Gaussian Process Regression Model for Crop Biophysical Parameter Retrieval from Multi-Polarized C-Band SAR Data
, Subhadip Dey 1, Narayanarao Bhogapurapu 1, Saeid Homayouni 2, Avik Bhattacharya 1 and Heather McNairn 3Round 1
Reviewer 1 Report
Line 84, weird ??
The application of Gaussian process region in remote sensing needs to be introduced in more detail in the background part. The contribution of this manuscript needs to be compared with the directly related achievements of predecessors
More data is needed to support the conclusion.Why these three crops were choosed.
It is better to compare the results of the same batch of data with the results of other methods
Author Response
Kindly find the Author's responses attached,
Author Response File: 
Author Response.pdf
Reviewer 2 Report
General comments:
The purpose of this work is to investigate Gaussian Process Regression (GPR) models to retrieve biophysical parameters of wheat, soybean and canola utilizing combinations of multiple polarizations from RADARSAT-2 C-band SAR data.
There are many sensitivity analyses demonstrated that many other polarimetric decompositions are sensitive to crop variables, why do not use other polarimetric decompositions?
Recommend to add more polarimetric decomposition parameters in this study.
The authors need to discussion weather the trained models can be used for other Radarsat-2 data acquired in other dates (generalization of this model) or the advantage and limitation of this study. There should be a separate section for discussion according to remote sensing instruction.
In general, the authors need to address the novelty of this study, Add more polarimetric parameters and discuss the limitations.
Detailed comments:
L102: wheat-> please specify winter wheat or spring wheat.
L78: why the use of GPR is novel in this study? Since there are papers that using GPR to retrieve vegetation biosphysical parameters (See paper Retrieval of Vegetation Biophysical Parameters Using Gaussian Process Techniques).
L109: 50 fields of various crops were selected ->total number of fields? how many fields for each type of crop?
Table 2 & Table 3: do not understand why WB and VWC are in one table (same HH/HV/VV values) but not for PAI.
L306: The HH backscatter value on 15 Jun is -8.27 dB-> The mean HH backscatter value?
Figure 3: The figures are confusing. add more description in the figure caption.
Figure 4: Keep the size of the X/ Y label same for all sub-figures. add (a),(b),(c).... in the figure captions.
Author Response
Kindly find the Author's response as attached,
Author Response File: Author Response.pdf
Reviewer 3 Report
Dear authors,
Thank you for such nice research on biophysical parameter estimation using SAR data. In such a nice written manuscript, I do not have a significant number of comments but have some suggestions. Please find them in the attached pdf. Keep up the good work, and have healthy days.
Comments for author File: Comments.pdf
Author Response
Kindly find the Author's response as attached,
Author Response File: Author Response.pdf
Round 2
Reviewer 2 Report
The authors did not add experiments using other PolSAR decomposition parameters, but explained why. I think it is acceptable.
Another suggestion is to put the novelty and the limitations discussion of the study in a Discussion section instead of Conclusion section.
Author Response
Kindly find our reply attached.
Author Response File: Author Response.pdf
