Initial Drift Correction and Spectral Calibration of MarSCoDe Laser-Induced Breakdown Spectroscopy on the Zhurong Rover

Round 1

Reviewer 1 Report

Authors utilized global iterative registration (MGR) procedure to compensate thermal drift of the spectrometer for LIBS. They compared this MGR procedure with their previous approach based on particle swarm optimization algorithm.

The manuscript novelty and impact fit the Remote Sensing. However, author missed the Stark shift effect which is rather important for LIBS due to high electron density of the plasma. It is rather useless to compare modeled LIBS spectrum with that acquired in the experiment if Stark shift is missed. Authors have to discuss it before the manuscript can be accepted.

Major revision.

 

Questions:

1.     Lines 91-92. “ … mixtures of IlmHem and ClinQzOrth as the calibration samples.” Please, rephrase the sentence and do not use rather specific names of the samples utilized in Ref. 11. It is better to describe as the mixture of minerals.

2.     Lines 244-245. Equation (1) problem. It seems authors missed a square root for RMSE. RMSE = (1/n Sum(lambda_1i-labda_2i))^0.5 I hope it is a typo.

3.     Figure 1 discussion. Lines 277-290. According to equation (1) at line 244-245 the RMSE should have units of nm (please, check my question 2 above). However, in fig.1 according to discussion the RMSE is expressed pixel units. Please, clarify it in the figure caption and discussion.

4.     Fig. 2 capture. What is the reference spectrum? Authors discussed that they utilized NIST’s ASD LIBS database for spectra modeling but this should be explained in the figure’s capture in my opinion.

5.     Fig. 3. What are RMSE units? Wavelength in nm or in pixels? Please, correct the title of the y-axis.

6.     The main question. According to the manuscript authors did not account on Stark shift in laser plasma when compared the measured spectrum lines positions with the NIST database values. Depending on chosen atomic/ionic line the Stark shift can be 0.1-0.4 nm for the electron density of 1*e^16 to 20*e^16 [Bengoechea, J., C. Aragón, and J. A. Aguilera. "Asymmetric Stark broadening of the Fe I 538.34 nm emission line in a laser induced plasma." Spectrochimica Acta Part B: Atomic Spectroscopy 60.7-8 (2005): 897-904]. This in an important issue for this manuscript.

LIBS plasma is known to be a nonuniform, so different plasma regions will provide different Stark shift [Aragón, Carlos, and Jose Antonio Aguilera. "Characterization of laser induced plasmas by optical emission spectroscopy: A review of experiments and methods." Spectrochimica Acta Part B: Atomic Spectroscopy 63.9 (2008): 893-916]. Moreover, time is also crucial for spectra modeling since local thermodynamic equilibrium (LTE) supposition is only valid in small plasma volume (fraction of a 1 mm3) at short time period (less than few microseconds). Careful Stark shift estimation needs LTE conditions so LTE is important feature to calculate/estimate anything in the spectrum. However, authors utilize non-gated detector so the plasma temperature and electron density will be averaged. Consequently, reliable data on plasma temperature and electron density is challenging if possible to achieve with non-gated detection. Finally, LTE is typically is not fulfilled at low pressures (not enough collisions of electrons with atoms/ions in the expanding plasma at pressures below 50 torr) [Aragón, Carlos, and Jose Antonio Aguilera. "Characterization of laser induced plasmas by optical emission spectroscopy: A review of experiments and methods." Spectrochimica Acta Part B: Atomic Spectroscopy 63.9 (2008): 893-916].

Summarizing my major question: Stark shift in plasma spectrum is a very important issue for the manuscript but authors did not discuss it. They just missed it. They have to provide discussion and estimation of the Stark shift impact before the manuscript will be accepted.

7.     Line 167. “The default values of electron temperature and electron density are 1 eV and 10^17 cm-3.” Why author chose these conditions? Please, estimate plasma temperature and electron density from your plasma spectra so this choice can be validated. Moreover, the plasma temperature and electron density will be rather different for samples of different origin and it is not convenient to use the same temperature and electron density for modeling spectra of different samples.

Author Response

Dear Reviewer,

We have revised all your comments point by point.
Please see the attachment for details.

Kind regards,

Liangchen Jia

Author Response File: Author Response.pdf

Reviewer 2 Report

The manuscript proposes approaches of spectral drift correction to perform accurate qualitative and quantitative analysis the Martian surface rocks and minerals by MarSCoDe instrument installed on Zhurong rover. The significant changes of Mars temperature and atmosphere can impact the obtained spectra. The authors consider two methods to correct the spectra in two different cases of calibration failure. In my opinion, the authors have carried out a good systematic study. They have shown the effectiveness of the applied correction methods. The subject of the publication is well suited to the Special issue theme “Advances in Exploring the Moon, Mars, and Asteroids Using Spacecraft Remote Sensing and Other Toolkits”. This work may be accepted for publication without further revision.

 

1.                What is the main question addressed by the research?

The main question this research addresses is the search and choice of appropriate approaches for establishing the relationship between the detector pixel numbers and spectral wavelengths to correct the drift of the spectrometer influenced by unique conditions associated with the extraterrestrial scientific mission. Although the purpose of this manuscript seems to be quite specific, it may be important for subsequent similar missions to study extraterrestrial objects.

 

2. Do you consider the topic original or relevant in the field? Does it

address a specific gap in the field?

The topic of this manuscript seems to me relevant in the case of the functioning of the MarSCoDe LIBS instrument carried by Zhurong rover. It is worth admitting that the solution of this problem is quite specific and is excessive in Earth conditions. This research is relevant and can be applied to various extraterrestrial rovers/instruments that do not use radioisotope thermoelectric generator energy to stabilize the spectrometer. In my opinion, this corresponds to the topic of the special issue.

 

3. What does it add to the subject area compared with other published

material?

This manuscript compares approaches for initial spectral drift correction under the changes of temperature and atmosphere. The authors consider spectra obtained under various environmental conditions on the Martian surface. One of the approaches for drift correction is proposed for the first time.

 

4. What specific improvements should the authors consider regarding the methodology? What further controls should be considered?

The authors state that Martian conditions also affect spectral lines intensities and other parameters, which can complicate the drift correction (“the relative intensity and number of characteristic lines also change, which makes it more difficult to correct the drift of on-board data”, lines 219-221). Maybe in the future the authors will be able to correct not only the wavelengths of the spectrometer, but also the intensities of the spectral lines obtained from different distances and under different environmental conditions in the study of Mars.. Previously, Melikechi, N., et al proposed an approach for distance correction of line intensities (Melikechi, N., et al. Spectrochimica Acta Part B: Atomic Spectroscopy 96 (2014): 51-60). Can this approach be used to reduce the number of iterations until the optimal spectral matching?

 

5. Are the conclusions consistent with the evidence and arguments presented and do they address the main question posed?

Yes, the authors have shown that the drift correction techniques are robust. The authors conclude which approach is better to use in different cases. The authors also quantify the effectiveness of drift correction in terms of RSME.

 

6. Are the references appropriate?

Yes, the references are appropriate, but in my opinion their quantity is rather small.

 

7. Please include any additional comments on the tables and figures.

In my opinion, the spectra in Figures 4 and 5 are difficult to perceive due to the large spectral range. Perhaps the authors should present more detailed regions of the spectrum before and after the correction for the three channels.

 

Also I wonder why the authors give the position of the spectral lines in Tables A1-A2 with large decimal fraction (up to 0.0001 nm) if  the spectrometer resolution varies at level of 0.01 nm.

Author Response

Dear Reviewer,

We have revised all your comments point by point.
Please see the attachment for details.

Kind regards,

Liangchen Jia

Author Response File: Author Response.pdf

Reviewer 3 Report

I suggest that this paper need major revision as the follow comments.

 

Q1. Is the proposed MGR correction approach same as the published Level 2B data product? What are their differences?

Q2. When standard spectra are matched to the NIST database, how to ensure that the matching standard wavelength is accurate? 

Q3. ChemCam LIBS data have also been drift corrected. What are the similarities or innovations between the two methods?

Q4. Are there any special conditions for the selection of the standard spectrum? Why is this spectrum chosen? 

Q5. Why use the Ti-alloy spectrum as a standard to correct other sample spectra? 

  

Some minor issues:

1. In page 2, line 64 and 89, in page 6, line 237, in page7, line 289, in page 11, line 398, it is suggested that “shift” be changed to “drift” in order to make the whole text consistent.

2. In page 2, line 78, line 82, line 84, line 87, line 89 and line 92, "et al" is suggested to be changed to “et al.”. 

3. In page 4, Table 1, “Montmo-rillonite” and “Nontron-ite” should be "Montmorillonite" and "Nontronite".

4.     In page 4, line 158, “sample” should be “samples”, which means plurality.

5.     In page 6, line 245, the first letter of “Where” should be small letter.

6.     In page 6, line 245, please add the meaning of “n” in equation 2.

7.     In page 8, Table 3, the note should end with a period. 

8.     In page 8, Table 4, the representation of the a₂ parameter is the same as in Table 2, for example -8.3685e^-07.

9.     In page 9, line 339,  and  should be carried into equation (3), rather than equation (4).

10.   In page 9, line 339, it is suggested that “Table 6 shows the mean error and RMSE” be changed to “Table 6 shows the AME and RMSE” to be consistent with the Table 6.

11.   In page 12, line 408 and line 409, “spectrum” should be “spectra”, which means plurality.

12.   In page 14, line 458, in page 15, line 459, Please explain the meaning of E_i and E_k in the Table A1 and Table A2.

Author Response

Dear Reviewer,

We have revised all your comments point by point.
Please see the attachment for details.

Kind regards,

Liangchen Jia

Author Response File: Author Response.pdf

Round 2

Reviewer 3 Report

The overall content is well revised, But the expression could be further shorten and refined to avoid using long sentences.