Ground Deformation of Wuhan, China, Revealed by Multi-Temporal InSAR Analysis

Round 1

Reviewer 1 Report

In this paper the authors describe the use of satellite radar interferometric results to detect and monitor ground subsidence in the area of Wuhan, China. Satellite data from different radar sensors (ENVISAT, ALOS1 and Sentinel 1) are processed and evaluated in combination with ground based measurements.

Two main comments concerning the whole “image” of the work:

1. The English text requires extensive proof reading from a native speaking English person. The text in many cases is written on a matter that the meaning is not clear.
2. The quality of the figures is very low. I am not sure if these are the final figures or I have received low resolution figures , but for sure they need a lot of improvement in order to be clear to the reader.

Analytical comments/suggestions/corrections:

1. Page2 line50. “supe-tall buildings” how do you define this , provide scale ie >100m tall
2. Page2 Line 69. You are referring first in Figure 1b and later on (line79) to figure 1a. It is better to present in the text the figures according to their numbering. First talk and describe figure 1a and then figure 1b
3. Page5 Table1. In some cases in table 5 it is not clearly described the key results.

There is a description of what these previous publications refer to, but not to the results

Constantini et al 2016 “…Cross validation between InSAR and leveling data with 5mm/year”.

It is not clear what “with 5 mm/yr” means… I assume that there is a discrepancy between InSAR and leveling of ~5mm/yr . Please clarify

Also, “..The average velocity of urban area from -80mm/yr to 40mm/yr “ the authors refer to vertical velocity or LOS ?? please clarify

Zhang et al 2019 “ …SBAS-based results describe urban constructions and infrastructures “ It is not clear what they describe as key result in this phrase. That subsidence occurred in urban areas? It is not clear what this sentence refers to.

4. Page 6 line 128. I assume that “…125 leveling benchmarks measured…” Please complete the phrase.

Also please clarify how many times the leveling benchmarks were measured. Based on a following paragraph the benchmarks were measured just 2 times in 2016 and 2017 so the results  are amplitudes for a period of 1 year and not … actual velocity deduced from multiple measurements. It would be nice to mention that (if this is the case) in order the reader to evaluate these results.

5. Page 6 line 136. “ PS-InSAR and SBAS-InSAR are the main techniques in time series processing” Please rephrase! PSInSAR and SBAS are interferometric techniques , and not techniques for time series processing. I assume the authors want to state that from these two techniques we produce time series of ground motion on the PS points. So please clarify this statement.
6. Page 7 line 165. “…3-D phase unwrapping algorithm…” which is this algorithm? SNAPHU another one… please state.
7. Page 7 line 188. “coherence threshold is set to 0.3.” A threshold of 0.3 is quite small , I would expect at least 0.6. However, please explain the reasons of selecting such a small threshold, and how this small value (0.3) reflects to the quality of the results.
8. Page 7 lines 189-190. With 3 Envisat images using stacking method, how many interferograms were produced (2 or 3) and which date was the “middle image” on table 2 tou present the 2 images but in this case I think that it is desirable to know the date of the third image.
9. Page 7 line 194. It is necessary to describe analytically how the LOS velocity was converted to vertical and not just refer to two previous publications [38,39]. Especially since the Envisat data appear to have significant different incidence angles ~23 and ~33 degrees. Moreover, the Sentinel and the ALOS data have also significant incidence angle so there is a not-tiny horizontal component .
10. Page 7 line 198. “All results are calibrated to a stable reference point “ In this line is the only time that the authors refer to the reference point used in their study. How this point was selected? Is there any knowledge for the absolute motion of this point? Is it subsiding less that other areas or not? How the “stable” characterization of the point can be justified ? Do you have any other points in the area that the absolute velocity is know from other techniques? For example a continuous GPS station or any other method that can support the characterization of a point as stable.

It is mandatory for these interferometric techniques and for the validation of the results to provide information of the reference point in order the reader to validate if the final velocity values are overestimated or underestimated

11. Page 7 line 203. “The subsidence of Hankou (including JH, HH, and QK) is more serious than that of Wuchang “ How the authors define the “more serious” provide values or range of values.

Actually, in many parts of the text the authors avoid to quantify their results. It is is more easy for the reader to have a value or a range of values in order to better understand the text. So, I propose to the authors that when they make a comment about increase or decrease , low or high to provide values.

12. Page 7 line 205. “The subsidence areas expanded to the six main subsidence zones in 2015-2019. “ It is unclear what the authors want to describe with this sentence. Which subsidence areas have been expanded.
13. Page 8 Figure 2. Please state in the figure as has been done in the caption that it is Vertical Velocity not just velocity. Also, personally I ve tried hard to see the reference point, and I am not sure that I ve seen it. So, please the authors to use a more discrete color or symbol for the reference point
14. Page 8 line 217. “ the annual subsidence velocity from -67mm to 17mm …” in this line but also in following figures, diagrams and text the authors refer to subsidence velocity using also positive values. I consider as subsidence only the negative vertical velocity values. Based on that I propose to use the term “vertical velocity” instead of “subsidence velocity” when they present or refer on values that have also positive sign. Of course, it is evident that the authors discuss the observed subsidence, but still I believe that “vertical” is more precise term.
15. Page 9 line 231. “…The spatial distribution of subsidence center is significantly…” spatial distribution and … center .. are not coincide properly. Probably correct to … “subsidence area …”
16. Page 9 line 239. “ …subsidence velocity from -52 mm/yr to 28 mm/yr based … “ Concerning the term subsidence velocity and positive values please refer to a previous comment. Another issue that I would like to raise in this point is the positive vertical velocity values (uplift) that in this case is quite significant (+28mm/yr). This is quite strong uplift but the authors do not make any comment or do not provide any explanation of the observed uplift. It is evident that the authors emphasize on the subsiding phenomena, but still an uplift of 28mm/yr is quite strong and may cause also problems to the urban environment. I propose to the authors to comment on these positive values although it is not the main scope of this work.
17. Page 9 lines 240-241. Please provide values or ranges in order the reader to better evaluate the results.
18. Page 9 line 250-251. “….time-varying 250 deformation velocity…” Although in this paragraph the authors refer to deformation as amplitude (mm) in this line authors discus the deformation velocity (mm/yr) the paragraph is not consistent.
19. Page 9 figure 3. Again the authors use the general term subsidence but there is also uplift (positive values). Also, please move the acronyms of the locations from the lower part of the diagram so to focus on the variations of amplitude. There is no meaning to scale the diagram from -280 to 80 when the TS range ends to about -160mm.
20. Page 10 Figure 4. Please change to “Vertical velocity”
21. Page 10 line 370. “ ….we calibrated the two results in the same temporal and spatial reference framework…” On the following lines there is an effort to explain this work. However, this whole paragraph is a bit confusing for me. Please rephrase and explain on a more analytical way how this was done.
22. Page 10 Lines 278-281. Here the authors state that the leveling results are from 2 epochs (see comment 4). Also, the authors state that the InSAR results “… will be different …” from the leveling results in areas where there is time-varying displacement . However, this is the case in many significant areas as these are presented in figure 3. So maybe it is better to clarify that there are differences mainly where the seasonal variations of the ground deformation is strong. Also, it may be nice the authors to make some comments and compare the pattern of deformation deduced by InSAR and levelling as well as the amplitude. This comment aims to improve the text and highlight the value of the leveling data.
23. Page 11 lines 302-304. It will be clearer for the reader to include some numerical values or ranges of motion velocities and not just to describe the differences between areas.
24. Page 12 line 320. “…The land subsidence in JH has been serious since 2007…” This is a very general statemen, how do the authors define the “serious” please be more analytical.
25. Page 12 Line 322. The profile AA’ has been marked as rectangle on the figure 6a and it is confusing. Please mark as a line and just make it bold.
26. Page 13 lines 337-339. “The maximum subsidence velocity reached about -42 mm/yr and -47 mm/yr on SH and BSZ, respectively during 2015-2017, the subsidence velocity and the coverage of subsidence areas have reduced significantly since 2018.” Please rephrase to make the sentence more easy for the reader to understand your conclusion concerning the changes during the first period 2015-2017 and since 2018.
27. Page 16 lines 392-393 “it could reveal a correlation between the geological condition of alluvial zones along Yangtze River and the results obtained by InSAR” What kind of correlation? This is a very general statement and requires a more detail description on how SAR results and geology are correlated (eg compaction ?? )
28. . Pages 17 and 18. In these two pages the authors corelate the observed deformation (mainly subsidence) with urbanization (building construction) and seasonal variations of the Yangtze river. Although this is a quite descriptive text I think that it is lacking of quantification: how long the “delayed effect on land subsidence “ is (line 432). What is the delay of subsidence when the water level changes? (line 463-464).
29. Page 18 figure 12. This is a nice figure that can be used to support the text. However, since it contains a lot of information it becomes quite “difficult” to read it. Maybe if the authors highlight some crucial zones that the data are well correlated, or mark some specific dates it would be much more efficient to the reader.
30. Page 19. The correlation of the water level changes and ground vertical motion is quite critical in such study. Is there a possibility for a numerical model of these data in order to better corelate water table horizon changes and land subsidence. A model will certainly increase the impact of this study. In any case this is a recommendation that is not necessary to be addressed in this present study.
31. Page 20 figure 13. One more time I refer to the use of term subsidence in the Y-axis instead of a more appropriate “vertical motion”
32. Page 21 line 495. “…for InSAR time series analysis…” personally I believe that the term time series analysis refers to a more mathematical and statistical approach of the data. At this point, I think that the authors refer to “time series interpretation” since what they present is how the ground deformation changes with time and they do not present any more detailed analysis on the time series.

Author Response

Point1

1. The English text requires extensive proofreading from a native speaking English person. The text in many cases is written on a matter that the meaning is not clear.
2. The quality of the figures is very low. I am not sure if these are the final figures or I have received low-resolution figures, but for sure they need a lot of improvement in order to be clear to the reader.

Response1: Thank you very much for the insightful comments on this manuscript! In the revised manuscript, we tried our best to improve the manuscript based on your constructive suggestions. Please see the specific responses below. The manuscript has been proofread by a native English speaker and the pictures are replaced by high-resolution figures.

Point2

Page2 line50. “supe-tall buildings” how do you define this, provide scale ie >100m tall

Response2:Thanks for your comments! Here the super-tall buildings are defined as height over 200 m, and we have added the details in Line 49 of the manuscript.

Point3

Page2 Line 69. You are referring first in Figure 1b and later on (line79) to figure 1a. It is better to present in the text the figures according to their numbering. First, talk and describe figure 1a and then figure 1b

Response3: Thanks for your comments! Done! We have revised them in the manuscript in Line 77&81.

Point4

Page5 Table1. In some cases in table 5, it is not clearly described the key results.

There is a description of what these previous publications refer to, but not to the results

Constantini et al 2016 “…Cross-validation between InSAR and leveling data with 5mm/year”.

It is not clear what “with 5 mm/yr” means… I assume that there is a discrepancy between InSAR and leveling of ~5mm/yr . Please clarify

Response4:Thank you for your suggestions. The 5 mm is the standard deviation of the differences between leveling and InSAR measurements. We have made revisions in table1 accordingly.

Also, “..The average velocity of urban area from -80mm/yr to 40mm/yr “ the authors refer to vertical velocity or LOS ?? please clarify

Response4:Thank you for your comment. The papers of Constantini et al. (2016) did not describe the direction of the average velocity.

Zhang et al 2019 “ …SBAS-based results describe urban constructions and infrastructures “ It is not clear what they describe as a key result in this phrase. That subsidence occurred in urban areas? It is not clear what this sentence refers to.

Response4:Thanks very much. The paper of Zhang et al 2019 describes the long time (2015-2018) subsidence in Wuhan and the urban constructions are the main driving factor subsidence in Wuhan. Done in table2.

Point5

Page 6 line 128. I assume that “…125 leveling benchmarks measured…” Please complete the phrase.

Also please clarify how many times the leveling benchmarks were measured. Based on the following paragraph the benchmarks were measured just 2 times in 2016 and 2017 so the results are amplitudes for a period of 1 year and not … actual velocity deduced from multiple measurements. It would be nice to mention that (if this is the case) in order the reader to evaluate these results.

Response5:Thanks for your suggestions! Done! Line 138-140.

Point6

Page 6 line 136. “ PS-InSAR and SBAS-InSAR are the main techniques in time series processing” Please rephrase! PSInSAR and SBAS are interferometric techniques and not techniques for time series processing. I assume the authors want to state that from these two techniques we produce time series of ground motion on the PS points. So please clarify this statement.

Response6:Thanks! Done! We are sorry for our ambiguous expression. We have revised the sentence in Line 150-151 accordingly.

Point7

Page 7 line 165. “…3-D phase unwrapping algorithm…” which is this algorithm? SNAPHU another one… please state.

Response7:Thanks for your suggestion. We added the 3-D phase (two in space and one in time) unwrapping algorithm in Line 179&180. The two-dimensional unwrapping algorithm is SNAPHU. The details of the 3-D unwrapping algorithm are introduced in the research [38].

Point8

Page 7 line 188. “coherence threshold is set to 0.3.” A threshold of 0.3 is quite small, I would expect at least 0.6. However, please explain the reasons for selecting such a small threshold, and how this small value (0.3) reflects to the quality of the results.

Response8:Thanks for your suggestion. We agree with your opinion. As the SLC images are not clipped in InSAR processing, the coherence values are high in an urban area and low in rural areas. The coherence value is the mean value of the whole interferograms in this paper and the average coherence value of all the interferograms is 0.328. In order to maximize the interferogram correlation and ensuring there were no isolated clusters within the resultant network of image-pairs. Thus we selected the coherence threshold value as 0.3.

Point9

Page 7 lines 189-190. With 3 Envisat images using stacking method, how many interferograms were produced (2 or 3) and which date was the “middle image” on table 2 tou present the 2 images but in this case, I think that it is desirable to know the date of the third image.

Response9:Thanks for your thoughtfulness. Done in table2. Three interferograms were processed for stacking.

Point10

Page 7 line 194. It is necessary to describe analytically how the LOS velocity was converted to vertical and not just refer to two previous publications [38,39]. Especially since the Envisat data appear to have significantly different incidence angles ~23 and ~33 degrees. Moreover, the Sentinel and the ALOS data have also a significant incidence angle so there is a not-tiny horizontal component.

Response10:We appreciate and understand your comments. We have converted all the PS/SDFP points from LOS into vertical with corresponding incidence angles. Done! Line 209&210.

Point11

10. Page 7 line 198. “All results are calibrated to a stable reference point “ In this line is the only time that the authors refer to the reference point used in their study. How this point was selected? Is there any knowledge for the absolute motion of this point? Is it subsiding less than other areas or not? How the “stable” characterization of the point can be justified? Do you have any other points in the area that the absolute velocity is known from other techniques? For example a continuous GPS station or any other method that can support the characterization of a point as stable.

It is mandatory for these interferometric techniques and for the validation of the results to provide information of the reference point in order for the reader to validate if the final velocity values are overestimated or underestimated.

Response11:We are sorry for our ambiguous expression. The reference point is a WUHN International GNSS Service (IGS) station, which is located in Wuhan University. Line 214-216.

Point12

Page 7 line 203. “The subsidence of Hankou (including JH, HH, and QK) is more serious than that of Wuchang “ How the authors define the “more serious” provide values or range of values.

Actually, in many parts of the text, the authors avoid quantifying their results. It is easier for the reader to have a value or a range of values in order to better understand the text. So, I propose to the authors that when they make a comment about increase or decrease, low or high to provide values.

Response12:Thanks! Done! We have provided the values to describe the subsidence in Hankou is severe than Wuchang. Done! Line 220-223.

Point13

Page 7 line 205. “The subsidence areas expanded to the six main subsidence zones in 2015-2019. “ It is unclear what the authors want to describe with this sentence. Which subsidence areas have been expanded.

Response13:We are sorry for the ambiguous expression in this sentence. We meant that the subsidence areas are larger during 2015-2019 than 2007-2010. The subsidence areas are mainly concentered in Hankou in 2007-2010; however, the subsidence areas were expanding in Hankou and Wuchang during 2015-2019, especially in Wuhcang. Line 224-225.

Point14

Page 8 Figure 2. Please state in the figure as has been done in the caption that it is Vertical Velocity, not just velocity. Also, personally, I've tried hard to see the reference point, and I am not sure that I've seen it. So, please the authors to use a more discrete color or symbol for the reference point

Response14: Thank you for your advice. We have revised Figure 2 to make it more clear.

Point15

Page 8 line 217. “ the annual subsidence velocity from -67mm to 17mm …” in this line but also in following figures, diagrams and text the authors refer to subsidence velocity using also positive values. I consider as subsidence only the negative vertical velocity values. Based on that I propose to use the term “vertical velocity” instead of “subsidence velocity” when they present or refer on values that have also a positive sign. Of course, it is evident that the authors discuss the observed subsidence, but still, I believe that “vertical” is a more precise term.

Response15: Thank you for your advice. We have replaced “subsidence velocity” with “vertical velocity” throughout the manuscript. Line 236.

Point16

Page 9 line 231. “…The spatial distribution of subsidence center is significantly…” spatial distribution and … center .. are not coincide properly. Probably correct to … “subsidence area …”

Response16:Thanks, Done. Line 249-250.

Point17

Page 9 line 239. “ …subsidence velocity from -52 mm/yr to 28 mm/yr based … “ Concerning the term subsidence velocity and positive values please refer to a previous comment. Another issue that I would like to raise in this point is the positive vertical velocity values (uplift) that in this case is quite significant (+28mm/yr). This is quite strong uplift but the authors do not make any comment or do not provide any explanation of the observed uplift. It is evident that the authors emphasize on the subsiding phenomena, but still an uplift of 28mm/yr is quite strong and may cause also problems to the urban environment. I propose to the authors to comment on these positive values although it is not the main scope of this work.

Response17:Thank you for your advice. These results were acquired from the papers which were published in 2019. Here we only used these results as a reference.

Point18

Page 9 lines 240-241. Please provide values or ranges in order the reader to better evaluate the results.

Response18: Thanks! Done. Line 258.

Point19

Page 9 line 250-251. “….time-varying 250 deformation velocity…” Although in this paragraph the authors refer to deformation as amplitude (mm) in this line authors discuss the deformation velocity (mm/yr) the paragraph is not consistent.

Response19:Thanks! Done. We have made a revision in Line 270.

Point20

Page 9 figure 3. Again the authors use the general term subsidence but there is also uplift (positive values). Also, please move the acronyms of the locations from the lower part of the diagram so to focus on the variations of amplitude. There is no meaning to scale the diagram from -280 to 80 when the TS range ends to about -160mm.

Response20:Thanks! Done. Figure 3.

Point21

Page 10 Figure 4. Please change to “Vertical velocity”

Response21:Thanks! Done. Figure 4.

Point22

Page 10 line 270. “ ….we calibrated the two results in the same temporal and spatial reference framework…” On the following lines there is an effort to explain this work. However, this whole paragraph is a bit confusing for me. Please rephrase and explain on a more analytical way how this was done.

Response22:Thank you for your advice. We derived the vertical velocity from InSAR results in Wuhan from June 2016 to June 2017, and we calibrated the InSAR-derived vertical velocity and leveling observations in the WGS84 coordinate system. Done! Line 286-289.

Point23

Page 10 Lines 278-281. Here the authors state that the leveling results are from 2 epochs (see comment 4). Also, the authors state that the InSAR results “… will be different …” from the leveling results in areas where there is time-varying displacement. However, this is the case in many significant areas as these are presented in figure 3. So maybe it is better to clarify that there are differences mainly where the seasonal variations of the ground deformation is strong. Also, it may be nice the authors to make some comments and compare the pattern of deformation deduced by InSAR and leveling as well as the amplitude. This comment aims to improve the text and highlight the value of the leveling data.

Response23:Thanks for your advice. Done. Line 300-304.

Point24

Page 11 lines 302-304. It will be clearer for the reader to include some numerical values or ranges of motion velocities and not just to describe the differences between areas.

Response24:Thanks. Done. Line 325-327.

Point25

Page 12 line 320. “…The land subsidence in JH has been serious since 2007…” This is a very general statement, how do the authors define the “serious” please be more analytical.

Response25: Thank you for your suggestions! The maximum subsidence velocity reached to -44 mm/a and -38 mm/a in 2007-2010 and 2015-2017, respectively. We have added values to describe the subsidence in JH. Line 344-346.

Point26

Page 12 Line 322. The profile AA’ has been marked as a rectangle on figure 6a and it is confusing. Please mark as a line and just make it bold.

Response26:Thanks! Done. Figure 6a.

Point27

Page 13 lines 337-339. “The maximum subsidence velocity reached about -42 mm/yr and -47 mm/yr on SH and BSZ, respectively during 2015-2017, the subsidence velocity and the coverage of subsidence areas have reduced significantly since 2018.” Please rephrase to make the sentence easier for the reader to understand your conclusion concerning the changes during the first period 2015-2017 and since 2018.

Response27: Thanks! Done Line 362-365.

Point28

Page 16 lines 392-393 “it could reveal a correlation between the geological condition of alluvial zones along Yangtze River and the results obtained by InSAR” What kind of correlation? This is a very general statement and requires a more detail description on how SAR results and geology are correlated (eg compaction ?? )

Response28:Thanks! We are very sorry for our ambiguous expression in this sentence. We intended to state that most of the land subsidence zones detected by InSAR are located on the carbonate rock belts and mucky soft soil layer.

Point29

Pages 17 and 18. In these two pages, the authors correlate the observed deformation (mainly subsidence) with urbanization (building construction) and seasonal variations of the Yangtze River. Although this is a quite descriptive text I think that it is lacking of quantification: how long the “delayed effect on land subsidence “ is (line 432). What is the delay of subsidence when the water level changes? (line 463-464).

Response29:Thanks! There are no available long-term monitoring data for the land subsidence, so we could not obtain the accurate timing for the delay effect. However, our observations suggest that the land subsidence often occurred after the rapid changes of Yangtze water levels.

Point30

Page 18 figure 12. This is a nice figure that can be used to support the text. However, since it contains a lot of information it becomes quite “difficult” to read it. Maybe if the authors highlight some crucial zones that the data are well correlated, or mark some specific dates it would be much more efficient to the reader.

Response30:Thanks! Done. Figure 12. Line 486.

Point31

Page 19. The correlation of the water level changes and ground vertical motion is quite critical in such a study. Is there a possibility for a numerical model of these data in order to better correlate water table horizon changes and land subsidence. A model will certainly increase the impact of this study. In any case, this is a recommendation that is not necessary to be addressed in this present study.

Response31: Thanks for your suggestions. Because the change of water levels is not keeping pace with land subsidence variations, and the delay effect is different in different zones. We could not model it well due to the lack of information regarding geology in the bank of Yangtze River.

Point32

Page 20 figure 13. One more time I refer to the use of term subsidence in the Y-axis instead of a more appropriate “vertical motion”

Response32:Thanks! Done. Figure 13.

Point33

Page 21 line 495. “…for InSAR time series analysis…” personally I believe that the term time series analysis refers to more mathematical and statistical approach of the data. In this point I think that the authors refer to “time series interpretation” since what they present is how the ground deformation changes with time, and they do not present any more detailed analysis on the time series.

Response33:Thanks! Done. Line 521.

Author Response File: Author Response.docx

Reviewer 2 Report

Manuscript remotesensing-980977.

This manuscript presents some investigations in Wuhan (China) with multi-sensor InSAR time series. The time series analysis over 8 years shows the spatial and temporal pattern of land subsidence in Wuhan. Results about land subsidence are correlated with leveling observations, the water level of Yangtze River, and groundwater table measurements. The authors analyses the different active phenomena and finally suggest that human activities are the main cause of this land subsidence in Wuhan.
The manuscript is easy and interesting to read. It perhaps includes too many local characteristics. However, it shows the possibilities for ground change analyses from a multi-sensor InSAR time series data set combined with other monitoring data (leveling, groundwater table level, river water level, etc.)
I suggest some minor changes.

- page 2, line 56: "earthquake monitoring due"
- Figure 1 a: The left-up map of China with the location of Wuhan could be a little larger.
- Figure 1 a: "Research area"
- Figure 1, a: External coordinate numbers are extremely small in many figures.
- Figure 1, b: The satellite image in the background doesn't help. Better to remove it.
Maybe better to use black lettering. And maybe reduce the area of the figure a bit (although the red square was partial)
- Figure 1, caption. Perhaps "affiliation" of points could be deleted.
- Figure 1, caption. Perhaps it is not necessary to include the data suppliers in this caption. Better to include it in 2.1.2.
- line 65. 113°41′-115°05′ 65 E, 29°58′-31°22′ N are limits for central China or for Jianghan Plain ?
- line 70: "elevation is low and" could be deleted
- line 73: -18ºC to 41ºC could be enough.
- line138. PS-InSAR or PSInSAR
- line 163: "Iterations continue until"
- line 301: "subsidence areas decreased"
- Figures: Authors say that warm colour represent subsidence and cold colours represent uplift. It would be true if cero value would correspond to green colour. But in most of the figures cold colours correspond mostly to no changes. Uplift is only dark blue. Then uplift is not clear in figures. It is nearly the same colour (light blue and dark blue) that small subsidence. I suggest to use green colour only close to cero, blue colour (from light to dark) only for uplift, and yellow-orange-red-brown to subsidence. It would help to distinguish between no changes, uplift and subsidence.
- Line 349: "There was 8 metro", "lines was under"
- Line 352: "lines passed through"
- Figure 10: The satellite image in the background doesn't help. Perhaps, better to remove it.
- Line 402: Perhaps "population was about 9.78)
- Line 403: Perhaps "sites was more than"
- Line 404: Perhaps "was intense", "was exploited"
- Figure 12, caption. Please, explain meaning of W1, W2, W3, W4 and complete the caption about the content of the figure.

Author Response

Response: Thank you very much for the insightful comments on this manuscript! In the revised manuscript, we tried our best to improve the manuscript based on your constructive suggestions. Please see the specific responses below.

Point1 page 2, line 56: "earthquake monitoring due"

Response1: Thanks! Done. Line 56.
Point2- Figure 1 a: The left-up map of China with the location of Wuhan could be a little larger.
Point3 - Figure 1 a: "Research area"
Point4- Figure 1, a: External coordinate numbers are extremely small in many figures.

Response2,3,4:Thanks! Done. Figure 1a.
Point5 - Figure 1, b: The satellite image in the background doesn't help. Better to remove it.
Maybe better to use black lettering. And maybe reduce the area of the figure a bit (although the red square was partially)

Response5:Thanks! Done. Figure 1b.
Point6 - Figure 1, caption. Perhaps "affiliation" of points could be deleted.
Point7 - Figure 1, caption. Perhaps it is not necessary to include the data suppliers in this caption. Better to include it in 2.1.2.

Response6,7: Thanks! Done. Figure 1b, Line 145-148.
Point8- line 65. 113°41′-115°05′ 65 E, 29°58′-31°22′ N are limits for central China or for Jianghan Plain ?

Response8: Thanks! These are the limits for Wuhan。
Point9- line 70: "elevation is low and" could be deleted

Response9: Thanks! Done. Line 82.
Point10 - line 73: -18ºC to 41ºC could be enough.

Response10: Thanks! Done. Line 86.
Point11 - line138. PS-InSAR or PSInSAR

Response11: Thanks! Done. Line 153.
Point12 - line 163: "Iterations continue until"

Response12: Thanks! Done. Line 177.
Point13 - line 301: "subsidence areas decreased"

Response13: Thanks! Done. Line 327.
Point14- Figures: Authors say that warm colour represent subsidence and cold colours represent uplift. It would be true if zero value would correspond to green colour. But in most of the figures cold colours correspond mostly to no changes. Uplift is only dark blue. Then uplift is not clear in figures. It is nearly the same colour (light blue and dark blue) that small subsidence. I suggest to use green colour only close to cero, blue colour (from light to dark) only for uplift, and yellow-orange-red-brown to subsidence. It would help to distinguish between no changes, uplift and subsidence.

Response14: Thanks a lot for your advices. The uplift is not obvious in Wuhan, and most of blue areas are stable. We revised the text accordingly.
Point15 - Line 349: "There was 8 metro", "lines was under"

Response15:Thanks! Done. Line 375-376.
Point16 - Line 352: "lines passed through"

Response16: Thanks! Done. Line 378.
Point17 - Figure 10: The satellite image in the background doesn't help. Perhaps, better to remove it.

Response17: Thanks! Done. Figure 10.
Point18 - Line 402: Perhaps "population was about 9.78)

Response18: Thanks! Done. Line 431.
Point19 - Line 403: Perhaps "sites was more than"

Response19: Thanks! Done. Line 433.
Point20 - Line 404: Perhaps "was intense", "was exploited"

Response20: Thanks! Done. Line 433-434.
Point21 - Figure 12, caption. Please, explain meaning of W1, W2, W3, W4 and complete the caption about the content of the figure.

Response21: Thanks! Done. Figure 12.

Author Response File: Author Response.docx

Reviewer 3 Report

The manuscript Ground Deformation of Wuhan, China, revealed by 2 Multi-temporal InSAR analysis is fine and interesting, but it must be improved before the publication. The figures are not well, I suggest doing it again, in order to be more readable and uniform.

The introduction lacks references to previous works and of an important concept:

Given that InSAR is a differential method, all the measurements produced are “relative” and not “absolute”, both in space and time [A]. This is because each measurement is referred temporally and spatially to respectively a unique reference image (“Master” image) and to a reference point arbitrarily supposed to be motionless. The master scene is chosen in order to maximize the total coherence of the interferometric stack and to keep the dispersion of the normal baseline values as low as possible.

For these reasons, the InSAR technique is not an ideally stand-alone technique for studying larger areas involved in phenomena such as subsidence and uplift zones, but it requires comparison and calibration with another kind of information such as GNSS. Moreover, an additional drawback of InSAR is that it cannot recognize constant movements involving ground reference points, such as crustal movements that occur over large areas, and the accuracies of low wavenumber components of the velocity field decrease with increasing distance from them [B].

[A] Farolfi, G., Del Soldato, M., Bianchini, S., & Casagli, N. (2019). A procedure to use GNSS data to calibrate satellite PSI data for the study of subsidence: An example from the north-western Adriatic coast (Italy). European Journal of Remote Sensing, 52(sup4), 54-63.

[B] Colesanti, C., Ferretti, A., Prati, C., and Rocca, F.  (2003). Monitoring landslides and tectonic motions with the Permanent Scatterers Technique. Engineering Geology, 68, 3-14.

Other suggested modifications:

In all the maps, replace with a bigger font size the coordinates around the frame and the legend. Remove or make less visible the label reported at the bottom of each map.
Put the legend of the color inside each map, I suggest with a line bar.

Figure 1. In figures A and B, replace with a bigger font size the coordinates around the frame and the legend. The map of China as well inside figure 1a is not readable.   Put a white background at the labels inside the figure 1.

Line 118-120. Are you meaning images for scenes?
Did you use only 3 ENVISAT images in the processing?
How many ascending and descending mages did you process?

Line 127. Please, provide a figure with a focus on leveling benchmarks. Is it a high-resolution leveling network? Please, draw the path or the graph of the leveling network.

line 193-194 Please, insert the formula after the phrase: all of the results are converted from line-of-sight into vertical [38,39].

Figure 3. Increase the thickness of the lines.

Figure 4c. Make of a different color and increase the thickness of the histogram line.

Figure 6. Reorganize all the figures. Put the three maps on the same line and the picture in a different figure also in one line.


Figure 7. Increase the font of the cartesian information and the thickness of the points.

Figure 8. Reorganize all the figures. Put the three maps on the same line and the color legend inside the maps.

Figure 9. Increase the font of the color legend inside the maps.

Figure 11. Move the pictures c and f to the left side.


The last correction regards the unit of measure of the velocity of displacements.
In astronomy, geology, and paleontology, the abbreviation yr for years and ya for years ago are sometimes used, combined with prefixes for thousand, million, or billion, but they are not SI units.
The SI unit for the year is not mm/yr, but mm/a where ‘a’ is annum in the SI units  (per annum means "each year").

I suggest the publication after an accurate minor revision.

Author Response

Response:

Thank you very much for the insightful comments on this manuscript! In the revised manuscript, we tried our best to improve the manuscript based on your constructive suggestions. Please see the specific responses below.

Ponits1 The introduction lacks references to previous works and of an important concept:

Given that InSAR is a differential method, all the measurements produced are “relative” and not “absolute”, both in space and time [A]. This is because each measurement is referred temporally and spatially to respectively a unique reference image (“Master” image) and to a reference point arbitrarily supposed to be motionless. The master scene is chosen in order to maximize the total coherence of the interferometric stack and to keep the dispersion of the normal baseline values as low as possible.

For these reasons, the InSAR technique is not an ideally stand-alone technique for studying larger areas involved in phenomena such as subsidence and uplift zones, but it requires comparison and calibration with another kind of information such as GNSS. Moreover, an additional drawback of InSAR is that it cannot recognize constant movements involving ground reference points, such as crustal movements that occur over large areas, and the accuracies of low wavenumber components of the velocity field decrease with increasing distance from them [B].

[A] Farolfi, G., Del Soldato, M., Bianchini, S., & Casagli, N. (2019). A procedure to use GNSS data to calibrate satellite PSI data for the study of subsidence: An example from the north-western Adriatic coast (Italy). European Journal of Remote Sensing, 52(sup4), 54-63.

[B] Colesanti, C., Ferretti, A., Prati, C., and Rocca, F.  (2003). Monitoring landslides and tectonic motions with the Permanent Scatterers Technique. Engineering Geology, 68, 3-14.

Response1: Thanks so much for your advices. Done. Line 58-68.

Other suggested modifications:

In all the maps, replace with a bigger font size the coordinates around the frame and the legend. Remove or make less visible the label reported at the bottom of each map.
Put the legend of the color inside each map, I suggest with a line bar.

Ponits2 Figure 1. In figures A and B, replace with a bigger font size the coordinates around the frame and the legend. The map of China as well inside figure 1a is not readable.   Put a white background at the labels inside the figure 1.

Response2: Thanks! Done, Figure 1a, 1b.

Ponits3 Line 118-120. Are you meaning images for scenes?

Response3:Thanks! Yes, it describes the amount of images used in data processing.
Ponits4 Did you use only 3 ENVISAT images in the processing?

Response4:Thanks! Yes, it is. Because the other Envisat images could not maintain good coherence.
Ponits5 How many ascending and descending mages did you process?

Response5:Thanks! All of the images are ascending images.

Ponits6 Line 127. Please, provide a figure with a focus on leveling benchmarks. Is it a high-resolution leveling network? Please, draw the path or the graph of the leveling network.

Response6:Thanks! I am so sorry to report that the Wuhan Institute of Geomatics just provided the location and relative results of leveling benchmarks. The path/graph of the leveling network was not available.
Ponits7 line 193-194 Please, insert the formula after the phrase: all of the results are converted from line-of-sight into vertical [38,39].

Response7:Thanks! Done. Line 209-210.

Ponits8 Figure 3. Increase the thickness of the lines.
Response8:Thanks! Done. Figure 3.
Ponits9 Figure 4c. Make of a different color and increase the thickness of the histogram line.
Response9:Thanks! Done. Figure 4c.

Ponits10 Figure 6. Reorganize all the figures. Put the three maps on the same line and the picture in a different figure also in one line.

Response10:Thanks! Done. Figure 6.
Ponits11  Figure 7. Increase the font of the cartesian information and the thickness of the points.

Response11:Thanks! Done. Figure 7.
Ponits12 Figure 8. Reorganize all the figures. Put the three maps on the same line and the color legend inside the maps.
Response12:Thanks! Done. Figure 8.
Ponits13 Figure 9. Increase the font of the color legend inside the maps.

Response13:Thanks! Done. Figure 9.
Ponits14 Figure 11. Move the pictures c and f to the left side.
Response14:Thanks! Done. Figure 11.
Ponits15 The last correction regards the unit of measure of the velocity of displacements.
In astronomy, geology, and paleontology, the abbreviation yr for years and ya for years ago are sometimes used, combined with prefixes for thousand, million, or billion, but they are not SI units.
The SI unit for the year is not mm/yr, but mm/a where ‘a’ is annum in the SI units  (per annum means "each year").

Response15:Thanks! Done. I have revised all units to the SI units.

Please see the attachment

Author Response File: Author Response.docx

Reviewer 4 Report

This paper provides an overview of the sinkholes and subsidence activity over the city of Wuhan in China, The paper fits the topic of the journal and it’s well organized despite minor errors with the adopted English grammar. This is acceptable given the authors are not native speakers and I would recommend a native speaker to revise the manuscript before in case of acceptance.

I realize this is the second round of review and I appreciate the effort of the authors in addressing previous reviewers’ comments. I made some comments below the authors should address before acceptance.

In my opinion, there is only little advancement provided by this paper compared to previous studies (The main novelty is the updated sentinel analysis up to 2019 which none of the other studies presents). I would have been nice to see what is the effect of the Coronavirus Pandemic on the subsidence level related to causes connected with human activities ( ex. Groundwater exploitation)

Given the advanced status of the review process, I only have one main concern related to the introduction of the paper where authors do not clearly highlight existing literature related to similar studies on tunneling, groundwater activities etc. This, in my opinion, does not allow the reader to understand what the advancements is proposed by the authors in this research paper. I recommend hence to update the introduction providing some background of existing studies. These are both old well-known studies and recent studies that attracted some interest in the scientific community. As an example:

Giardina, G., Milillo, P., DeJong, M. J., Perissin, D., & Milillo, G. (2019). Evaluation of InSAR monitoring data for post-tunneling settlement damage assessment. Structural Control and Health Monitoring, 26(2), e2285.

Baer, G., Magen, Y., Nof, R. N., Raz, E., Lyakhovsky, V., & Shalev, E. (2018). InSAR measurements and viscoelastic modeling of sinkhole precursory subsidence: Implications for sinkhole formation, early warning, and sediment properties. Journal of Geophysical Research: Earth Surface, 123(4), 678-693.

More comments below:

44 – you mean sinkholes?

61-63 wrong citation reference [17-26] (cite quasi PS and squeesar)

Perissin, D., & Wang, T. (2011). Repeat-pass SAR interferometry with partially coherent targets. IEEE Transactions on Geoscience and Remote Sensing, 50(1), 271-280.

Ferretti, A., Fumagalli, A., Novali, F., Prati, C., Rocca, F., & Rucci, A. (2011). A new algorithm for processing interferometric data-stacks: SqueeSAR. IEEE transactions on geoscience and remote sensing, 49(9), 3460-3470.

80-81 Please refer to table 1

Figure 2 Did you decompose in vertical vs horizontal displacements? If these are Line of sight measurements please refer in the figure caption as movements away (negative values) or toward (positive values) the sensor  

Figure 5 please list the standard deviation of the InSAR vs Leveling data

Figure 6 Profile A-A’ please make the letters larger since it’s barely visible. Also do you average values along the squared profile?

Figure 7 please speceify this is related to the profile A-A’ in figure 6

Figure 9 what is the buffer that you used for the metro lines?

Author Response

Response:

Thank you very much for the insightful comments on this manuscript! In the revised manuscript, we tried our best to improve the manuscript based on your constructive suggestions. Please see the specific responses below. The manuscript has been proofread by a native English speaker.

points1 In my opinion there is only little advancement provided by this paper compared to previous studies (The main novelty is the updated sentinel analysis up to 2019 which none of the other studies presents). I would have been nice to see what is the effect of the Coronavirus Pandemic on the subsidence level related to causes connected with human activities ( ex. Groundwater exploitation)

Response1: Thanks very much for your suggestions. Compared with the previous studies, we used both ALOS-1 and Envisat images to estimate the land subsidence in Wuhan during 2007-2010, and these images span a longer time interval. Sentinel-1A/B images we used in this study span 4 years. In contrast, the previous publication only used 1-year of Sentinel-1A/B observations. We have utilized the changes in SAR backscattering intensity images to infer new building constructions and discussed their impact on land subsidence. Finally, we discussed all of major factors affecting land subsidence in Wuhan. Hence, our paper has improved previous studies and our current knowledge on understanding land subsidence in Wuhan. Your suggestion about the effect of the Coronavirus Pandemic on the subsidence in Wuhan is a very interesting and intriguing, which will be subject to near-future research.

points2 Given the advanced status of the review process, I only have one main concern related to the introduction of the paper where authors do not clearly highlight existing literature related to similar studies on tunneling, groundwater activities etc. This, in my opinion does not allow the reader to understand what the advancements is proposed by the authors in this research paper. I recommend hence to update the introduction providing some background of existing studies. These are both old well known studies and recent studies that attracted some interest in the scientific community. As an example:

Giardina, G., Milillo, P., DeJong, M. J., Perissin, D., & Milillo, G. (2019). Evaluation of InSAR monitoring data for post‐tunnelling settlement damage assessment. Structural Control and Health Monitoring, 26(2), e2285.

Baer, G., Magen, Y., Nof, R. N., Raz, E., Lyakhovsky, V., & Shalev, E. (2018). InSAR measurements and viscoelastic modeling of sinkhole precursory subsidence: Implications for sinkhole formation, early warning, and sediment properties. Journal of Geophysical Research: Earth Surface, 123(4), 678-693.

Response2:Thanks! Your suggestion is very helpful. Done. Line 57, Reference [16-17].

More comments below:

points3 44 – you mean sinkholes?

Response3: Thanks!  There are 29 sinkholes recorded from 1977 to 2014.

points4 61-63 wrong citation reference [17-26] (cite quasi PS and squeesar)

Perissin, D., & Wang, T. (2011). Repeat-pass SAR interferometry with partially coherent targets. IEEE Transactions on Geoscience and Remote Sensing, 50(1), 271-280.

Ferretti, A., Fumagalli, A., Novali, F., Prati, C., Rocca, F., & Rucci, A. (2011). A new algorithm for processing interferometric data-stacks: SqueeSAR. IEEE transactions on geoscience and remote sensing, 49(9), 3460-3470.

Response4: Thanks!  Done!  Line 74. Reference [23,30].

points5 80-81 Please refer to table 1

Response5: Thanks! Done! Line 92.

points6 Figure 2 Did you decompose in vertical vs horizontal displacements? If these are Line of sight measurements please refer in the figure caption as movements away (negative values) or toward (positive values) the sensor  

Response6: Thanks! I had decomposed the deformation into vertical direction. Line 209-210.

points7 Figure 5 please list the standard deviation of the InSAR vs Leveling data

Response7: Thanks! Done. Figure 5.

points8 Figure 6 Profile A-A’ please make the letters larger since it’s barely visible. Also do you average values along the squared profile?

Response8:Thanks! Done. Figure 6. I have replaced the squared profile by bold line.

points9 Figure 7 please speceify this is related to the profile A-A’ in figure 6

Response91:Thanks! Done. Figure 7.

points10 Figure 9 what is the buffer that you used for the metro lines?

Response10:Thanks! The width of buffer is 1km used for metro lines.

Please see the attachment

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

The text has been improved on both directions, scientific and presentation.

All the comments were appropriate addressed by the authors.

The only comment concerns the reference point used in the interferometric analysis. In lines 214-216 refer to the use of continuous GNSS station, i think that it would be nice to present the velocity vector of this station(Veast, Vnorth, Vup). Especially the vertical component. Also, considering that this station dates back to 1998, i have the feeling that there is a quite reliable estimation of the vertical velocity field. So please add this information to your text and make a short comment.

Author Response

Response: Thank you very much for your careful read and insightful comment on this manuscript! In the revised manuscript, we tried our best to improve the manuscript based on your constructive suggestion.

note1 The only comment concerns the reference point used in the interferometric analysis. In lines 214-216 refer to the use of a continuous GNSS station, i think that it would be nice to present the velocity vector of this station (Veast, Vnorth, Vup). Especially the vertical component. Also, considering that this station dates back to 1998, i have the feeling that there is a quite reliable estimation of the vertical velocity field. So please add this information to your text and make a short comment.

Response1 Thank you so much for your valuable comments! We have obtained the time series of the vertical motion at the reference point (IGS station – WUHN) during 2007-2019 from the Nevada Geodetic Laboratory (http://geodesy.unr.edu/NGLStationPages/stations/WUHN.sta). The vertical velocity did not changed significantly during the InSAR monitoring period. We have added the GPS time-series plot to Figure 2 (Figure 2d), and added a brief explanation to the text (Line 216-219).

Please see the attachment.

Author Response File: Author Response.docx