A Unified Physically Based Method for Monitoring Grassland Nitrogen Concentration with Landsat 7, Landsat 8, and Sentinel-2 Satellite Data

Round 1

Reviewer 1 Report

The presented study presents an interesting approach for N% concentration monitoring in grasslands from multiple sensors including Landsat 7, Landsat 8, and Sentinel-2 by integrating multiple RTMs combined with bottom-of-atmosphere (BOA) spectral bidirectional reflectance distribution function. Severe ambiguities in terms of results presentation and discussion are mentioned. More specifically, the findings are documented but there is limited explanation about the reasons and factors that contributed to  these results and how  are these results linked to the previously conducted studies. Moreover, Conclusion should be merged to one paragraph, focusing mainly on the proposed method, the short but clear documentation of the results and the innovation of the current study and its contribution the field of N% monitoring in grasslands.  The more descriptive parts are suggested to be moved to the discussion section.

The authors are suggested to take into consideration the above comments and provide a revised version of their work.

Some additional suggestions are given as follows:

Page 1 Line 5 RTMs. Please provide the whole term (Radiative Transfer Modelling,) since it  appears first time in the manuscript.

Page 2 Lines 32-32 How is this achieved through optical sensing? Please elaborate more.

Page 2 Lines 36-37 How wach of them( Landsat 7 (L7), Landsat 8 (L8), and Sentinel-2 37 (S2)) contribute to  overcome the cloud coverage and image acquisition time ?

Page 2 Line 38 Explain the reason these sensors use remain a challenging task.

Lines 50-51 The verb is missing. Please reform the sentence accordingly, since at its state it is hard to make sense.

Line 72 reference is missing

Line 79 You are mentioning the biophysical and biochemical variables, but you are not refereeing to some of them for instance.

Line 107 Reform the sentence so as to  introduce the reader better to your approach. For Example begin the paragraph like “In the current work, a machine learning model using a Gaussian Process Regressor (GPR)  has been developed for……. “ or likewise.

Line 127 Please be more specific about the sampling locations you used

Line 212 Which were the most informative biochemical and biophysical variables you found for estimating N% ?

Line 215 Add lines to Table 2. There is no reference to the Table 2 in the text.

Figure 3 Should be moved to 3.1 section

Lines 248-250 How is this attributed?

Lines 260-261 Hows is this attributed?

Line 268-270 How is this attributed? Please elaborate shortly in 2-3 lines

Lines 294-295 Reform “760 nm to 890 nm region” to “the region between 760 nm to 890 nm”. Do the same for the 450-60nm region

Line 299-300 Why is that? Please explain more.

Line 305 “longer wavelengths”. Please provide the range

Lines 392-393  Chance the expression other methods with alternative methods, since it is more suitable

Lines 394 Replace as to since

Line 398 the expression  “can yield improved and timely monitoring of grassland N%.” does not make sense. Please reform

Lines 400-406 The whole paragraph needs to be totally reconstructed, since it is vague. At this point of the manuscript the author is recommended to summarize both the findings and the limitations of their presented study, highlighting the novel points of their work, and provide further possibilities of potential expansion of their study in the near future.

Line 408 Remove “In conclusion”

Author Response

 

Dear Reviewer,

Thank you very much for your kind feedback on our manuscript. We have addressed the raised issues and have attached the detailed responses to your queries as a Word document (Please see attachment).

Best regards,

Mohammad Dehghan-Shoar

Author Response File: Author Response.pdf

Reviewer 2 Report

While the topic is of interest, the manuscript (ms) is difficult to follow, and overally stays mostly too vague or imprecise. I added my comments within the PDF. Here are some general remarks:

1) terminology, please use more correct and widely accepted terminology. Terms as "sensor invariant model" or "sensor-specific models" are vague concepts that need better explanation, or probably more correct terminology. At several places, the writing needs to be more precise. 

2)  The workflow is hard to follow. Please add a flowchart so the conducted steps are clearer.

3) Why no maps are provided? During the whole ms the authors mention about N% monitoring from satellite data, but no maps are given? It undermines the credibility of the presented work.

4) The authors should clarify better whether they use a (numerical) inversion approach, or a hybrid model, or both? The impression arose that the authors are mixing up concepts. 

5) Note that validation results are rather low, which undermines the credibility of the presented method. 

6) Apparently the authors use GPR, but they do not explain the method. 

Altogether, I believe the ms should be better elaborated before it is suitable for submission. I am sure the authors have the capacity to improve their work to a relevant ms. 

Comments for author File: Comments.pdf

Author Response

 

Dear Reviewer,

Thank you very much for your kind feedback on our manuscript. We have addressed the raised issues and have attached the detailed responses to your queries as a Word document (Please see attachment).

Best regards,

Mohammad Dehghan-Shoar

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report

The manuscript has been improved, but I am concerned about some parts of your Discussion and Conclusion. To me, that undermines the credibility of the conducted work. The discussion section about SAR does not make sense when it comes to N% retrieval, while SAR is to some extent sensitive to vegetation structure (e.g. LAI), there is no evidence that SAR data can be used for biochemical variables such as N%. That also does not make sense given the nature of the SAR signal. That you are suggesting sensitivity towards biochemical variables, gives the impression of a poor understanding. If, however, being convinced that SAR data can be used for N% monitoring, please back it up with references. If not, I highly recommend removing it.

The same is true for your Conclusion, where you refer again to microwave imagery? Please base your conclusion on evidence, and avoid speculation. If not able to provide evidence, do not mention it.

As a general remark, please realize that by making questionable statements, it reflects on your work as a whole. As a general guideline: only write about something when fully sure, and supported by evidence. 

 

 

Author Response

Dear Reviewer,
Thank you for providing your insightful feedback on our manuscript. We have considered your comments and agree that the sections discussing SAR and microwave imagery in the Discussion and Conclusion were inappropriate and have been removed (lines 486-498 and 521-524). The corrected sections have been highlighted in the submitted draft of our manuscript.

We appreciate your constructive criticism, which has been invaluable in improving this manuscript. Once again, we thank you for taking the time to review our manuscript and for your kind feedback.
Sincerely,
Mohammad Dehghan-Shoar