Lunar Surface Resource Exploration: Tracing Lithium, 7 Li and Black Ice Using Spectral Libraries and Apollo Mission Samples

Round 1

Reviewer 1 Report

see attached file.

Comments for author File: Comments.pdf

Author Response

I would like to express my gratitude to you for your efforts in improving my work and providing me with the opportunity to publish it. Your anonymous feedback has helped me meet the quality standards required by the journal.

Furthermore, it should be noted that the necessary improvements did not require me to provide additional information beyond what was already available. This made it easy and enjoyable for me to incorporate all of the requested enhancements.

Since 2009, I have been working in Antarctica as the principal investigator of the PALANTARICE project, which focuses on spectroscopy and remote sensing of snow and ice. The new version incorporates detailed information about the ice spectroscopy and the sources of the lapilli used in this research.

Additionally, I have addressed the formal,  linguistic and grammatical issues highlighted by you in the English language of the text. In response to the your comments, we have incorporated all suggestions and corrections into the current version (2) of the document. The modifications made to enhance the manuscript are:

Reviewer #1: “When mentioning specific samples in the text….”

( In line 149_199 and Table 1).

Section 2, Materials and Methods, includes crucial information regarding the samples that aids in comprehending the results and the motivation behind this work. A subsection is dedicated to discussing each sample individually, covering its weight, the mission it was collected on, composition, and most importantly, the exposure age. The exposure age is calculated using mainly 38Ar and 81Kr and is discussed for all Apollo mission samples. In some cases, the absolute age has been included if available. Table 1 has also been updated to include a column for exposure time and altitude from LOLA_LRO

(In line 315_ 321; Figures 6 and 7; line 333-343).

Additionally, the results section discusses the relationship between exposure age, 7Li concentration, and sample topography to determine whether 7Li is cosmogenic or geogenetic. This discussion includes two graphs to provide a new perspective on the importance of the chosen samples as ground truth for our Lithium models

Reviewer #1: “When describing how to convert spectral information into Clementine images, a more detailed explanation of the definition and calculation process of spectral ratios (C1-C5) can be provided….”

The methodology for obtaining the spectral indices (C1_C5) and deriving the 5 types of spectra from the 19 spectra downloaded from RELAB has been extensively detailed.

(In lines 211-229).

1. This includes a description of the principal component analysis (PCA) conducted on the 19 RELAB spectra. The analysis synthesized the 19 spectra into just 5 Type spectral curves, which are presented in Figure 1. These curves represent 5 different substrates, each with a varying average concentration of Lithium and 7Li

(In new Table 2)

2. Additionally, a table is provided with the variables used to calculate each spectral index. The spectral index captures the most extreme reflection characteristics of each Type spectral curve and is composed of a coefficient that multiplies a subtraction of bands, as detailed in Equation 1. The highest wavelength peaks and lowest valleys were selected using a spectroscopy program to manage the spectral curves

(In lines 254_264).

3. The Linear Regression method used in this case was 'BackSteepwise' and employed a statistical criterion (AKAIKE) to determine the smallest number of variables that best fit the data. The explanatory variables of the model were automatically chosen with this criterion. Linear regression is a simple yet informative method that can provide valuable insights into the structure of data and the variables involved. It is important to note that subjective evaluations should be excluded unless clearly marked as such

Reviewer #1: “The "MDT variables" mentioned in the text (line 187) should be clearly defined….”

(In Figure 1 and line 273-285)

We have provided a more detailed description of the SAGA program and included a link to its page. The software is open-source and has significant scientific potential, particularly in the quantitative analysis of topography and geomorphological processes. We have also included a description of the topographic index of wind exposure, its meaning, and a map of the model. The previous version did not include this mapping due to an oversight

Reviewer #1: “Mentioning the setup of an ice spectral library (0, 2, 4, and 6 ppm of very fine….”

(In Figures 3 and 4; lines 294-310)

I specialize in spectral ice libraries. I have experience working with Vis_NIR libraries, as well as Raman and cryo-Raman, for soils, sediments, snow, and ice. In the original version, I did not include all the information related to the ice libraries because I thought I had to explain too many things. However, I have now included this information, which clarifies the origin of the andesitic nature lapilli and the ice libraries. As stated in the article, Vis_NIR libraries were obtained both directly from glaciers and constructed in the laboratory under controlled conditions. In this study, we utilized lapilli from the black glacier of Deception Island and the Johnsons Glacier of Livingston Island. However, we created the spectra in the laboratory to ensure controlled concentrations and temperature during the experiment

Reviewer #1: “Emphasize the lack of "ground truth" under dirty ice conditions for result comparison or validation….”

To fulfill this request, we have done the following:

(In  new Table 3).

1. The spectral ice indices are constructed using the equation 1. We have included a table with the variables involved in formulating the indices. The raster calculator of QGIS was used to extend the indices over the entire Clementine image. The procedure is the same as for the C1_C5 indices

(In Figure 14; lines 554-570) .

2. The results of the dirty ice model have been better described. We have included a figure that shows the mapping of the three indices accompanied by the frequency histogram. It can be seen that both the ICELx1 and ICELx2 indices do not have positive values. Only the ICELx3 index does. The values that the index takes range from 0 to 0.4; therefore, the probability of finding something similar to the Antarctic black glaciers is no more than 40%, but the truth is that there is a chance, and higher spectral and spatial resolution images would be needed to investigate areas of high probability. It is not far-fetched to think that an impact crater created by an asteroid with a significant amount of ice has mixed with the lunar regolith and that this regolith is mixed with these ices. Therefore, we believe that the edges of craters or areas where there is residual material from the impact are preferential areas to look for this type of ice-mixed regolith

Reviewer #1:  “Regarding the mapping of the C5 index involving the ClemVis2 and ClemNir2….”

(In lines 440-455) .

The selection of the highest wavelength of the peaks and the lowest of the valleys, as well as the bands of Clementine that could be related to the wavelengths of the peaks and valleys, was done with a spectroscopy program used to manage spectral curves, Spectroscopy_ProTools 1.2. It is not an author's choice.

I have a spectrum with a series of troughs and peaks, and the program looks for the highest and lowest. I could do this with an Excel spreadsheet and look for the highest and lowest coefficients, but it is better to do it with the appropriate program because it offers alternatives if the wavelength that marks the valley or peak could not be used because that band is not available in your image. In this case I chose the first option. I insist that the Clementine image has many limitations because it is an image obtained with photometric filters, but it still represents global information of incalculable value and for a first screening I think it is appropriate to use it.

However, I believe that by including the table explaining how the C1_C5 spectral indices were calculated and by quoting this table, any doubts about the assignment of the C5 index are removed  

Reviewer #1:  “The discrepancy between the data 0.48 (line 308)…..”

The discrepancy has been corrected. Regarding why lithium correlates better with NIR bands than with visible bands, I believe it is because the origin of lithium is geogenetic, and its concentrations are higher in basaltic rocks, which are fine-grained and contain ferromagnesian minerals; these minerals have a strong response in the thermal infrared bands. I have tried to introduce this discussion.

Reviewer #1: “Standardization of all figures in the manuscript needs to be organized:….”

In the new version, all figures have been reviewed and modified based on these guidelines.

Reviewer #1: “It is recommended to add a column in Table 2 listing…”

To introduce this correction, which I consider very important, I have decided to rename all variables in the Clementine image to their real names; for example, ClemVis1 is now called Band1, Band2...etc., and ClemNir1 is now called Nir1. I hope this makes the document clearer and easier to follow.

Reviewer #1: “Parameters presented in Table 3 should have specific explanations”

The parameters in Table 3, which correspond to the summary of the linear regression, have been explained in more detail.

I hope that these improvements will help to better understand the document.

sincerely

Author Response File: Author Response.pdf

Reviewer 2 Report

Exploring the concentration and distribution of Lithium, Lithium-7 isotope and the possible presence of black dirty ice in the lunar surface is meaningful for future lunar exploration missions as well as helps to understand the lunar geology process. However, the description of the method is too simple, and there is a lack of innovation in methods. Moreover, there are many grammar and spelling errors in the manuscript. Thus, I suggest major revision for this manuscript.

1.        The authors conducted PCA to explore the concentration of Lithium, Li 7. But the detailed method is not very clear. I suggest the authors add some materials such that readers can reproduce the work.

2.        In section 2.2 and 2.3, the authors using existing GIS tools and lunar topographic data (such as LOLA DEM) to investigate the distribution of Lithium and relief patterns as well as dirty ice. But it lacks methodology innovation. It seems that the author only used these existing tools without making any corresponding improvements.

3.        Grammar and spelling errors:

Line 24, “To explore the possibility of find ice” should be “To explore the possibility of finding ice”

Line 17-19, Line 45-47, grammar error or unclear sentences

Line 37, “in this south”, unclear expressions.

Line 43, PSD may be “PDS”?

Line 227, “mmainly” maybe “mainly”

Line 269, “overt” maybe over time.

Line 184, “existence of patters” maybe “existence of patterns”

Line 187, MDT need to provide the full name.

Line 195~196, wrong () for DTMs.

Equation 1, Do peak and valley intensity indicate gray value?

Line 314, Linear Regression Analyses (LRM) is a spelling error.

In figure 1, the numbers in the figures are not clear. Does ‘0,757101‘ in the left top subfigure indicate the gray value of the image?  If so, it is better to use ‘0.757101’.

Figure 7, Lack of horizontal and vertical coordinate axis information

Line 415-417, “China’s missions such as Lunar Reconnaissance Orbiter (LRO) [29], Chandrayaan-1 [68], or Chang’e Program [12].” China did not conduct LRO or Chandrayaan-1 missions.

Line 421, lunar geology Several, miss period(.)

have many grammer errors, need to be improved

Author Response

Below I list the changes I've made to the document to address your suggestions for improvement. The lines on which you can find these changes are specified in the response document. I hope that the new version will be easier to understand and that all the linguistic and formal problems have been fixed. Sincerely.

Dear Reviewer of the article remotesensing-2903505 "Lunar Surface Resources Exploration: Tracing Lithium, 7Li, and Black Ice using Spectral Libraries and Apollo Missions' Samples",

I would like to express my gratitude to you for your efforts in improving my work and providing me with the opportunity to publish it. Your anonymous feedback has helped me meet the quality standards required by the journal.

Reviewer #2: “The authors conducted PCA to explore the concentration of Lithium, Li….”

The methodology, both for obtaining the spectral indices (C1_C5) and for deriving the 5 types of spectra from the 19 spectra downloaded from RELAB, has been extensively detailed. For this we have included:

(In lines 211-229).

1. A detailed description of the Principal Component Analysis (PCA) performed on the 19 RELAB spectra. This analysis synthesized the 19 spectra into just 5 type spectral curves, which are shown in Figure 1. These curves are representative of 5 different substrates, each with a different average concentration of lithium and 7Li.

(In new Table 2)

2. A table of the variables used to calculate each spectral index. The spectral index captures the most extreme reflectance characteristics of each type of spectral curve and is composed of a coefficient that multiplies a subtraction of bands, as detailed in Equation 1. The selection of the highest wavelength of the peaks and the lowest of the troughs was done with a spectroscopy program used to manage the spectral curves... it is not a choice of the author.

Reviewer #2: “In section 2.2 and 2.3, the authors using existing GIS tools and lunar topographic data (such as LOLA DEM) to investigate the distribution of Lithium and relief patterns as well as dirty ice. But it lacks methodology innovation. It seems that the author only used these existing tools without making any corresponding improvements”

I must politely disagree with this, as the methodology used is completely new to lunar exploration. In the extensive literature reviewed, there are no studies using this method or approach. The authors of this paper developed the ice libraries. It's true that the tools used already exist, but the goal of this work is not to develop new tools, but to use what is available to create a new method and new information. I think we have done that. However, we are working on a neural network to do the same exploration, both to look for the most significant spectral features and to classify the images.

Review# 2:  “Grammar and spelling errors:….”

The linguistic and grammatical aspects have been thoroughly reviewed and we hope that the new version will correct all errors known to us.

Author Response File: Author Response.pdf