "in"
line 244: The appearance and application of these exponential terms is not immediately clear to me. Could you please clarify or elaborate on this point?
line 373: "When impurities are added in realistic low quantities, we assume the extinction coefficient of snow is unchanged..." - Although this is certainly a valid approximation for most realistic impurity mixing ratios, there may be situations where high impurity loads appreciably affect the total extinction coefficient, particularly in the near-IR spectrum. For example, very high dust loads can flatten out the 1.03um ice absorption feature (e.g., Fair et al, 2022, https://tc.copernicus.org/articles/16/3801/2022/) . Hence, it might be useful to provide a rough upper limit to the "low quantity" that applies for this assumption to hold, or in general to define the limits of applicability for model users.
line 466: Briefly, what are the user-controlled inputs to this "atmospheric_incident_spectrum" function?
line 492: The accuracy of the delta-Eddington technique in handling diffuse incident light was also assessed more recently by Dang et al (2019, https://doi.org/10.5194/tc-13-2325-2019)
line 576: "benefice" -> "benefit"
p.29 and Fig. 11: The comparison provided is useful because it shows how the default representations of BC in each model affect the simulated albedo. Because the BC optical properties are slightly different in each model, one could further explore the sources of differences between TARTES and SNICAR by imposing identical BC properties. This could be accomplished, e.g., by directly importing the BC MAE values from the SNICAR optics library into TARTES, similar to how dust MAE is used in TARTES. I am not requesting this for manuscript revisions, but merely highlighting it as an informative sensitivity study for the future.
line 605: "67um": Typo, should be "67nm"
line 605: 67nm appears to be close to the monodisperse radius of maximum MAE. Because the SNICAR BC properties are mass-weighted from a lognormal distribution, the MAE value at the effective radius corresponding to the maximum MAE for a monodisperse distribution (i.e., the trough in Fig. 12) will always be less than the monodisperse MAE at that radius, since the weighted average incorporates lower MAE values from both sides of that radius. This could explain the phenomenon described in this part of the text. But other model factors could also contribute to differences in the absolute simulated albedos shown in Fig 12. For example, are identical ice refractive indices used in each model simulation?
line 624: "This rates" -> "This rate"
line 638: "model as TARTES" -> "model such as TARTES"
line 654: For which wavelength or wavelength range does g=0.82 apply?
line 655: "These values are the defaults in TARTES v2.0" - Actually, it might be helpful to include a table of all of the default parameter settings in TARTES v2.0, but I leave this to the authors to decide.
Finally, I agree with a comment from the other referee that it would be useful to highlight differences and improvements between versions 2.0 and 1.0 of TARTES.
Citation: https://doi.org/10.5194/egusphere-2024-1176-RC2 - AC2: 'Reply on RC2', Ghislain Picard, 23 Aug 2024
Interactive discussion
Status: closed
-
RC1: 'Review Comment on egusphere-2024-1176', Anonymous Referee #1, 23 Jul 2024
The authors provide a very comprehensive description of TARTES model, which serves as a very good technical documentation for the model. They also compared the TARTES model results with a few other widely-used snow radiative transfer models (AART, DISORT-Mie, SNICAR-ADv3), which generally show reasonable agreement (within 0.02 for snow albedo). Overall, the manuscript is well written. I have a few minor comments for the authors to address.
Minor comments:
- It is not clear what the specific improvements in TARTES version 2 are compared to version 1, which needs to be clarified.
- It would be good to also summarize/discuss some underlying assumptions that may limit the TARTES model applications or the cautions users need to take when applying the model.
- Lines 53-78: These widely-used snow albedo/radiative transfer models have been reviewed and described in details in He and Flanner (2020; https://doi.org/10.1007/978-3-030-38696-2_3), which would be a resource to mention and refer the interested readers to.
- Equation (4): please introduce the delta function and F0 variable.
- It would be good to have a table listing all the input and output variables (with units) for TARTES.
- Equations (36-37): please give the mathematical expressions of “A” and “B”.
- Line 244: “… we incorporate the exponential terms in the vector X”. This sentence is not very clear to me.
- Equations (53-55): It seems that B is not included in these equations, which however should be, right?
- Line 300: For the negative albedo case, did the authors also reset other relevant quantities (e.g., internal light absorption energy, actinic flux, etc.)? If so, reset to what values?
- Line 320: Some studies (e.g., Peltoniemi, 2007: https://doi.org/10.1016/j.jqsrt.2007.05.009; He et al., 2017: https://doi.org/10.1002/2017GL072916) have quantified the impact of treating snow grains as densely packed medium on snow albedo, which is worth discussing briefly here.
- Section 3.2: It is not very clear what the key differences in model physics and/or parameters between the Python and Fortran versions, which needs to be clarified.
- Line 506: It should note that this conclusion here is applied to semi-infinite snowpack tested in this study.
- It would be good to add a short paragraph to discuss future plans for TARTES model improvements/developments.
Citation: https://doi.org/10.5194/egusphere-2024-1176-RC1 - AC1: 'Reply on RC1', Ghislain Picard, 23 Aug 2024
-
RC2: 'Comment on egusphere-2024-1176', Mark Flanner, 26 Jul 2024
This manuscript provides a comprehensive technical description of the TARTES v2.0 radiative transfer model for snowpack. TARTES is widely used by the community and is embedded in the CROCUS snow thermodynamic model. A unique and valuable feature of TARTES is the representation of ice particle asphericity via two parameters that can vary continuously and are not tied to any particular shapes, thus enabling the representation of the "optical shape" of snow via a continuum. The manuscript is well-written and appropriately includes both technical descriptions and comparisons against other snowpack radiative transfer models. The introduction is well-referenced and provides useful background to the topic. I have only minor comments and am happy to recommend publication of the manuscript in GMD.
Minor comments:line 30: "understand" -> "understanding"
lines 34-36: "SSA ... advantageously replaces the grain size as it can be rigorously defined and calculated for any porous medium" - While I agree about the advantages of SSA, "calculating" the surface area of complex shapes and porous media is often non-trivial. To me, the main advantage of SSA is simply that it is a well-defined physical quantity, whereas the meaning of effective radius can be unclear for complex shapes and porous media.
line 107: "an" -> "a"
line 123: The latter part of the sentence also refers to diffuse illumination, so perhaps change the first party to "... illuminated by a beam source and diffuse light..."
line 163: "into" -> "in"
line 244: The appearance and application of these exponential terms is not immediately clear to me. Could you please clarify or elaborate on this point?
line 373: "When impurities are added in realistic low quantities, we assume the extinction coefficient of snow is unchanged..." - Although this is certainly a valid approximation for most realistic impurity mixing ratios, there may be situations where high impurity loads appreciably affect the total extinction coefficient, particularly in the near-IR spectrum. For example, very high dust loads can flatten out the 1.03um ice absorption feature (e.g., Fair et al, 2022, https://tc.copernicus.org/articles/16/3801/2022/) . Hence, it might be useful to provide a rough upper limit to the "low quantity" that applies for this assumption to hold, or in general to define the limits of applicability for model users.
line 466: Briefly, what are the user-controlled inputs to this "atmospheric_incident_spectrum" function?
line 492: The accuracy of the delta-Eddington technique in handling diffuse incident light was also assessed more recently by Dang et al (2019, https://doi.org/10.5194/tc-13-2325-2019)
line 576: "benefice" -> "benefit"
p.29 and Fig. 11: The comparison provided is useful because it shows how the default representations of BC in each model affect the simulated albedo. Because the BC optical properties are slightly different in each model, one could further explore the sources of differences between TARTES and SNICAR by imposing identical BC properties. This could be accomplished, e.g., by directly importing the BC MAE values from the SNICAR optics library into TARTES, similar to how dust MAE is used in TARTES. I am not requesting this for manuscript revisions, but merely highlighting it as an informative sensitivity study for the future.
line 605: "67um": Typo, should be "67nm"
line 605: 67nm appears to be close to the monodisperse radius of maximum MAE. Because the SNICAR BC properties are mass-weighted from a lognormal distribution, the MAE value at the effective radius corresponding to the maximum MAE for a monodisperse distribution (i.e., the trough in Fig. 12) will always be less than the monodisperse MAE at that radius, since the weighted average incorporates lower MAE values from both sides of that radius. This could explain the phenomenon described in this part of the text. But other model factors could also contribute to differences in the absolute simulated albedos shown in Fig 12. For example, are identical ice refractive indices used in each model simulation?
line 624: "This rates" -> "This rate"
line 638: "model as TARTES" -> "model such as TARTES"
line 654: For which wavelength or wavelength range does g=0.82 apply?
line 655: "These values are the defaults in TARTES v2.0" - Actually, it might be helpful to include a table of all of the default parameter settings in TARTES v2.0, but I leave this to the authors to decide.
Finally, I agree with a comment from the other referee that it would be useful to highlight differences and improvements between versions 2.0 and 1.0 of TARTES.
Citation: https://doi.org/10.5194/egusphere-2024-1176-RC2 - AC2: 'Reply on RC2', Ghislain Picard, 23 Aug 2024
Peer review completion
Journal article(s) based on this preprint
Model code and software
TARTES pre-v2.0 source code Ghislain Picard and Quentin Libois https://doi.org/10.5281/zenodo.12103855
Interactive computing environment
Snowtartes web application Ghislain Picard https://snow.univ-grenoble-alpes.fr/snowtartes/
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 339 | 134 | 29 | 502 | 17 | 14 |
- HTML: 339
- PDF: 134
- XML: 29
- Total: 502
- BibTeX: 17
- EndNote: 14
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
Ghislain Picard
Quentin Libois
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(948 KB) - Metadata XML