the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 11 Aug 2022
A modern-day Mars climate in the Met Office Unified Model: dry simulations
Abstract. We present results from the Met Office Unified Model (UM), a world-leading climate and weather model, adapted to simulate a dry Martian climate. We detail the adaptation of the basic parameterisations and analyse results from two simulations, one with radiatively active mineral dust, and one with radiatively inactive dust. These simulations demonstrate how the radiative effects of dust act to accelerate the winds and create a mid-altitude isothermal layer during the dusty season. We validate our model through comparison with an established Mars model, the Laboratoire de Météorologie Dynamique Mars Planetary Climate Model (PCM), finding good agreement in the seasonal wind and temperature profiles, but discrepancies in the predicted dust mass mixing ratio and conditions at the poles. This study validates the use of the UM for a Martian atmosphere, it highlights how the adaptation of an Earth GCM can be beneficial for existing Mars GCMs and provides insight into the next steps in our development of a new Mars climate model.
-
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
(13863 KB)
-
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(13863 KB) - Metadata XML
- BibTeX
- EndNote
- Final revised paper
Journal article(s) based on this preprint
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2022-718', Anonymous Referee #1, 23 Aug 2022
This manuscript provides an introduction to the modern Mars version of the Met Office United Model (UM). The paper describes the foundational adjustments that were made to the model to adapt it to modern Mars and then evaluates its performance relative to the LMD Mars Planetary Climate Model (PCM) and compares UM simulations with (RA) and without (RI) radiatively active dust. The model is still in development and currently lacks CO2 and H2O cycles. The model broadly reproduces the martian dust cycle and annual thermal and dynamical conditions. It will be useful to the scientific community to have another Mars GCM, but with terrestrial-derived physical parameterizations to help better understand the physics of the martian climate. I recommend the manuscript for publication after a moderate revision.
General comments:
- Substantially more discussion is needed regarding several aspects of the modeled dust particles and dust cycle in general.
- First, only a single sentence is used to describe the dust optical properties used. Based on the reference, I assume this implies that a terrestrial dust composition and optical properties are used. But that reference also refers to other references with those details. The authors should summarize the relevant information in the text and also compare and contrast that with known properties of martian dust (chemistry, optical properties, albedo, etc.). Additionally, discuss how these differences could impact the simulated climate.
- Second, how is dust lifted in the model? It seems (based on a few sentences throughout the text) that the model only lifts dust through wind stress-driven saltation. Is that correct? Observations have shown that dust devils likely supply ~50% of dust into the atmosphere, so it should be mentioned explicitly if there is no parameterization for dust devils. I have some additional comments below that relate to this.
- Given the paper is focused on the dust cycle and its impact on the climate, I think another figure or two on the modeled dust cycle would be quite helpful. Specifically, a figure showing the globally averaged aerosol optical depth as a function of season for RA and RI simulations, perhaps with the PCM overplotted, and a second figure identifying model grid points favored for dust lifting and deposition. See Gebhardt et al. (2019) for similar figures: https://agupubs.onlinelibrary.wiley.com/doi/full/10.1029/2019JE006253. I think such figures could replace Figures 1 and 2 if desired.
- I don’t think Figures 6-9 are structured in the most effective manner. The difference plot in the center is intuitively for the differences between the left and right columns, but that isn’t the case here. And differences between the RA simulation and the PCM are discussed in the text, which leaves the reader having to mentally try and create such a difference plot. I don’t think the authors need to double the number of plots necessarily (one set comparing RA and RI, another comparing RA and PCM), but I do think the visual comparison between RA and RI and RA and PCM should be separated.
Specific comments:
Line 98: Are spectral files for the ancient Sun (3.8gya) being used here?? Or is that just a typo?
Table 1: Maybe change the title from ‘orbital parameters’ to ‘physical constants’ or something else as the gas constant and scale height are mentioned? Additionally, is the scale height prescribed as a constant? How does that work? That value seems high relative to Mars (~11 km).
Section 2.3: There’s no mention of surface albedo or thermal inertia here or elsewhere. Are martian values used as boundary conditions? Something else?
Line 155: Is an effective dust radius of 3um assumed as it was in Madeleine et al. (2011)? Some other value? Specify it here.
Line 167: Does it really take 40 years to reach steady state? Why so long, especially without H2O or CO2 cycles?
Line 221: “meaned” reads oddly to me. “Averaged”?
Line 237: should “to” be “in” here?
Lines 239-240: it looks like 1-2 m/s and 1 K differences to my eyes?
Line 251: “in each month” should be “in this month”?
Line 252: This difference perhaps could be somewhat ameliorated by tuning the dust cycle, however? Was any tuning done to the dust cycle? This goes back to general comment #1. Lifting efficiency factors, etc., are typical tuning parameters for Mars GCMs.
Line 256: “reverse” should be “reverses”
Figure 7 caption: specify that negative values indicate southward winds.
Section 4.2. titles: include Ls range with each month title (e.g., “Month 9: Ls 270-300°”).
Line 314: “RI” here should be “RA”
Figure 9: I appreciate the reason to use different color bar scales for RA and PCM here, but it’s still quite confusing by eye when the RA plots have deeper red colors. Can the color bars be adjusted to help the reader visually intuit that RA often has less dust? Additionally, percent difference (as is used in the text) might be more intuitive for the difference plot in the center in this case.
Section 4.2.3 and Figure 9: One consistent difference (at least for months 6-12) between RA and RI seems to be that RA lifts dust farther south than RI? Is that correct? Is that a feedback effect between radiatively active dust and the Hadley circulation perhaps? This may be worth some discussion in the text.
Line 322: I’m confused by the sentence beginning with “The high vertical uplifting…”. This sentence makes it sound like the UM has a “rocket dust storm” parameterization, which I don’t think is true. So what is really meant by “high vertical uplifting”? What physical mechanism is bringing higher dust mixing ratios to high heights in UM-RA? Months 6 and 9 seem to natively produce a high altitude dust layer (Heavens et al., 2011; Guzewich et al., 2013), which I think would be quite novel for a free-running dust simulation! More discussion is needed here.
Section 4.2.3: calling back to my general comment #2, does dust opacity peak in month 9 in UM-RA and UM-RI? That is the typical pattern for free-running dust simulations in Mars GCMs, but the real Mars often has a “solstitical pause” in activity near southern summer solstice outside of years with solstitical global dust storms. I imagine PCM reflects this since it uses the Montabone dust climatology.
Line 347: The sentence beginning with “Dust abundance is higher…” is essentially repeated in the next paragraph. Similarly, saying a simulation is “higher” by a negative number is confusing.
Line 378: “Once again there is a difference…” of zero?
Line 380: Missing “Fig.”
Line 381: specify this difference is only in RA and RI, not in PCM simulations
Section 4.3.3: Dust devils aren’t mentioned at all! The most likely reason for this disparity, in my opinion, is that UM doesn’t have a dust devil parameterization. Wind stress lifting is very low during the early portion of the year, while real Mars (and hence the PCM) maintains a background dust haze through dust devils. Kahre et al. (2017) and references therein discuss this in detail.
Line 471: typo “whata”
Citation: https://doi.org/10.5194/egusphere-2022-718-RC1 - Substantially more discussion is needed regarding several aspects of the modeled dust particles and dust cycle in general.
-
RC2: 'Comment on egusphere-2022-718', Anonymous Referee #2, 14 Sep 2022
This manuscript documents two simulations from a new Mars global circulation model, ported from the UK Met Unified Model. The addition of new, independent global models is welcome in the community and should always be encouraged. Its development will increase the capabilities of the modeling community. In advocating for the development of a new Mars model, the manuscript also makes a strong case for the need for a Mars Atmosphere Inter-comparison Project. CMIP has been extremely successful for the assessment of future climates for Earth, and a MAIP might hold similar promise for Mars.
Generally, the model proves capable in simulating many of the large-scale features of the Martian climate. The model reproduces a dust cycle, which bears a reasonably good resemblance to that of Mars, without the need for forcing the simulation to observations. This is an impressive advancement. The manuscript is well-written and organized; however, a couple of discussion points are neglected, and most importantly, a key process is not included in the model. See main comments below and annotated PDF for minor comments.
Major comments:
I.) Most critical of the major comments: I understand that porting a terrestrial model to Mars is a substantial undertaking, but a Mars model lacking the CO2 cycle—and therefore having surfaces pressure being up to 20% too large for a given grid cell—seems like a massive (literally and figuratively) potential source of error. The authors recognize the need for a CO2 cycle on lines 521–531. A non-exhaustive list of the potential problems in simulating a realistic Martian climate without the CO2 cycle include: 1.) incorrect tidal amplitudes, since a given radiative forcing will, for a large part of the year according to Fig. 5, be working on more mass than actually exists. Because the tides are so important for closure of, for example, momentum budgets, getting this wrong makes the entirety of the presented results less robust. 2.) The reality of the radiative influence of dust may also suffer. If X amount of dust is lifted, the mixing ratio of X amount of dust would be less than reality if there was an incorrectly excessive amount of non-dust mass in the atmosphere. This comparatively thinner dust layer may not necessarily change the surface temperature because the total extinction would be nearly the same, but it might change the vertical temperature profile based on changed vertical distribution of dust. 3.) Similar arguments could be made on the impact spurious mass might have on various wave modes, but without running experiments, it is hard to know the non-linear changes on a model never adapted for Mars before, which is the point. 4.) Finally, the CO2 cycle is associated with a flow of the atmosphere from one pole to the other as the polar caps sublimate and deposit. This flow, while comparatively smaller the zonal winds of the Hadley cell itself, is still missing, and with it one of the important potential mechanisms for dust lifting in the high latitudes.While overall, the simulated climate looks reasonable; as plotted in sigma coordinates, the effect of an incorrect surface pressure is obviated. I am left to wonder if the model would look even better if this neglected process were included; conversely, I am concerned that not including this process is hiding other issues that are not yet apparent. I would strongly encourage the authors to investigate the feasibility of incorporating the CO2 cycle in the present version rather than saving for a future manuscript. At least, a simplified parameterization as noted used by other models around line 525, could be attempted. The process by forcing the atmospheric mass to a prescribed surface pressure or enforcing mass sources/sinks as need in the poles at the appropriate times of year might be sufficient.
II.) I missed a discussion on dust lifting. How is the process parameterized? How is the surface dust reservoir calculated? There is a brief mention in the results section on Line 320 that the UM calculates reservoirs and the horizontal motion, but this warrants a more complete description in the methods. Only on line 449, the fact that the dust reservoir is infinite is finally mentioned. This all needs to be organized into a specific section in the methods. It is impressive, as noted around Line 340, that a dust cycle is reproduced in the model without forcing, but it is difficult to assess how robust the cycle is without knowing how lifting is parameterized. One particular detail that appears contrary to the observed dust cycle is that month "9" of Ls~270 should have reduced dust MMR than months 6 or 12 (Montabone et al., 2015), but that is not the case.
III.) Is there any sensitivity to model resolution (predominantly horizontal but vertical as well)?
IV.) The results and discussion focused on the zonal-mean structure. This is an excellent way to put the bulk climate into context but is not the full picture. As the goal is a comparison of the UM MCGM to the PCM, at least some investigation of the non-zonal mean structure is warranted. The addition of at least a few plots showing the column optical depth for each season on a lat/lon figure would show that dust is being transported within the circulation in a realistic way. Similarly, plots of the surface temperature and pressure would demonstrate how poorly or how well the model manages to capture the true climate without a CO2 cycle (Major comment I).
-
AC1: 'Comment on egusphere-2022-718', Danny McCulloch, 26 Nov 2022
We thank the reviewers for their time and valuable comments. We have amended/adjusted the manuscript, where alterations can be found in green text and old text features a strike-through. A detailed response with the original comment and specific replies is provided as a supplement which features a description of the added content.
We also thank the editor for their understanding in extending the deadline for the final response.
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2022-718', Anonymous Referee #1, 23 Aug 2022
This manuscript provides an introduction to the modern Mars version of the Met Office United Model (UM). The paper describes the foundational adjustments that were made to the model to adapt it to modern Mars and then evaluates its performance relative to the LMD Mars Planetary Climate Model (PCM) and compares UM simulations with (RA) and without (RI) radiatively active dust. The model is still in development and currently lacks CO2 and H2O cycles. The model broadly reproduces the martian dust cycle and annual thermal and dynamical conditions. It will be useful to the scientific community to have another Mars GCM, but with terrestrial-derived physical parameterizations to help better understand the physics of the martian climate. I recommend the manuscript for publication after a moderate revision.
General comments:
- Substantially more discussion is needed regarding several aspects of the modeled dust particles and dust cycle in general.
- First, only a single sentence is used to describe the dust optical properties used. Based on the reference, I assume this implies that a terrestrial dust composition and optical properties are used. But that reference also refers to other references with those details. The authors should summarize the relevant information in the text and also compare and contrast that with known properties of martian dust (chemistry, optical properties, albedo, etc.). Additionally, discuss how these differences could impact the simulated climate.
- Second, how is dust lifted in the model? It seems (based on a few sentences throughout the text) that the model only lifts dust through wind stress-driven saltation. Is that correct? Observations have shown that dust devils likely supply ~50% of dust into the atmosphere, so it should be mentioned explicitly if there is no parameterization for dust devils. I have some additional comments below that relate to this.
- Given the paper is focused on the dust cycle and its impact on the climate, I think another figure or two on the modeled dust cycle would be quite helpful. Specifically, a figure showing the globally averaged aerosol optical depth as a function of season for RA and RI simulations, perhaps with the PCM overplotted, and a second figure identifying model grid points favored for dust lifting and deposition. See Gebhardt et al. (2019) for similar figures: https://agupubs.onlinelibrary.wiley.com/doi/full/10.1029/2019JE006253. I think such figures could replace Figures 1 and 2 if desired.
- I don’t think Figures 6-9 are structured in the most effective manner. The difference plot in the center is intuitively for the differences between the left and right columns, but that isn’t the case here. And differences between the RA simulation and the PCM are discussed in the text, which leaves the reader having to mentally try and create such a difference plot. I don’t think the authors need to double the number of plots necessarily (one set comparing RA and RI, another comparing RA and PCM), but I do think the visual comparison between RA and RI and RA and PCM should be separated.
Specific comments:
Line 98: Are spectral files for the ancient Sun (3.8gya) being used here?? Or is that just a typo?
Table 1: Maybe change the title from ‘orbital parameters’ to ‘physical constants’ or something else as the gas constant and scale height are mentioned? Additionally, is the scale height prescribed as a constant? How does that work? That value seems high relative to Mars (~11 km).
Section 2.3: There’s no mention of surface albedo or thermal inertia here or elsewhere. Are martian values used as boundary conditions? Something else?
Line 155: Is an effective dust radius of 3um assumed as it was in Madeleine et al. (2011)? Some other value? Specify it here.
Line 167: Does it really take 40 years to reach steady state? Why so long, especially without H2O or CO2 cycles?
Line 221: “meaned” reads oddly to me. “Averaged”?
Line 237: should “to” be “in” here?
Lines 239-240: it looks like 1-2 m/s and 1 K differences to my eyes?
Line 251: “in each month” should be “in this month”?
Line 252: This difference perhaps could be somewhat ameliorated by tuning the dust cycle, however? Was any tuning done to the dust cycle? This goes back to general comment #1. Lifting efficiency factors, etc., are typical tuning parameters for Mars GCMs.
Line 256: “reverse” should be “reverses”
Figure 7 caption: specify that negative values indicate southward winds.
Section 4.2. titles: include Ls range with each month title (e.g., “Month 9: Ls 270-300°”).
Line 314: “RI” here should be “RA”
Figure 9: I appreciate the reason to use different color bar scales for RA and PCM here, but it’s still quite confusing by eye when the RA plots have deeper red colors. Can the color bars be adjusted to help the reader visually intuit that RA often has less dust? Additionally, percent difference (as is used in the text) might be more intuitive for the difference plot in the center in this case.
Section 4.2.3 and Figure 9: One consistent difference (at least for months 6-12) between RA and RI seems to be that RA lifts dust farther south than RI? Is that correct? Is that a feedback effect between radiatively active dust and the Hadley circulation perhaps? This may be worth some discussion in the text.
Line 322: I’m confused by the sentence beginning with “The high vertical uplifting…”. This sentence makes it sound like the UM has a “rocket dust storm” parameterization, which I don’t think is true. So what is really meant by “high vertical uplifting”? What physical mechanism is bringing higher dust mixing ratios to high heights in UM-RA? Months 6 and 9 seem to natively produce a high altitude dust layer (Heavens et al., 2011; Guzewich et al., 2013), which I think would be quite novel for a free-running dust simulation! More discussion is needed here.
Section 4.2.3: calling back to my general comment #2, does dust opacity peak in month 9 in UM-RA and UM-RI? That is the typical pattern for free-running dust simulations in Mars GCMs, but the real Mars often has a “solstitical pause” in activity near southern summer solstice outside of years with solstitical global dust storms. I imagine PCM reflects this since it uses the Montabone dust climatology.
Line 347: The sentence beginning with “Dust abundance is higher…” is essentially repeated in the next paragraph. Similarly, saying a simulation is “higher” by a negative number is confusing.
Line 378: “Once again there is a difference…” of zero?
Line 380: Missing “Fig.”
Line 381: specify this difference is only in RA and RI, not in PCM simulations
Section 4.3.3: Dust devils aren’t mentioned at all! The most likely reason for this disparity, in my opinion, is that UM doesn’t have a dust devil parameterization. Wind stress lifting is very low during the early portion of the year, while real Mars (and hence the PCM) maintains a background dust haze through dust devils. Kahre et al. (2017) and references therein discuss this in detail.
Line 471: typo “whata”
Citation: https://doi.org/10.5194/egusphere-2022-718-RC1 - Substantially more discussion is needed regarding several aspects of the modeled dust particles and dust cycle in general.
-
RC2: 'Comment on egusphere-2022-718', Anonymous Referee #2, 14 Sep 2022
This manuscript documents two simulations from a new Mars global circulation model, ported from the UK Met Unified Model. The addition of new, independent global models is welcome in the community and should always be encouraged. Its development will increase the capabilities of the modeling community. In advocating for the development of a new Mars model, the manuscript also makes a strong case for the need for a Mars Atmosphere Inter-comparison Project. CMIP has been extremely successful for the assessment of future climates for Earth, and a MAIP might hold similar promise for Mars.
Generally, the model proves capable in simulating many of the large-scale features of the Martian climate. The model reproduces a dust cycle, which bears a reasonably good resemblance to that of Mars, without the need for forcing the simulation to observations. This is an impressive advancement. The manuscript is well-written and organized; however, a couple of discussion points are neglected, and most importantly, a key process is not included in the model. See main comments below and annotated PDF for minor comments.
Major comments:
I.) Most critical of the major comments: I understand that porting a terrestrial model to Mars is a substantial undertaking, but a Mars model lacking the CO2 cycle—and therefore having surfaces pressure being up to 20% too large for a given grid cell—seems like a massive (literally and figuratively) potential source of error. The authors recognize the need for a CO2 cycle on lines 521–531. A non-exhaustive list of the potential problems in simulating a realistic Martian climate without the CO2 cycle include: 1.) incorrect tidal amplitudes, since a given radiative forcing will, for a large part of the year according to Fig. 5, be working on more mass than actually exists. Because the tides are so important for closure of, for example, momentum budgets, getting this wrong makes the entirety of the presented results less robust. 2.) The reality of the radiative influence of dust may also suffer. If X amount of dust is lifted, the mixing ratio of X amount of dust would be less than reality if there was an incorrectly excessive amount of non-dust mass in the atmosphere. This comparatively thinner dust layer may not necessarily change the surface temperature because the total extinction would be nearly the same, but it might change the vertical temperature profile based on changed vertical distribution of dust. 3.) Similar arguments could be made on the impact spurious mass might have on various wave modes, but without running experiments, it is hard to know the non-linear changes on a model never adapted for Mars before, which is the point. 4.) Finally, the CO2 cycle is associated with a flow of the atmosphere from one pole to the other as the polar caps sublimate and deposit. This flow, while comparatively smaller the zonal winds of the Hadley cell itself, is still missing, and with it one of the important potential mechanisms for dust lifting in the high latitudes.While overall, the simulated climate looks reasonable; as plotted in sigma coordinates, the effect of an incorrect surface pressure is obviated. I am left to wonder if the model would look even better if this neglected process were included; conversely, I am concerned that not including this process is hiding other issues that are not yet apparent. I would strongly encourage the authors to investigate the feasibility of incorporating the CO2 cycle in the present version rather than saving for a future manuscript. At least, a simplified parameterization as noted used by other models around line 525, could be attempted. The process by forcing the atmospheric mass to a prescribed surface pressure or enforcing mass sources/sinks as need in the poles at the appropriate times of year might be sufficient.
II.) I missed a discussion on dust lifting. How is the process parameterized? How is the surface dust reservoir calculated? There is a brief mention in the results section on Line 320 that the UM calculates reservoirs and the horizontal motion, but this warrants a more complete description in the methods. Only on line 449, the fact that the dust reservoir is infinite is finally mentioned. This all needs to be organized into a specific section in the methods. It is impressive, as noted around Line 340, that a dust cycle is reproduced in the model without forcing, but it is difficult to assess how robust the cycle is without knowing how lifting is parameterized. One particular detail that appears contrary to the observed dust cycle is that month "9" of Ls~270 should have reduced dust MMR than months 6 or 12 (Montabone et al., 2015), but that is not the case.
III.) Is there any sensitivity to model resolution (predominantly horizontal but vertical as well)?
IV.) The results and discussion focused on the zonal-mean structure. This is an excellent way to put the bulk climate into context but is not the full picture. As the goal is a comparison of the UM MCGM to the PCM, at least some investigation of the non-zonal mean structure is warranted. The addition of at least a few plots showing the column optical depth for each season on a lat/lon figure would show that dust is being transported within the circulation in a realistic way. Similarly, plots of the surface temperature and pressure would demonstrate how poorly or how well the model manages to capture the true climate without a CO2 cycle (Major comment I).
-
AC1: 'Comment on egusphere-2022-718', Danny McCulloch, 26 Nov 2022
We thank the reviewers for their time and valuable comments. We have amended/adjusted the manuscript, where alterations can be found in green text and old text features a strike-through. A detailed response with the original comment and specific replies is provided as a supplement which features a description of the added content.
We also thank the editor for their understanding in extending the deadline for the final response.
Peer review completion
Journal article(s) based on this preprint
Data sets
UM datasets and scripts for plot reproduction Danny McCulloch https://doi.org/10.5281/zenodo.6974260
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 551 | 144 | 10 | 705 | 6 | 2 |
- HTML: 551
- PDF: 144
- XML: 10
- Total: 705
- BibTeX: 6
- EndNote: 2
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
Denis Sergeev
Nathan Mayne
Matthew Bate
James Manners
Ian Boutle
Benjamin Drummond
Kristzian Kohary
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(13863 KB) - Metadata XML