A Modified Geometrical Optical Model of Row Crops Considering Multiple Scattering Frame
Round 1
Reviewer 1 Report
The paper entitled "A modified geometrical optical model of row crops considering multiple scattering frame" presents a comprehensive work that has a strong novelty in terms of the possibilities of application of the proposed model. Its applicability is of interest, but it isn't very easy to follow the steps carried out to make the process reproducible. This study is well prepared and potentially well-received work by the audience. However, it needs further major revision before being published or even its rejection if the calculation above is not clarified.
Major comments:
The whole approach is interesting, and the methodology seems to me to be correct, although I miss the addition of the intercomparison with RAMI. This is referred to in the text on three occasions [4-6], and it is noticeable that it is neither used nor argued why it had not been used.
Without going into full detail, I have been impressed by the results obtained. So much so that I took the time to convert one of the shown plots into data (randomly selected, Fig. 11 e). To my surprise, the result obtained was very different from the one indicated in the manuscript. We are talking about going from a correlation coefficient of 0.83 (as indicated in the manuscript) to 0.53 (calculated from the retrieved data), and an error of more than double (RMSE from 0.018 to 0.039). Attached you can find a couple of plots about this issue. I neither recommend the publication of this paper nor a complete review to be performed until this issue is clarified since it is affecting the whole idea and main results.
Comments for author File: 
Comments.pdf
Author Response
Please see the attachment
Author Response File: Author Response.pdf
Reviewer 2 Report
The submitted manuscript titled as „A modified geometrical optical model of row crops considering multiple scattering frame“ represents a very complex study, specifically directed to solve the problem with underestimation in biophysical modelling based on reflectance data in row crops. The main issue was that geometrical optical (GO) approach so far ignored a multiple scattering contribution of canopy reflectance in near infrared spectral range. It is a mathematically based approach which gives a physical base for RS inversion in crop growth monitoring. The authors developed simulation of reflectance based on unique geometric characteristic of row crops.
Research falls in the scope of the journal. The study is comprehensive, concise with detail description of the used methodology. Aims are well defined, while results represent advance to current knowledge in the field of physical modelling of canopy reflectance of row crops.
Introduction is concise, while progressively giving insight to the topic supported by relevant literature review.
Study is correctly designed with in detail and step-by-step elaborated methodology. Statistical methods and mathematical derivation of model are very well explained.
Comments can be found embedded in the text of the manuscript (.pdf).
Comments for author File: Comments.pdf
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Reviewer 3 Report
In this manuscript, the authors evaluated and established a row model to accurately estimate the reflectance of row crops which have geometric characteristics different from forestry. For that, they introduced a new method considering the multiple scattering framework. With the introduction of the new methodology, the authors found that can address an underestimation problem in row crops due to ignoring multiple scattering calculations in the geometrical optical approach.
I consider that the study is interesting and provides a potential mechanism for remote sensing inversion based on the physical model. Thus, I suggest that this manuscript should be considered for publication in Remote Sensing after minor revision.
Particular comments
In my opinion, the article follows an unconventional structure. According to the journal requirements, the manuscript should have the following sections:
Introduction, M+M, Results, Discussion, and Conclusions.
L242. The spelling of “conponents” is wrong.
The numbers of figure and where they appear it should be more consistent. This item need attention.
Figure 5 should be after the subtitle: “3.2.1. Field measurement data in Zhangye”.
I think the description of figure 6, 7, and 8 that appears in lines 324-328 it is not necessary because appear again in the result section where it is more consistent with the figures.
Regard the Figures 6, 7, and 8, what is the meaning of VZA (º)? The letter size and numbers on the axes could be smaller. The size could be like that Figure 9. (same comments for the Figure 11.).
In the description of field measurements, it would be interesting include total number of measurements throughout the crop cycle and time of measurement.
The Figure 11 in the main text appear before to the Figure 10.
L386. The spelling of “compuer” is wrong, it should be “computer”.
It would be clearer if Table 4 appears after line 414.
The legend of Figure 7 could appear down of the figure.
L515: Did you calculate correlation coefficient (R) or determination coefficient (R2)?
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Round 2
Reviewer 1 Report
I have randomly selected other figures (in this case fig 11 c and 11 f). The first one is correct, but in the second (11f) I can see that the RMSE is much higher (RMSE = 0.02157016 instead of 0.0083)
What are the reasons for these calculation errors? Again the same? I am checking them myself by choosing graphs at random, which makes me think that more plots/figures/results may be affected.
On the other hand, using RMSE does not seem valid if we are comparing data at different scales, this measure would have to be normalised (e.g. NRMSE in any of its variants). Otherwise, we are comparing apples and oranges.
I suggest that all calculations are reviewed in depth, appropriately shown as a result, and compared in some way that they are comparable, as well as the discussion part should be significantly improved in this regard. I have found hardly any changes since the previous review.
This journal deserves a thorough review, and that is why I continue to recommend major changes, changes that were also requested before and not carried out in full, only avoiding the errors exposed, without going into detail about the possible underlying problem.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
