Review Reports

Round 1

Reviewer 1 Report

I would recommend the authors to recheck the entire manuscript but mostly in the introduction while the name are systematically written to start a sentence. This is not necessary as there is a reference number attached. So better to change the beginning and delete the names

line 59 The PSI method used to monitor (instead of Ng et al.used PSI method to monitor)

idem or same idea at line 71/73/74/76/81/85/103/109/

Line 85 is "large" really appropriate, ? i di believe that "thicker" is better

Line 97 GA ?? model ? where is it defined before ??

Line 100 GRU abbreviation is never defined before ??

Figure 1: the colour of the two polygones in the legend are the same while they are different on the map

the scalebar and the text is really difficult to read on figure 1

in the caption legend of figure 1 - it is written that the background image is a remote sensing image from which satellite, indicate the origin, year etc of the imagery used

Line 146 - there is a range between 280-450 m of thickness while on the cross-section is it very homogeneous (>480 m) so can you explain where those differences are reported. this is very important as the more the thickness is reduced, the lower the land subsidence should be expected

Line 197 - add the word that (that were collected)

Line 198 add processed (were processed to obtain)

Line 211 - ok but i'm surprised for the two years of groundwater data used to monitor the land subsidence using the three wells ?? 2 years is really nothing takin into account the time interval studied with InSAR (2003-2020) How are you able to tackle the variations before that were observed using InSAR

Line 218 - add that (Considering that the ASAR)

Line 220 - SBAS methodology (add methodology)

Line 220 - move the reference number [6] after the name of the author

Line 231 - the coherence threshold of 0.4 - This is surprising and quite very low as such. How are you thus managing the quality of your points and values through time. Generally a coherence level above 0.7 is much better. On the counter part it will reduce the number of points but your ROI is in a urban area so this is not a big challenge

Line 301 - i know that the way it is written it is correct but generally the symbol yr is preferred instead of y

Figure 6 - May be good to add that the measurement given are the LOS displacement rate

the entire legend is very small for the reader - it should be good to size up the text for the reader

Line 326 - The area has a superficy of  443.4 km²

Line 336 - The contour lines .... 2019 were compared with ... (replace in the text accordingly)

Line 338. After (Figure 9). Put a dot and start a new sentence

Line 346 - Is the "groundwater level depth" representing the piezometric water levels ??

Line 426 - The remaining data

Line 436-437 - land subsidence mutation - what do you mean ?

Main remark, most of the clay layers and aquitards above the main aquifers will participate to the land subsidence. The authors have choosen to concentrate their approach using only the third one while aquitards level are also present below and they are not counted in the model.

The issue that is still not explained is the amount of groundwater pumped in the third aquifer versus the others ?? Overpumping rates are expected to account for the land subsidence but this is not really pushed forward neither in the data nor in the analysis. Let's say that one aquifer is mainly pumped is probably true but not proved although the second and fourth aquifer are also used and should also induces an effect on the compaction in the aquitards ? How can you tackle this point here ?

 

 

 

Author Response

We appreciate the constructive comments provided by the Reviewers. We have revised the manuscript carefully by incorporating these comments and suggestions. The following are our point-to-point responses to the review comments.

 

Point 1: I would recommend the authors to recheck the entire manuscript but mostly in the introduction while the name are systematically written to start a sentence. This is not necessary as there is a reference number attached. So better to change the beginning and delete the names

 

Response 1: Thank you for your comments, we have rechecked the whole manuscript and deleted the names when references number were attached.

 

Point 2: line 59 The PSI method used to monitor (instead of Ng et al.used PSI method to monitor)

 

Response 2: Thank you for your comments, we have revised it in the paper: The PSI method used to monitor the land subsidence in Jakarta, Indonesia with ALOS PALSAR images from 2007 to 2010. 

 

Point 3: idem or same idea at line 71/73/74/76/81/85/103/109/

 

Response 3: Thank you for your comments, we have checked the full text and revised it at Line 81-122 in the paper: A striking linear relationship between compaction rate and peat thickness on basis of ~250 boreholes found in Mississippi Delta [11]. It was found that subsidence is mainly caused by primary compaction of the Holocene strata in Po Delta [12]. Land subsidence was concentric around locations of intense groundwater withdrawal in Houston–Galveston, Gulf Coast region of the United States [13]. Subsidence with the value up to 140 mm/yr due to peat compaction and oxidation was quantified in built-up areas in the Rhine-Meuse delta in the Netherlands on basis of borehole data and dry bulk density, organic matter and CO₂ respiration [14]. Previous studies showed that the main causes of land subsidence in the Binhai New Area are intensive groundwater extraction, the natural consolidation of cohesive soil and building load [15]…

 

Point 4: Line 85 is "large" really appropriate, ? i di believe that "thicker" is better

 

Response 4: Thank you for your comments, we have replaced "large" with "thicker" in the paper.

 

Point 5: Line 97 GA ?? model ? where is it defined before ??

 

Response 5: Thank you for your comments, GA means Genetic Algorithm, we have given the description at Line 107: Artificial neural network model such as back propagation (BP) model, coupled Genetic Algorithm (GA) and BP model.

Point 6: Line 100 GRU abbreviation is never defined before ??

 

Response 6: Thank you for your comments. We have given the full description at line 109-113: Long Short-Term Memory (LSTM) and Gated Recurrent Unit (GRU) belongs to recurrent neural network, which can learn the characteristics of time series data, share parameters during the time domain using a cyclic structure (memory unit), record the context of the sequence and process long-term data with nonlinear characteristics.

Point 7: Figure 1: the colour of the two polygones in the legend are the same while they are different on the map

 

Response 7: Thank you for your comments, the colour of the two polygones in the legend are different, it may not be obvious, and it has been modified. The modified figure 1 can find at Response 8.

 

Point 8: the scalebar and the text is really difficult to read on figure 1

 

Response 8: Thank you for your comments, we have modified the scalebar and the text. Figure 1 after modification is as follows.

 

Point 9: in the caption legend of figure 1 - it is written that the background image is a remote sensing image from which satellite, indicate the origin, year etc of the imagery used

 

Response 9: Thank you for your comments, we have revised it in the caption legend of figure 1 : the background image is a Sentinel-2 image acquired on Sep. 16, 2020.

 

Point 10: Line 146 - there is a range between 280-450 m of thickness while on the cross-section is it very homogeneous (>480 m) so can you explain where those differences are reported. this is very important as the more the thickness is reduced, the lower the land subsidence should be expected

 

Response 10: Thank you for your comments, the Quaternary sediments thickness ranges from 280 to 450 m, the thickness of some areas is relatively small due to the existence of bedrock in the study area.

 

Point 11: Line 197 - add the word that (that were collected)

 

Response 11: Thank you for pointing this out, it has been added: and 66 descending TOPS-mode Sentinel-1 images with a revisit time of 12 days from 2015 to 2020 that were collected.

 

Point 12: Line 198 add processed (were processed to obtain)

 

Response 12: Thank you for pointing this out, it has been added : These images were processed to obtain the distribution of land subsidence by using InSAR technology.

 

Point 13: Line 211 - ok but i'm surprised for the two years of groundwater data used to monitor the land subsidence using the three wells ?? 2 years is really nothing takin into account the time interval studied with InSAR (2003-2020) How are you able to tackle the variations before that were observed using InSAR

 

Response 13: Thank you for your comments. There were few groundwater level data collected in this study. The discontinuous groundwater level monitoring data were deleted, and three representative groundwater wells monitoring data with obvious seasonal change of groundwater level were chosed to make the comparative study with InSAR data. In the future, government should construct more groundwater wells to monitor the long-term water level, which will be adopted to analyze the relations between the subsidence and groundwater variations.

 

Point 14: Line 218 - add that (Considering that the ASAR)

 

Response 14: Thank you for pointing this out, it has been added: Considering that the ASAR images number from 2003 to 2010 was limit.

 

Point 15: Line 220 - SBAS methodology (add methodology)

 

Response 15: Thank you for pointing this out, it has been added : SBAS methodology was proposed by Berardino et al..

 

Point 16: Line 220 - move the reference number [6] after the name of the author

 

Response 16: Thank you for pointing this out, the reference number [6] has been moved: SBAS methodology was proposed by Berardino et al. [6], which can guarantee high land subsidence monitoring accuracy with limited images.

 

Point 17: Line 231 - the coherence threshold of 0.4 - This is surprising and quite very low as such. How are you thus managing the quality of your points and values through time. Generally a coherence level above 0.7 is much better. On the counter part it will reduce the number of points but your ROI is in a urban area so this is not a big challenge

 

Response 17: Thank you for your comments, we have revised the reason at Line 245-248: the reason for setting the coherence threshold to 0.4 is that the time interval of Envisat ASAR images are too long. Envisat ASAR images are acquired by C-band sensors. Short wavelength and long time interval will lead to incoherence. After many experiments, the results obtained by setting the coherence threshold to 0.4 have been verified to be good, so we set it to 0.4.

 

Point 18: Line 301 - i know that the way it is written it is correct but generally the symbol yr is preferred instead of y

 

Response 18: Thank you for pointing this out, we have checked the full text and replaced “mm/y” with “mm/yr” in the paper.

 

Point 19: Figure 6 - May be good to add that the measurement given are the LOS displacement rate

the entire legend is very small for the reader - it should be good to size up the text for the reader

 

Response 19: Thank you for your comments, Figure 6 is the vertical direction displacement rate projected from the measurements in line of sight (LOS), and we explained it at Line 315-317: The vertical displacement is considered to be more appropriate for comparison with ground leveling measurements. Thus, the measurements in line of sight (LOS) are directly projected into the vertical direction.

The legend of figure 6 was modified, the modification figure 6 is as follows.

 

Point 20: Line 326 - The area has a superficy of  443.4 km²

 

Response 20: Thank you for your comments, it has been added at Line 344 : The area has a superficy of 443.4 km².

 

Point 21: Line 336 - The contour lines .... 2019 were compared with ... (replace in the text accordingly)

 

Response 21: Thank you for your comments, we have revised it at Line 354-356 : The contour lines of the groundwater level depth of the second and third confined aquifer groups in 2013, 2016 and 2019 were compared with the subsidence rate distribution from 2015 to 2020.

 

Point 22: Line 338. After (Figure 9). Put a dot and start a new sentence

 

Response 22: Thank you for your comments, we have put a dot and start a new sentence.

 

Point 23: Line 346 - Is the "groundwater level depth" representing the piezometric water levels ??

 

Response 23: Thank you for your comments, we have explained it at Line 221: "groundwater level depth" refers to the buried depth of groundwater and it is the distance from groundwater to the surface.

 

Point 24: Line 426 - The remaining data

 

Response 24: Thank you for your comments, we have revised it at Line 451 : The remaining data from September 2019 to December 2019 were used to validate the subsidence model.

 

Point 25: Line 436-437 - land subsidence mutation - what do you mean ?

 

Response 25: Thank you for your comments, “land subsidence mutation” means that the land subsidence changes greatly in a short time, we have revised it at Line 460-461: The maximum error occurred at well #1 in September 2019 with the reason of land sub-sidence changed greatly in September and October 2019.

 

Point 26: Main remark, most of the clay layers and aquitards above the main aquifers will participate to the land subsidence. The authors have choosen to concentrate their approach using only the third one while aquitards level are also present below and they are not counted in the model.

 

Response 26: Thank you for your comments. There are four aquifer groups which are composed with sandy sediments and interbeded with clay or silt. There is no distinct boundary of aquifer or aquitard. The groundwater level of each aquifer group is the same. The main pumping layer is the third aquifer group. The analysis in Section 5.1.1 showed that the distribution of land subsidence is more similar to the groundwater level distribution in the third confined aquifer group than other aquifer groups. The groundwater level data in the third confined aquifer group was used as the input data of the model as well as the cumulative thickness of the compressible layers, which calculated on basis of the collected borehole with around ~300 m. In future work, more data will be collected to make the model more reliable.

 

Point 27: The issue that is still not explained is the amount of groundwater pumped in the third aquifer versus the others ?? Overpumping rates are expected to account for the land subsidence but this is not really pushed forward neither in the data nor in the analysis. Let's say that one aquifer is mainly pumped is probably true but not proved although the second and fourth aquifer are also used and should also induces an effect on the compaction in the aquitards ? How can you tackle this point here ?

Response 27: Thank you for your comments. Only the pumping rate in 2002 were collected, which showed that the third aquifer group was 3.5×10⁷ m³/a. The number was 1.5~1.8×10⁷ m³/a larger than the second and forth aquifer groups, respectively. Considering the groundwater level can reflect the pumping volume, we chose the groundwater level of the third aquifer group as one of the input data of the model. We have added the pumping volume at Line 171-173 in Section 2. The analysis in Section 5.1.1 showed that the distribution of land subsidence is more similar to the groundwater level distribution in the third confined aquifer group than other aquifer groups. The cumulative thickness of the compressible layers, which calculated on basis of the collected borehole with around ~300 m. In future research, more groundwater level in different aquifer groups and deeper boreholes will be collected.

Author Response File: Author Response.pdf

Reviewer 2 Report

 jmse-1948761-peer-review-v1

 

The manuscript entitled “Land subsidence evolution and simulation in the western coastal area of Bohai Bay, China” addresses an interesting and up-to-date subject, which adheres to JMSE journal policies.

 

The manuscript contains genuine work, and presents a good InSAR application and case study related to a multiannual monitoring of land subsidence. The manuscript is also well written and has the publication potential after some medium improvements that are required:

 

-        The most lacking aspect of the manuscript is the integration of the levelling data. That aspect should be extensively improved, as that precise geodetic type of measurements are the validation (also the RMSE) of the results obtained by InSAR techniques. Did your team measured that data, or were them provided by authorities or other researchers? What kind of instrumentation were there used, as well as the type of measurement (maybe precision geometric levelling?). How was it ensured that the levelling benchmarks were stable and not affected by the subsidence (if they are displaced, it can invalidate or erroneous the data)?

-        At R51, you should change “GPS” to DGPS or GNSS

 

-        You should mention if there there any limitations of the InSAR technique, as all methods have certain advantages and disadvantages

Author Response

We appreciate the constructive comments. We have revised the manuscript carefully by incorporating these comments and suggestions. The following are our point-to-point responses to the review comments.

 

The manuscript entitled “Land subsidence evolution and simulation in the western coastal area of Bohai Bay, China” addresses an interesting and up-to-date subject, which adheres to JMSE journal policies.

 

The manuscript contains genuine work, and presents a good InSAR application and case study related to a multiannual monitoring of land subsidence. The manuscript is also well written and has the publication potential after some medium improvements that are required:

 

Point 1: -        The most lacking aspect of the manuscript is the integration of the levelling data. That aspect should be extensively improved, as that precise geodetic type of measurements are the validation (also the RMSE) of the results obtained by InSAR techniques. Did your team measured that data, or were them provided by authorities or other researchers? What kind of instrumentation were there used, as well as the type of measurement (maybe precision geometric levelling?). How was it ensured that the levelling benchmarks were stable and not affected by the subsidence (if they are displaced, it can invalidate or erroneous the data)?

 

Response 1: Thank you for your comments, we have revised it in the manuscript, the information was given at Line 217-218 in Section 3.2.2 : the levelling data were provided by authorities, and these data were obtained from precision geometric levelling.

 

Point 2: -        At R51, you should change “GPS” to DGPS or GNSS

 

Response 2: Thank you for your comments, we have replaced “GPS” with “GNSS”.

 

Point 3: -        You should mention if there there any limitations of the InSAR technique, as all methods have certain advantages and disadvantages

 

Response 3: Thank you for your comments, we have revised it at Line 53-62 in Section 1: The InSAR technique uses microwave radar signals to illuminate the ground surface and then records the amplitude and phase of the signals backscattered from the surface. Comparing the change in the phase enables the displacement of the ground surface to-wards or away from the sensor to be determined. The precision of the phase observations is often expressed in terms of the coherence between the two SAR images, with values close to one indicating high-quality measurements. The radar wavelength of the SAR instrument, atmospheric conditions during each image acquisition and land cover can affect the interferometric coherence. Loss of coherence can limit the coastal areas for which conventional InSAR can provide useful data.

Author Response File: Author Response.pdf

Reviewer 3 Report

This paper analyses the land subsidence due to groundwater overexploitation and its affection to buildings using remote sensing data from various sources along a coastal strip in China. In addition, the authors created a neural network model to simulate subsidence, with a high correlation between measured data and simulations.

The manuscript is clear, relevant for the field and presented in a well-structured manner. The cited references are mostly recent publications and relevant. The methods section is well explained and results are adequately presented. Figures are of quality. Conclusions are consistent with the evidence and arguments presented in the manuscript.

Overall, I consider it a novel and interesting research. Hence, I recommend to Accept in present form.

Author Response

Thank for your comments, we have checked the spell of the manuscript and further improved the paper.

Round 2

Reviewer 2 Report

 

The revised manuscript demonstrates the author’s commitment in improving the overall paper, thus obtaining a cohesive and interesting article.

In my opinion, the manuscript should be considered for publication.