the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 23 May 2023
Sensitivity analysis of a Martian atmospheric column model with data from Mars Science Laboratory
Abstract. An extensive sensitivity analysis was performed for a horizontally homogeneous and hydrostatic 1-D column model at the Mars Science Laboratory (MSL) location. Model experiments were compared with observations from the Curiosity Rover Environmental Monitoring Station humidity (REMS-H) device. Based on our earlier column model investigations, model surface temperature and pressure, dust optical depth (τ) and column precipitable water content (PWC) were the initial parameters that we investigated by our sensitivity analysis. Our analysis suggests that the most sensitive initial parameters for the column model temperature profile are τ and the surface temperature. The initial value of PWC does not affect the temperature profile of the model, but it is the most important parameter for the humidity profile. The initial value of τ also seems to have some effect on the humidity profile of the model. Based on our analysis, variations in surface pressure initialization are negligible for the model’s humidity and temperature predictions. The model simulations are generally in good agreement with the observations. Our analysis suggest that a slightly different shape of the model’s initial humidity profile could yield better results in the predicted water vapor volume mixing ratios at 1.6 m.
- Preprint
(9913 KB) - Metadata XML
- BibTeX
- EndNote
Status: final response (author comments only)
-
CC1: 'Comment on egusphere-2023-846', Franck Montmessin, 10 Aug 2023
This article presents a sensitivity study conducted with a 1D-model developed to simulate a column of atmosphere on Mars. The model includes a variety of physical processes intended to represent those to which the column is submitted at various timescales (convection, radiation, exchanges with the regolith).
The main ambition is to address the impact that some parameters have in the predictions of the model; namely temperature, relative humidity and water vapor mixing ratio. As this model has already been applied to interpret data produced by atmospheric sensors on board the Mars Science Laboratory rover (Curiosity), this work should enable the model to be used more effectively in the future, and its main limitations to be better understood.
While this paper represents a solid and valuable effort to explore the behavior of a model used to interpret Curiosity's atmospheric and surface data, it does not answer any particular scientific question and will primarily serve as a reference for the future use of the 1D model.
For this reason, the scientific contribution of the manuscript seems rather weak, while its technical value is beyond doubt.
This consideration aside, the article is concise and well-structured, but suffers from several flaws which are listed below:
-It is not clear how the conclusions drawn from this study will impact on the future use of the 1D model. Some conclusions could have been avoided, as they merely confirm things already known and presented in the introduction (the negligible role of sensible and latent heat on surface temperature), while others could have been used to extend the study to arguably more representative cases (MCD). The question that should be addressed is how the findings will change the strategy for the interpretation of MSL data.
-The role of regolith has long been an open question in the Mars water cycle community, since several Martian climate models have successfully reproduced the main features of the Mars water cycle in the absence of regolith. It is understood that the 1D model used here is based on the assumption that regolith plays an active and important role in the concentration of water vapor near the surface, which should deserve some more justification, especially in the context of contradicting results from 3D climate. Another unquestioned phenomenon concerns condensation and the formation of fogs. This is not mentioned in the manuscript, something that should be clarified by the authors. In particular, it would be interesting whether there is a competition between adsorption and condensation in the early morning
-half of the graphs show a comparison between various model results as a function of altitude. Yet they should only emphasize the altitude at which the measurements are made (1.6 m) and not show T and VMR profiles up to 5 km while most of the diurnal variations occur in the first hundreds of meters .
Specific comments (numbers refer to line numbers in the text):
26: 1) one of its unique features, compared to Earth, is also its 95% composition.
28: 2) sensible heat is negligible for the surface, but not for the atmosphere (matters for the BL)
87: “and average of the T” remove of
129+: PWC should be expressed in precipitable microns, pr-um.
141: the few ChemCam observations could have been expanded by many more data from orbiters
Fig4: 1) is condensation included in the model?
2) limit altitude axis to below 250 m
3) since the text emphasizes the lack of reliability of VMR for Low RH data; RH plots should be added to let the readers see when VMR should be ignored. Alternatively, the authors could mark points when unreliable.
202: nighttime H2O VMR for Ls 271° after dusk is not well reproduced at all. Any comment?
206: It should be stressed that Chemcam cannot directly measure the H2O VMR at 1.6 m, and its value is essentially an extrapolation based on strong assumptions that the MCD, for instance, could contradict. It is clear from the graphs that H2O vmr deduced from RH measurements are made in a layer marked by a strong yet very shallow gradient. Chemcam has no sensitivity to that region of the atmosphere.
216: Since it was already discussed in previous works using the same model, why is it the MCD has not been employed to initialize the moisture profile?
222: “initial” implies these parameters can evolve during the run. “Fixed” parameters see more appropriate
Citation: https://doi.org/10.5194/egusphere-2023-846-CC1 - AC3: 'Reply on CC1', Joonas Leino, 12 Mar 2024
-
RC1: 'Comment on egusphere-2023-846', Anonymous Referee #1, 13 Feb 2024
General Comments
The authors report experiments with some one-dimensional atmospheric column model, at the site of the Mars rover Curiosity. That includes model experiments in the warm season around solar longitude 271 and cold season around solar longitude 90. They use Curiosity data on the near-surface temperature, near-surface pressure, atmospheric dust, and atmospheric precipitable water content as initialization condition of the model. The authors study how sensitively their model results depend on these initialization conditions. This is done by increasing and decreasing the model initialization condition by certain values, re-running the model, and inter-comparing model data. Also, the authors use Curiosity data for evaluating their model experiments. The latter is done by comparisons between Curiosity data and model data.
Major revision of the manuscript is needed before publication. There are the following major comments.
Line 152-154: “On top of that, the VMR cycles (Figs. g and h) include the ChemCam-derived VMR (marked by x) estimated from the PWC assuming a well-mixed moisture profile (McConnochie et al., 2018)”. Is the ChemCam-derived VMR in Figs. 4-7 (g) and (h) identical or similar to the initialization condition of the model moisture profile (Lines 109-113)? If yes, agreement between your model and ChemCam in Figs. 4-7 (g) and (h) may be trivial. And, it may not provide any new information. Please clarify (or remove the comparison with ChemCam VMR in Figs. 4-7 (g) and (h)).
Line 150-152: “Modeled cycles of diurnal temperature (Figs. c and d) […] at 1.6 m include model runs with […] together with REMS-H values (black spheres).” Are the REMS-H temperature values in Figs. 4-7 (c) and (d) similar to the initialization condition of the model surface temperature (Line 102-103 and Line 129)? If yes, agreement between your model and REMS-H in Figs. 4-7 (c) and (d) may be trivial. And, it may not provide any new information. Please clarify (or remove the comparison with REMS-H in Figs. 4-7 (c) and (d)).
Line 9-10, lines 206-210 and lines 216-217: “Our analysis suggest that a slightly different shape of the model’s initial humidity profile could yield better results”, “… the model daytime humidity should be increased at low altitudes …”, “… good agreement with the experiments made by Savijärvi et al. (2019a), as initially "low-moist layer…"”, “The moisture profile from the MCD (Savijärvi et al., 2019a, Fig. 8) suggests … . Such an initialization of the moisture profile could work here as well … .” Yes, you made some interesting suggestions. Please make these suggestions happen. A revised version of the manuscript must include the following
- low-moist-layer model experiments, following Savijärvi et al. (2019a) or similar
- model initialization with the MCD moisture profile.
The authors have certainly the capacity to do that, as Dr. H Savijärvi is co-author of the manuscript.
That is a major comment. More model experiments and some basic rewriting and/or extension of the article are needed.Specific Comments
Line 2-3: “Model experiments were compared with observations from the Curiosity Rover Environmental Monitoring Station humidity (REMS-H) device.” You could add the comparisons with ChemCam (unless removing them, following the major comment on line 152-154).
Line 15: “testing new numerical algorithms”. Please mention briefly what numerical algorithms you mean here.
Line 38-39: “The REMS-H device humidity measurements will be re-evaluated, which will modify the calibration coefficients. Thus, the humidity values will change somewhat, but they still serve in their current form in the sensitivity analysis performed here.” What is this information based on? Personal communication or is there any reference? Some more clarification would be helpful.
Line 50-51: “we focus on parameters whose sensitivity has not been studied before”. Please provide all such parameters.
Line 58-59 and line 80: “The predicted quantities are horizontal wind components, potential temperature and mass mixing ratios of water vapor and ice.” (line 58-59) and “driven by the predicted ground heat flux” (line 80). Are lines 58-59 and line 80 consistent? Is the ground heat flux also a predicted quantity? Are comparisons between REMS wind data and your model possible (or not)?
Line 67-68: “and the surface transfer coefficients are defined with the same stability functions as above the lowest model layer.” Please explain how stability functions depend on the height. Otherwise, that may not be immediately clear to an external reader.
Line 74: “trace gases are not taken into account”. Please specify exactly what trace gases you mean here.
Line 81: “at eight levels” Are these sub-surface levels? They seem to be different from the model grid points in lines 57-58. Please clarify.
Line 86: “median of the first measurements of RH”. How many of the first measurements of RH? What does that mean for data accuracy?
Line 87: “average”. Are these 5 minute averages? Do you calculate medians, arithmetic means, or something else? Please give some more details.
Line 87-88: “Here we use only the last measurements of P as the stable sensor (LL type) needs long warm-up time”. How many of the last P measurements do you use? What does that mean for data accuracy? Do you calculate the median, arithmetic means, or something else? Please give some more details.
Line 102: “The model’s surface temperature and pressure”. Do you mean here the temperature and pressure exactly at the surface (at zero meters altitude) or at the lowest model level, which is 0.3 meters (as follows from line 58)?
Line 101-103: “The hourly REMS observations, described above, are used to initialize the column model. The model’s surface temperature and pressure are initialized with the sol-averaged values, calculated from the hourly REMS observations of the previous sol.” What REMS measurements are used? Is the REMS ground temperature sensor, air temperature sensor, or REMS-H sensor temperature used for initializing the surface temperature and REMS-P for the surface pressure? See also the below comment on Line 129.
Line 103-104: “lapse rate of 1 K/km”. Provide evidence why a lapse rate of 1K/km is reasonable here. Is that consistent with measured or theoretical lapse rates on Mars?
Line 105-108: Do you use a single value or daily mean value of the Mastcam dust optical depth? Please explain.
Line 109-113: Do you use a single value or daily mean values of ChemCam PWC or not. Please explain.
Line 113: Is the surface pressure ps based on REMS-P? Please make that clear.
Line 129: “for REMS-H mean temperature”. Does that mean the REMS-H sensor temperature is used for initializing surface temperature? If yes, please state that clearly in the manuscript. Explain also why not using ground temperature sensor and air temperature sensor data. See also the above comment on Line 101-103.
Line 132-145: These paragraphs may need some rewriting. Their structure should be more logical. Take for instance the paragraph from line 132-138.
- There are the sentences “As there are lots of data gaps in the measurements, some sols may miss essential observations for determining the sol-averaged T and P. The seasonal pressure 135 cycle is well known at the MSL site, as there are more than 3000 sols of pressure data. Thus, the sol average pressure can be estimated relatively accurately, even from some other Martian year.” That does not make clear what is the reason for varying the model initialization by ± 10 Pa.
- There is the sentence “The reported accuracies of the REMS-P pressure and REMS-H temperature sensors are ± 3.5 Pa and ± 5 K (Martínez et al., 2017).” Maybe, that can be used as a minimum value for varying the model initialization (also, there could be questions such as the following. Are data uncertainties randomly distributed? Is the data uncertainty much smaller if calculating data averages over many REMS measurements?). Maybe, you do not want to use just the minimum value. That is why you eventually select ± 10 Pa and ± 10 K.
Please think about how to rewrite the paragraphs from lines 132-145.Line 149: “local time (LT)”. Please use local true solar time, if not done yet. Change LT to LTST (consistent with Figs. 4-7 (c)-(d) and (g)-(h); their x-axis caption is “LTST (h)”).
Line 148-149: “at 06 (black), 08 (blue), 10 (red) and 12 (orange) local time (LT)”. Please explain why selecting these times.
Line 149: “up to 5 km”. Please explain why selecting 5 km as the upper limit.
Line 172-182: High dust seems to give higher near-surface temperatures at night and cooler near-surface temperatures during the day. Is that correct? If yes, that may be consistent with the effects on the near-surface-temperature, known from dust storms. Any consistency with dust storms may be pointed out in the paragraph from lines 172-182 (if any). And, the paragraph may be rewritten, accordingly.
Line 201-202: “The nighttime VMR derived from the REMS-H, in Figs. 4g and 4h, is relatively close to the model simulation in both seasons”. There seems to be some dis-agreement in the first half of the night, around 18-24 LTST, in Figs. 4g and 4h. More explanation is needed.
Technical Corrections
Line10: Change “our analysis suggest” to “our analysis suggests”?
Line 52: Change “summarized and discussed” to “discussed and summarized”?
Line 55: Does the model have a name? That is just to make sure. If not, it is alright.
Line 76: “The long-wave radiation scheme is described using a fast broadband emissivity approach.” Does it mean that “The long-wave radiation scheme uses a fast broadband emissivity approach”?
Line 81: Change “Savijärvi et al. (2016, 2019a, b, 2020); Savijärvi and Harri (2021)” to “Savijärvi et al. (2016, 2019a, b, 2020) and Savijärvi and Harri (2021)”?
Line 91: “The REMS-H VMR values are most accurate at minimum VMR, which usually occurs during the night at maximum RH.” REMS-H measures RH, not vmr. Right? Does it mean the following? REMS-H is most accurate at maximum RH. The maximum RH occurs at night and thus may coincide with minimum vmr. That may be misunderstandable. Please rephrase. Also, you could provide some more explanation for external readers, on why maximum RH and minimum vmr occur at night.
Line 92-93: “Thus, Figure 2 shows the REMS-H maximum RH (black) and derived VMR (purple) at the same time of sol during Martian year (MY) 32”. You take the daily maximum of RH. Then, you convert the daily maximum of RH into VMR. Right? If so, it is self-explaining that the daily maximum of RH and its derived vmr are at the same time of sol. But, they do not occur at the same time on any sol. Right? That may be misunderstandable. Some rewriting may be needed.
Line 94-95: “the warm perihelion period is at around Ls 220°–280°”. The red curve in Fig. 2 seems to have some dip from LS220-280.
Line 98-99: “reach a minimum around the southern hemisphere winter solstice.” Please add the related solar longitude (Ls 90°).
Line 105: “(Lemmon, 2014)”. Another very recent publication may be relevant here
M.T. Lemmon, S.D. Guzewich, J.M. Battalio, M.C. Malin, A. Vicente-Retortillo, M.-P. Zorzano, J. Martín-Torres, R. Sullivan, J.N. Maki, M.D. Smith, J.F. Bell, The Mars Science Laboratory record of optical depth measurements via solar imaging, Icarus, Volume 408, 2024.
https://doi.org/10.1016/j.icarus.2023.115821.That is just to let you know.
Line 127: “The diurnal surface pressure cycle is not predicted in the model”. External readers could have the following questions. What does that mean exactly? Why does the model need surface pressure initialization then? Please make that clear.
Line 144: Change “cf. 3” to “cf. Fig 3”.
Line 151: Change “VMR” to “water vapor VMR”?
Caption of Fig. 4: Change “default (-)” to “default (continuous line)” or similar?
Caption of Fig. 4: Change “VMR” to “water vapor VMR”?
Caption of Fig. 4: Change “local time” to “local true solar time”?
Figure 4-7: Change “MSL” to “REMS-H” in the legend of sub-figures and the caption of Fig. 4?
Line 159: External readers may need some help for seeing the temperature inversion in Figs. 4(a) and (b) (temperature increases with altitude, close to the surface, …). Please add some more details.
Line 159-160: “while at 12 LT it is no longer present”. The inversion is already not present at 10 LTST. Right? If right, please rephrase.
Line 161: “At 08 LT (blue line) convection has already started as solar radiation has started to strongly heat the surface of Mars.” External readers may need some help. Make clear that can be seen from the lower end of the blue curve in Figs. 4-7 (a) and (b) (temperature has changed from increasing with altitude to decreasing with altitude). A close look is needed.
Line 162: Change “On top of the stronger convection in the warm season” to “In addition to the stronger convection in the warm season”?
Line 183-184: “The humidity profiles of both seasons (e.g. Figs. 4e and 4f) display a well-mixed layer in the boundary layer (BL). At 06–08 LT, the well-mixed layer is very shallow and grows thereafter due to strong convection in both seasons.” At 10 LTST (red curve), there seems to be a shallow well-mixed layer from ca. 100-500 meters in Fig. 4e) and 100-800 meters in Fig. 4f). That can be seen from the water vapor mass mixing ratio not changing with altitude. A similar feature is not obvious for 6 and 8 LTST (black and blue curve). Please clarify.
Line 188-189: “Increased solar radiation near the surface in the morning”. Increased solar radiation near the surface means model initialization with less dust. Right? Please say that clearly.
Line 183-192: Please do not move back and forth between Figs. 4 and 5 in this paragraph.
Line 195: “which is at least partly due to the fact that they are a function of temperature.” Some more explanation is needed. Does it mean that adsorption is a function of temperature? How does it change with increasing surface temperature?
Figure 4-7: Why do you use mass mixing ratio in Figs. 4-7 e) and f) and volume mixing ratio Figs. 4-7 g) and h)?
Figure 4-7: Why do the model data have some gap from 0-1 LTST in Figs. 4-7 (c)-(d) and (g)-(h)?
Line 205: Change “marked by x” to “marked by x in Figs. 4-7 (g) and (h)”?
Line 208: Change “sphere” to “(sphere, Fig. 5g)”?
Line 209: “as initially "low-moist layer" in the model increased 1.6 m VMR values”. Make clear that humidity values were increased at low altitude relative to the well-mixed model experiment in Savijärvi et al. (2019a). That may not be immediately clear to an external reader.
Citation: https://doi.org/10.5194/egusphere-2023-846-RC1 - AC1: 'Reply on RC1', Joonas Leino, 12 Mar 2024
-
RC2: 'Comment on egusphere-2023-846', Anonymous Referee #2, 11 Mar 2024
This paper reports a number of sensitivity analyses conducted using a one-dimensional atmospheric column model and comparisons with data acquired by REMS: pressure, near-surface temperature, and VMR at 1.6 m. The authors investigated the impact of dust optical depth, precipitable water content (PWC), surface temperature, and surface pressure values on the model results by varying these parameters in different ranges chosen based on observations. The comparison is conducted for two Ls values: 90° and 271°. Although the manuscript is generally well-written and provides new and valuable information for column model simulations on Mars, further analyses and comparison (listed below) are needed before publication.
Major comments:
-Vertical profiles of Figures 4-7: In the discussion, the authors refer to altitudes that are not shown (e.g., lines 175-182) and that are key to understanding the performance of the model. Additionally, as the comparison is with near-surface data, I would like to see the model results near the surface. If the authors want to keep the sensitivity analysis up to an altitude of 5km, then additional figures focused on the 0-1000 m range should be added as it is hard to distinguish the different curves in that range.
-The manuscripts states that one of the most sensitive initial parameters for the column model temperature profile are the dust opacity and surface temperature. Here, I would like to see a comparison with the MCD. Also, what would be the effect if part of the aerosol opacity is due to water ice? For Ls=90° simulations, a notable % of the total opacity should be ice whose single scattering albedo is close to 1. Would it be possible to add in the model a diurnal cycle of the aerosol opacity?
-Conclusions section: ‘An earlier study by Savijärvi et al. (2019a), large-scale model moisture profile from the MCD (Fig. 8 in Savijärvi et al. (2019a)) and our sensitivity experiments (Figs. 5g and 5h) suggest that the model’s initial humidity profile at the MSL site should vary with the season to provide a better moisture prediction near the surface.’. I think the authors should address this in this study. Why not taking the MCD profiles and see if the simulations improve with those model profiles? I don’t think that “…the model’s initial humidity profile at the MSL site should vary with the season to provide a better moisture prediction near the surface..” is demonstrated in this work, and it is not clear what this study contributes beyond the cited work. This point is also mentioned at the end of the abstract but again no demonstration of how the humidity profile can affect is given. Please use MCD profiles and see of results change.
-Conclusions section: The authors basically summarize the findings in the sensitivity analysis (some previously reported in previous works) but they discuss vaguely the reasons behind and do not detail the impact of their results. For instance, page 14: “We found that the initial value of surface temperature affects the entire temperature profile with a slightly larger effect at Ls 90.”. Why is that? Also, is this the case at all altitudes? It is complicated to say below 1000 m from the figures.
-I believe the manuscript would benefit from the addition of more data from other Ls in the comparison.General comments:
-Why the authors are not included in the comparison data from MEDA??
-Include the errors in the observations, as otherwise, it is hard to figure out how well the model reproduces the data.
-It is confusing to use in the paper terms like ‘profile initialized…’ for parameters that do not change during the run. For the model parameters that do not change during run, please just use ‘fixed profile…’ or ‘fix values of …’
-Section 2.2: please add information about the sampling when describing the REMS data.Citation: https://doi.org/10.5194/egusphere-2023-846-RC2 - AC2: 'Reply on RC2', Joonas Leino, 12 Mar 2024
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 347 | 87 | 32 | 466 | 23 | 17 |
- HTML: 347
- PDF: 87
- XML: 32
- Total: 466
- BibTeX: 23
- EndNote: 17
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1