SNOWTRAN: A Fast Radiative Transfer Model for Polar Hyperspectral Remote Sensing Applications

Round 1

Reviewer 1 Report

Please see attached document for comments. Main suggestion is to show additional cases in another season (e.g. summer) of the SNOWTRAN versus satellite observations of reflectance.

Comments for author File: Comments.pdf

The text describing the equations needs to be improved for clarity.

Author Response

Please, see the attachment.

Author Response File: Author Response.docx

Reviewer 2 Report

This is a well written and organized paper.

Author Response

Thanks for your work with our paper.

Reviewer 3 Report

This is a good topic to compare two Hyperspectral data , PRISMA and ENMAP for analysis, in snow covered regions using snowtran software.

1. Abstract: what do you mean by DOME C, pls clarify , is this study site,,, so mentioned it earlier (what is dome c.  study site , pls clarify to readers. I assume its concordia research station. so kindly write there in ( )...

Page no; 2 Line 59- 60.

2. open source or licensed > pls mention
3. is this code based on GUI based ?? mention.
4. Availability  ?? pls mentioned webpage if available.

5. Page 3 line 90:  missing or incomplete sentence... pls be careful with missing or incomplete sentence.

6. page 1 line 36- when ever start new information like PRISMA or ENMAP, kindly provide detailed information like where it is acquired like ITALIAN Space Agency (ASI) for PRISMA... and few parameters or specifications of datasets.

7. Page 2, provide some information about SNOWTRAN software- like its code based or GUI based, open source or license, and if available freely or on request or on websites...????

8 fig 1 for enmap or PRISMA??

9 Figure 3 - represents curves for  ENMAP or PRISMA  >>???

10. Figure 4, - there are 4 sub figures but authors mentioned only about a, b and c, one sub figure between a and c is not properly captioned. (marked in Annotated PDF).

11. Actually Authors need to provide detailed captions to each figures, if they are mentioning it to ENMAP or PRISMA... nowhere it is done.. so each figure captions must be thoroughly checked.

12 for comparative purpose, authors must provide similar processing work for both sites.  i can see- for DOME C-  the following is missing- like figure 6 and 7.  "The retrieved spatial distribution of snow grain diameter and "....The spatial distribution of the parameter k., (like figure 7) .

13. Where is the PRISMA based spatial distribution of snow, and parameters k for both sites...

Authors must present , similar processing over both sites using both datasets, ENMAP and PRISMA to compare and relate or show differences...  the major flaws of the present study . if authors incorporate  this after revision

Comments for author File: Comments.pdf

Author Response

Please, see the attachment.

Author Response File: Author Response.docx

Reviewer 4 Report

In this work, the author has introduced a novel forward model for snow surface optical properties that considers atmospheric factors, enabling rapid and comprehensive simulation of typical snow characteristics. Although the model has more uncertainties in the retrieval of complex snow surface parameters, it still does not hinder the validity of the model when retrieving snow optical properties parameters for hyperspectral remote sensing data. In the latter half of the manuscript, the concept of the effective radius ratio is proposed, significantly enhancing the interpretability of snow grain size inversion in optical remote sensing. Undoubtedly, this is a challenging yet promising endeavor. The results of this paper are quite intriguing, yet there remain a few minor confusions:

Comments:

Based on these models, what is the feasibility of using EnMAP and PRISMA data to invert the pollutant content of polar snow surface?

The second is whether the set of theories can be extended to conventional optical remote sensing data. What are the uncertainties? What issues need to be noted?

What are the assumed types of pollutants in the dirty snow simulation results in Figure 3? This is important because the difference between black carbon and dust at 50 ppm (Table 1) on the reflectance of snow cover is considerable.

The results of the new snow forward model are very encouraging, as you can see in Figure 4, for pure snow, the vast majority of the simulation results of the SNOWTRAN spectra and EnMAP in the range of 350-2500 nm are highly consistent, and only distortion is in the range of 418-500 nm. What is the potential cause of this phenomenon, and is it due to EnMap calibration uncertainty?

In my view, the concept of effective radius ratio significantly enhances the interpretability of optical remote sensing inversion of snow grain size, as it confirms that the inverted snow grain size can represent the size of grains in a certain thickness of the snowpack. The effective radius ratio greatly extends the interpretability of grain size inversion based on analytic asymptotic theory, marking a groundbreaking effort. Can I then assume that the greater the spatial heterogeneity of the K value, the greater the variability in the vertical structure of the snowpack in a snowfield? This is because if the snowfield were vertically uniform, the K value for every pixel in satellite imagery of the snowfield would be identical. However, I still have the following confusion and hope the author can clarify:

The author mentions that the penetration depth at 2200 nm is less than that at 1235 nm (L 405-409), which seems contradictory to previous understanding. Traditionally, it is believed that longer waves (2200 nm) have stronger penetration than shorter waves (1235 nm). If understood in this way, the lower snow reflectance at 2200 nm would result in more photons penetrating deeper into the snow, implying a greater penetration depth.

If, as described by the author, the penetration depth at 1235 nm is greater than at 2200 nm, and 2200 nm only reflects information from shallow snow (small grain size), then theoretically K1 should be less than K2. However, why is it calculated that K1=1.1 & K2=0.4 (L 418)? Is there a possibility of misused variable subscripts? I suggest marking out the region used for calculating the average equivalent optical grain size in Figure 6. Additionally, what are the potential causes for d1235 being greater than d1030? If following the framework proposed by the author, since 1235 nm is more reflective of the grain size information of the snow's surface layer compared to 1030 nm, shouldn't it be d1235 < d1030?

Others:

1.      Line 131, Eq. 9, How versatile are the results of the particle volume calculations in Eq. 9?  Can I use it when calculating typical particle morphologies? Such as sphere, fractal, cylinder.

2.      If the author could provide a table of the physical variables used in the text in the form of a table in the appendix, this would help readers improve their reading experience. For example, line 124, β, ε, ρ, σ, g₀ and g_∞. Due to the large number of equations and mathematical notations involved in the article, I suggest providing a parameter cross-reference table in the appendix section.

3.      Line 402, Does radii refer to particle diameter?

Author Response

Please, see the attachment.

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

Thank you for your thorough responses.

I believe most of the current English language/formatting issues will be resolved during copyediting.

Reviewer 3 Report

Revised version  is now much more improved than

original one.

Very minor