A New Strategy for Forest Height Estimation Using Airborne X-Band PolInSAR Data

Round 1

Reviewer 1 Report

This work is focused on the derivation of the three stage inversion algorithm based on RVOG model for forest height estimation. The main novelty of the proposed method is coming up with a new strategy for the estimation under X band PolInSAR data. First, they introduce the frequency segmentation, then they choose the ground phase according expanding the PolInSAR coherent coherence domain and judging the       sign of the ground phase. At last they obtain the volume coherent coherence according the segmentation of fitting line intercepted by coherent coherence curves. The algorithm is validated with actual x band airborne PolInSAR data through comparison with other methods. This is an interesting approach since the conventional methods are not as comprehensive as this method, thus leading to the degradation of estimation results under the X band PolInSAR data..

The paper is well-organized and presents the algorithm quite clearly. However, some minor revisions should be considered as follows:

1. Page 1, Line 20. I think the term “Extinction coefficient curve (ECC)” should be fixed extinction coherence coefficient curve.

2 Page 2, Line 69: I suggest changing the sentence: “Antropov et al. explored the multitemporal behavior of fully PolSAR parameters at L-band in relation to the stem volume of boreal forests” to “Antropov et al. explored the multi-temporal behavior of fully PolSAR parameters at L-band in relation to the stem volume of boreal forests”.

3. The use of “process” and “processing” should be unified.

4. Line 120, Line 343, Line 347, etc.: The parentheses should be indented in the proper characters.

5. Page 3, Line 121: The sentence “Figure 1. Distribution of PolInSAR scattering dots …” should be “Figure 1. Distribution of PolInCC dots …”

6. All the phases of “PolInCC” in the paper are recommended to change as “PolInCCs”. Since your research is on a series of PolInCCs.

7. Page 9, Line 329: The first letter of a sentence should be capitalized

8. Page 13, Line 457: The sentence “After the distribution area of theVCC is  constrained by a given range” should be “After the distribution area of the VCC is  constrained by a given range”

9. Page 16, Line 540: The sentence “top-tobottom overlaps may occur with large terrain fluctuations” should be “top-to-bottom overlaps may occur with large terrain fluctuations”

In addition, some questions should be taken into account carefully as below:

1. The distribution range of the polarization interference coherence coefficient in the complex unit cell of a single pixel is greatly related to the estimation method of the coherence coefficient and the size of the sliding window. How do you estimate the coherence coefficient and how large the sliding window is in the article?

2 Since the method of your paper is validated only on airborne SAR system, I suggest you expanding your research to the spaceborne platform in the future.

Author Response

Please see the attachment.

Author Response File: Author Response.doc

Reviewer 2 Report

The manuscript uses X-band to retrieve height and it is challenging. The author can get accurate results using X-band, so it is a good algorithm. The idea of manuscript is innovative and the authors have obtained an accurate result of tree height. The manuscript uses an azimuth sub-aperture algorithm which is equal to increasing the penetration depth and add samples in the unit circle. This idea is the key idea and it interests me much.

But the manuscript lacks a whole model which combines sub-aperture and RVoG together. I have several suggestions:

(1) The manuscript lacks a theorical model which should consider azimuth sub-aperture and RVoG together. It means we can get formulas after doing azimuth sub-aperture and take it into RVoG. Then, we can get a theorical explanation that why your algorithm can improve accuracy.

(2) We need an explanation that why you choose this criteria to get topographic phase in Formula (12) and (13). If you have read a paper introducing this, please give a quotation.

(3) Format of title in sub-image is wrong, such as Fig 7,8, and 10. Title (a) and (b) should be located under the corresponding figure.

(4) If you have a tree height result of whole image obtained by your algorithm, please add it. If calculating the whole image will cause much time, please add some words to describe this.

Author Response

Please see the attachment.

Author Response File: Author Response.doc

Round 2

Reviewer 2 Report

Thank the authors for giving a detailed answer. According to the relevant content, the authors added experiments to explain. The article has a relatively clear exposition of the stated views.