Volume Structure Retrieval Using Drone-Based SAR Interferometry with Wide Fractional Bandwidth

Round 1

Reviewer 1 Report

This work proposed a noval procedure for three-dimensional structure retrieval by expoiting the  frequency dependency in untrawideband InSAR imagery.  Comments and suggestions for paper revision are as follows:

(1) It is suggested that the authors remove Figure 1 from the manuscript for better briefness, as it is not much helpful (or even leading to misunderstanding) for readers to catch the ideas between the existent and the currently proposed techniques. 

(2) In its current content, it is very difficult for readers to understand teh results as presented in Figure 9 between range 173 m and 193 m. It is reluctant to say that "the measured parameters are in good agreement with the simulated parameters" ( the last paragraph on page 12  and the first paragraph on page 13).

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Reviewer 2 Report

1) It seems not very robust by using RMS as the match criterion of two shapes (coherence trend and coherence model in this paper), since different shapes may lead to similar RSM. Are there other more robust ways for better comparison?

2)The scattering models of the simulation data in this paper are assumed to be homogeneous scatterers, and if they are heterogeneous scatterers, what changes will occur in the coherence model, and what impact will it have on the comparison of the coherence trend and the coherence model? please add some explanation. 

3) The scattering mechanism simulated in this paper is different from the actual morphology of forest and vegetation. If possible, it is recommended to construct a simplified simulation scattering model of forest and vegetation to demonstrate the effectiveness in forest and vegetation height extraction.

4) From the implementation process of this paper, each SAR pixel needs to be compared with the coherent model at different heights to invert the height of the scatterer. If there is no prior information about the height of the scatterer, each pixel needs to try in this way, which will be computationally intensive, and it is recommended to add feasible solutions.

5)In Figure 10(a), there is an obvious difference between the coherence trend and the coherence model, please explain the reason for the difference.

6) The content of Part 4.4 does not seem to have much to do with the extraction of structural information expressed in the title of the paper, so please consider whether to keep it.

7) Does the speckle noise affect the extraction results of the approach proposed in this paper?

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Reviewer 3 Report

Plase see attached document

Comments for author File: Comments.pdf

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Reviewer 4 Report

This is a unique paper on tree measurement using drones and polarization interference SAR.

Please add additional explanations regarding the following points.

1. In this research, you are dealing with multiple scattered light, but the reflectance of that scattered light differs depending on the tree species and leaves. How do you handle these errors?

2. The shape of the tree changes depending on environmental location conditions and individual differences in growth. For example, it shows the crown of a tree as conical, spherical, and so far. How do you handle the change in scattered light due to this difference in tree shape?

3. In the field of mathematics, there is an approach called ``leaf layer theory'' when dealing with multilayer phenomena. Is there anything that can be applied to this research?

Author Response

Please see the attachment

Author Response File: Author Response.pdf