Comparison on Quantitative Analysis of Olivine Using MarSCoDe Laser-Induced Breakdown Spectroscopy in a Simulated Martian Atmosphere

Round 1

Reviewer 1 Report

It is very important to test and understand the capability of MarSCoDe LIBS measurement and quantitative analysis. This paper focuses on the procedure of preprocessing and quantitative analysis on the MarSCoDe LIBS with the most commonly univariate linear regression (ULR), multivariate analysis of ordinary MLR, PCR, PLSR, Ridge, LASSO and Elastic Net, and ANN analysis of back-propagation (BP). But the quality of this article still needs to be improved, such as some figures are not clear, contains some mistakes. I suggest that this paper need major revision as the following comments:

Q1. Is the LIBS data preprocessing method same as the procedure of the published Level-2 data product? What are their differences?

Q2. In the paper, how did you do the validation? For example, in the abstract “R2 value on the calibration and validation of all the methods is close to 1.” It does validation with the testing set or validation the true concentration?

Q3. In the Section 2.1, “These can be used for sample classification, composition quantitative, and even 3D characterization.” How to do 3D characterization and how to operate in in-orbit detection?

Q4. In the Section 2.1, “The laser was emitted at a frequency of 3 Hz after an autofocus was performed…measured with an integration time of 1 ms”. Are the parameters set the same as the in-orbit detection in the simulation test?

Q5. Generally, the continuum background removal of LIBS is after the radiation calibration. What is considered in this paper?

Q6. In the Section 2.3.2, how is the blank determined when calculating the limit of detection?

Q7. In the Section 3.2, page 18-19, this paper lacks some experimental discussions on the validation.

Q8. Some figures are not clear enough.

Q9. Can this data and model be used for in-orbit quantitative analysis of olivine?

 

Some minor issues:

1.      In page 2, line 62, Please supplement the date of the observed optical emission.

2.      In page 3, line 147, “laser-to-sample distance of approximately 3 m”. How does the author consider using this distance?

3.      In page 5, line 167. The two olivine words remained consistent in case.

4.      In page 6, line 227, “the LIBS spectra may contain higher levels of continuum signals”. The Martian depression lowers the continuous signal, so this sentence should best be described clearly.

5.      In page 7, line 273, “Two overlap spectra between the adjacent channels … new spectral counts on the given wavelength are computed by a linear interpolation”. “selected with the range of wavelength”, how to determine the range value? And “computed by a linear interpolation”, adjacent channels have different sampling intervals, how do you compute the overlap region?

6.      In page 8, line 338. “??” should coincide with the symbols in equation.

7.      In page 9, line 356. The writing of word “Lasso” should be consistent in this article.

8.      In page 14, line 502. The word “NIST” should be written with full word using the first time.

9.      In page 16, line 574, “with the RMSE value of 0.00 and 1.33, respectively”. Please confirm correct?

10.   In page 17, Table 4 and Table 5. Lack of unit on the values

11.   In page 18, Figure 8. It also needs to add the unit.

12.   In page 18, line 606 and Table 4-5. “PLS” should be “PLSR”. Please check the full article.

Author Response

Special thanks to you for your good comments. We have studied comments carefully and have made correction in the revised manuscript. 

We have studied reviewer’s comments carefully and have made revision which marked up using the “Track Changes” function and marked in red in the paper. We have tried our best to revise our manuscript according to the comments. Attached please find the revised version and response to reviewers' comments, which we would like to submit for your kind consideration.

Thank you and best regards.

Yours sincerely,

Author Response File: Author Response.docx

Reviewer 2 Report

The present manuscript is too lengthy and boring as the authors have mainly focused their attention on Mars Surface Composition Detector (MarSCoDe) mounted on the front deck of the Zhurong rover, a detail of which is already available in the literature. The authors have described too many tasks for preprocessing the LIBS Spectra, like background subtraction, random signal denoising, continuum background removal called baseline correction), spectral drift correction, wavelength calibration, radiation calibration, and multi-channel spectra merging and normalization.And it does not add any new results covenant for its publication. Moreover, to test the capability of MarSCoDe LIBS measurement, the authors comprehensively studied some olivine samples (the primary igneous rock of Mars) with gradient concentrations and experimented in a simulated Martian atmosphere. The manuscript focuses on the procedure of quantitative analysis using the most commonly univariate linear regression (ULR), multivariate analysis. In addition to the information provided in the manuscript, the authors should give a detailed calculation of the LOD using the (multivariate) PLSR method.  The authors built the calibration model using the data of the same sample used for the testing. LIBS data from a sample with a different concentration (measured by other techniques like AAS or IC-MS) is necessary to validate the method. This Manuscript will also help in the study of planetary analogs having olivine found on the earth surface.

In the summary and conclusion section, the first paragraph is not related to the conclusion of the present work; it is related to the plan of the manuscript and available in the literature. Please check the spelling of olivine instead of olive.

 

Therefore, the manuscript may be accepted for publication after considering the above points.

Author Response

Special thanks to you for your good comments. We have studied comments carefully and have made correction in the revised manuscript. 

We have studied reviewer’s comments carefully and have made revision which marked up using the “Track Changes” function and marked in red in the paper. We have tried our best to revise our manuscript according to the comments. Attached please find the revised version and response to reviewers' comments, which we would like to submit for your kind consideration.

Thank you and best regards.

Yours sincerely,

Author Response File: Author Response.docx

Round 2

Reviewer 2 Report

My only concern is related to the detailed calculation of the LOD using the (multivariate) PLSR method, which is not provided in the manuscript. Anyway, in my opinion, the manuscript has been improved and the results collected deserve publication.

Author Response

Dear Reviewers:

 Special thanks to you for your good comments. Those comments are all valuable and very helpful for revising and improving our paper, as well as the important guiding significance to our researches.

Attached please find the revised version and response to reviewers' comments, which we would like to submit for your kind consideration.

Thank you and best regards.

Yours sincerely,

Author Response File: Author Response.docx