Slow Deformation Time-Series Monitoring for Urban Areas Based on the AWHPSPO Algorithm and TELM: A Case Study of Changsha, China

Round 1

Reviewer 1 Report

Dear authors,

The paper requires major changes before publication can be considered. As a starting point, the comments and questions below should be carefully considered before resubmitting. 

Questions/Comments:

1.      It is better to change to title. It does not reflect the contents of the manuscript very well.

2.       In abstract, line 2, PSI should be put in the parentheses as abbreviation. Before writing the abbreviation, the terms should be introduced, e.g. Permanent scatterer interferometry (PSI). Please do it for all the abbreviations which appear for the firs time in the manuscript.

3.      English proof-reading of the manuscript is recommended.

4.      In page 1, line 41, it is mentioned that the deformation sequences with a sub-millimeter accuracy is obtained from the PS time series analyses. What is the quality of the generated PSI data?

5.      On page 12, line 337, H is missing. “ H denotes …”

6.      What are the deformation values for each of the PS points (PS1 to PS4) shown in figure 12?

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 2 Report

General comments:

In this study, TerraSAR - X data was used to invert the time series of historical deformation in the Tangerine Island region of Changsha City, China. An improved method was proposed to overcome the DS identification difficulty, where Sentinel 1A data and SBAS - InSAR method were used for cross validation. The main innovation of this paper is to propose a novel time-series InSAR processing based on Adaptive Window Homogeneous Pixel Selection and Phase Optimization algorithm. This method introduces the physical thermal expansion component into the traditional linear model to compensate the lack of homogenous samples and low phase quality. However, only the SBAS results rather than the field measurements are used for verification, and only two points are considered, so the accuracy of this method needs further verification.

 

Specific comments:

1.     Figure 1, I suggest authors to highlight the improvement steps in the technical flow chart.

2.     The labels such as A/B/D/E/G/I etc. Figure 4 and Figure 8 and others need consistency to decrease the confusion.

3.     In this study, the phase optimization refers to interferogram filtering, and the new method is compared with traditional Goldstein, rather than with any improved one, please clarify the considerations.

4.     In the result part, it’s better to analyze the annual velocity map and thermal expansion coefficients map with same regions, which can improve the importance of this research.

5.     In the analysis and discussion section, it is suggested to combine 5.1 and 5.2. Besides, as for the correlation analysis, the precipitation is highly correlated with temperature, so some related statement should be mentioned carefully.

6.     Accuracy analysis, only the SBAS results from two points were used. It’s better to collect in-situ measurement, such as land subsidence measurements and underground water level data.

7.     The English expression needed improvement thoroughly.

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 3 Report

In response to the problems of the conventional DSI technique in monitoring deformation, the paper proposes a time-series InSAR deformation estimation method for urban areas that considering the deformation information, and the work is demonstrated in a concise and clear way, and the results are very encouraging. The paper contains new idea and can be accepted after a minor revision:

1. Did you use your own programs to run the InSAR processing?

2. The accuracy of satellite InSAR deformation monitoring can hardly reach sub-millimeter level, and it is suggested to be revised to "sub-centimeter level".

3. Why the thermal expansion linear model can completely describe the deformation you aim at? There may be other reasons for deformation for components.

4. DSI may be better suited for mining areas or landslides, is it suitable for the urban areas?

5. Is area D in figure 8 and figure 9 the same area?

6. How to ensure that the PS1 and PS3 in figure 12 are on the bridge?

7. P3 line 99 "target image elements", not properly expressed.

8. P3 line 122-123 "After the homogeneous candidates being identified, the random decorrelation noise will cause the interferometric phases not consistent"，I suggest change it to "After the homogeneous candidates being identified, the random decorrelation noise will cause the interferometric phases inconsistency. "

Author Response

Please see the attachment.

Author Response File: Author Response.docx