Radiometric Improvement of Spectral Indices Using Multispectral Lightweight Sensors Onboard UAVs

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Review of drones-2746671 “Radiometric Improvement of Spectral Indices Using Multi-spectral Lightweight Sensors Onboard UAV” by P. Andrés-Anaya et al.

This a very good topic. The author knows the area of research very well since BRDF is complex and the radiometric management of data sometimes is forgotten. Particularly, in this manuscript, a methodology is presented to improve the reflectances of different bands, and thereby improve spectral indices such as NDVI, from UAV-borne multispectral sensors. These improvements address the complex problem of BRDF in depth which is very interesting. Furthermore, the automated process is compared with the results of one of the most powerful and well-known commercial photogrammetric software, Metashape, providing interesting results for agronomic applications. The study is comprehensive and the results seem reasonable.

Major issues:

·        The title mentions the radiometric improvement of Spectral Indices, but only shows the improvement of NDVI.

·        The abstract should emphasize two aspects that are crucial in this contribution:

-        This approach is of particular interest in precision agriculture.

-        The BRDF effects are due to large changes in the relative position between the sensor, the Sun and the vegetation, as well as the possible presence of shadows inside the crop.

·        The abstract mentions that this process is automated and integrated in the QGIS Photogrammetry Tools plugin, but the plugin is not mentioned again in the rest of the document. Please provide more information if this plugin has been published, otherwise remove the information from the abstract.

·        The novel contribution of hotspots in the proposed methodology is not understood. Are hotspots simply detected and that information removed or is there some kind of novelty in the transformation of value? Neither do not understand the assertion that the original model does not verify the reciprocity principle. Explain in more detail.

·        Regarding the methodology and the pipeline, a final step of calculation of final vegetation indices including BRDF should be added. In addition, this last step shouldn’t be confused with the first NDVI calculation, since the last step of radiometric processing is done to discriminate between crop and non-crop pixels, so that only crop pixels are used when modelling BRDF.

Minor remarks:

·        Remove text from the template: Acknowledgements and supplementary material.

·        Check subscripts and superscripts: Line 157 (CaCO₃), 158 (cm³), 232 (R^W), etc.

·        Line 154: 400 C or 40° C?

·        Check double spacing: Line 179 and 555.

·        It is recommended to add a figure of the reflectance panels and Ground Control Points used.

·        Line 335: Nir NIR

Comments on the Quality of English Language

Minor editing of English language required

Author Response

The authors sincerely appreciate the positive evaluation of our work and the valuable suggestions, which have contributed to enhancing the quality of the manuscript.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

Accept in present form

Author Response

The authors sincerely appreciate the positive evaluation of our work.

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

In the article, the authors tried to present a method for correcting multispectral data. But in my opinion, they chose the wrong method to confirm the adequacy of their method.

Comparing the method accurately described in the article, the correction method, with software with unknown mathematics inside is quite strange. Naturally, the authors draw the only possible conclusion in such a situation - their results are different!

In my opinion, the authors could easily confirm the advantage of their method if they removed the same section of the underlying surface in the center and along the edge of the field of view. And on these two frames, you can compare which method will give the most identical results, both for reflectivity and for calculating NDVI.

And last note, when remote sensing from a satellite, the meaning of the NDVI index is clear. Its value roughly corresponds to the amount of green mass in the pixel projection. But what determines the NDVI index for a single leaf of a plant? If we talk about using it for agriculture, is it easier to average the index value over a fairly large area in order to return to the classical understanding of NDVI? In my opinion, the authors need to formulate the task they are solving more precisely.

Author Response

The authors are sincerely appreciative of the reviewer's comments and valuable suggestions, which have undoubtedly contributed to enhancing the quality of the manuscript.

Author Response File: Author Response.pdf

Reviewer 4 Report

Comments and Suggestions for Authors

For my full feedback please see the pdf document with the comments in. 

In general though you should include an independent dataset if you want to say that one method is better than another. By showing that they are different doesn't mean that one is better than another - for me that is the key weakness in your paper. As a result I haven't done a detailed review of the statistical analysis as I don't think it adds anything meaningful.

 

Having said that, I can see the use of a method and software that can calculate BRDF corrected reflectances from UAV data. On your choice of BRDF model, I would highlight more clearly that the model chosen is not physically consistent - for some applications this may be acceptable but for others it will not be.

Comments for author File: Comments.pdf

Comments on the Quality of English Language

The English in the paper is ok - I have made corrections to it where it has come up (I am a native English speaker but I appreciate that writing a scientific paper in a language that isn't your native language is challenging) although in some cases I did not understand the point you were trying to make (I have made this clear).

Author Response

The authors are sincerely appreciative of the reviewer's comments and valuable suggestions, which have undoubtedly contributed to enhancing the quality of the manuscript.

All the comments and suggestions in the PDF have been attended.

Author Response File: Author Response.pdf

Round 2

Reviewer 4 Report

Comments and Suggestions for Authors

Thanks for taking on board the comments provided originally. Most of the comments/edits left are changes to the English.

Comments for author File: Comments.pdf

Comments on the Quality of English Language

Author Response

Dear Reviewer,
we again deeply appreciate the rigour and effort you have put into this second review. All the issues raised have been rectified. Honestly, thank you for making such thorough revisions. It is not easy to find reviewers like you.

Best wishes,

The authors

Author Response File: Author Response.docx