Analysis of the Contribution Rate of the Influencing Factors to Land Subsidence in the Eastern Beijing Plain, China Based on Extremely Randomized Trees (ERT) Method
Round 1
Reviewer 1 Report
The paper is ready for publication.
Author Response
- The paper is ready for publication.
Response: Thanks for the comments of the reviewer.
Reviewer 2 Report
I am referring to the original review report, since this is a resubmission.
_ "comprehensive analysis of the contribution rate to land subsidence in terms of its influencing factors combining InSAR method and Extremely Randomized Trees (ERT) machine learning approaches have been rarely published, which is the main the novelty of the paper." I don't see this clearly stated in the introduction.
_ "Hydrogeological data such as ground water level in different aquifers after 2015 are not available due to the Beijing Institute of Hydrogeology and Engineering Geology do not published the relevant data to the society after 2015." Add something about this in the section 2.2 (or were do you think it fits better) to motivate the choice of your radar images.
_ "We describe the hydrogeology asset including groundwater level in different aquifers and quaternary thickness of compressible layer of the study area in more detail" First, why is this included in 2.2 and not in the study area section? Second, I don't see any "detail" in this description. A description of the stratigraphic and hydrogeological context includes a groundwater level map, cross sections, etc... I think that the authors are perfectly able to find a reference cross section for the hydrogeological asset of the city, considering the huge amount of literature available.
_line 188 time is incorrect, it is temporal baseline
_line 193 "The spatial filters were used to eliminate the phase noise". Which spatial filters?
_any reference for equation 1? I still believe that it is not correct to estimate the components with one singl orbit, but a reference can strenghten the text.
Author Response
- The paper is ready for publication. I am referring to the original review report, since this is a resubmission.
_ "comprehensive analysis of the contribution rate to land subsidence in terms of its influencing factors combining InSAR method and Extremely Randomized Trees (ERT) machine learning approaches have been rarely published, which is the main the novelty of the paper." I don't see this clearly stated in the introduction.
Response: Previous studies mainly focused on the contribution rate of influencing factors to the cumulative land subsidence, such as reference [11] and [21]. However, quantitative analysis of the contribution rate of the influencing factors on uneven land subsidence has not been recorded. As an efficient ensemble learning method, the Extremely Randomized Trees (ERT) method can obtain more reliable results by using all of the sample sets in the training process. Therefore, comprehensive analysis of the contribution rate of the influencing factors to cumulative and uneven land subsidence is the main novelty of the paper.
- _ "Hydrogeological data such as ground water level in different aquifers after 2015 are not available due to the Beijing Institute of Hydrogeology and Engineering Geology do not published the relevant data to the society after 2015." Add something about this in the section 2.2 (or were do you think it fits better) to motivate the choice of your radar images.
Response: We add “Ground water level in different aquifers after 2015 are not available due to the Beijing Institute of Hydrogeology and Engineering Geology do not published the relevant data to the society after 2015” in section 2.2 and highlighted in red in Line 166-168.
- 3. _ "We describe the hydrogeology asset including groundwater level in different aquifers and quaternary thickness of compressible layer of the study area in more detail" First, why is this included in 2.2 and not in the study area section? Second, I don't see any "detail" in this description. A description of the stratigraphic and hydrogeological context includes a groundwater level map, cross sections, etc... I think that the authors are perfectly able to find a reference cross section for the hydrogeological asset of the city, considering the huge amount of literature available.
Response: We describe the hydrogeology asset (groundwater level in different aquifers and quaternary thickness of compressible layer of the study area) in section 2.2 for that all the auxiliary data are described in this section 2.2. The reference cross section for the hydrogeological asset of the Beijing city is WGS84 geographical coordinate system.
- 4. _line 188 time is incorrect, it is temporal baseline.
Response: We have rephrase “time” to “temporal” and highlighted in red in Line 188 and 190.
- _line 193 "The spatial filters were used to eliminate the phase noise". Which spatial filters?
Response: This spatial filters refers to spatial low-pass filter which has been integrated in GAMMA software.
- _any reference for equation 1? I still believe that it is not correct to estimate the components with one singl orbit, but a reference can strenghten the text.
Response: Reference [21] shows that the horizontal deformation in the Beijing Plain can be neglected and equation 1 can be found in Reference [21].
Reviewer 3 Report
add coordinates to the map
Author Response
- add coordinates to the map.
Response: We have added coordinates in Figure 1. Other graphs are all about the study area, which coordinates can be found in Figure 1.
Reviewer 4 Report
In reviewer's opinion the Authors do not pay attention to the precision of their statements. The authors use terms that are not identical interchangeably. An engineer should be precise. This can be seen in the Methodology part of the text.
example 1:
The term "deformation" means several indicators such as subsidence, slope change, horizontal displacement, compression / tension, curvature.
The authors designate terrain subsidence , not deformations.
example 2:
land subsidence is not the same as deformation velocity - equation 1
Despite these inaccuracies in the initial part of the text, I believe that the article is very interesting, correct in terms of content and of great practical importance.
I believe that after the amendments to the Methodology chapter it can be published.
Author Response
- In reviewer's opinion the Authors do not pay attention to the precision of their statements. The authors use terms that are not identical interchangeably. An engineer should be precise. This can be seen in the Methodology part of the text.
example 1:
The term "deformation" means several indicators such as subsidence, slope change, horizontal displacement, compression / tension, curvature.
The authors designate terrain subsidence , not deformations.
example 2:
land subsidence is not the same as deformation velocity - equation 1
Despite these inaccuracies in the initial part of the text, I believe that the article is very interesting, correct in terms of content and of great practical importance.
I believe that after the amendments to the Methodology chapter it can be published.
Response: The deformations in this paper means the ground deformation information along the line-of-sight (LOS) direction. The land subsidence is deformation information along the vertical direction. We changed some statements in the paper and highlighted the changed parts in red.
This manuscript is a resubmission of an earlier submission. The following is a list of the peer review reports and author responses from that submission.
Round 1
Reviewer 1 Report
The paper analyzes subsidence in the Beijing plain, one of the most common targets for InSAR. Thus, the first comments regards the novelty of the paper that is, to me, not evident (considering what it has been published before). This includes the choice of ENVISAT and RSAT data and not of new constellations.
Secondly, the quality of the writing must be improved. The English editing strongly needs an improvement.
Considering that the target is the discrimination of the contribution of different geological and hydrogeological factors, the description of the local hydrgogeological asset is poor.
The same comments is valid for the insar methodology. The SBAS approach used here is not described and the vertical reprojection is 1) not motivated and 2) based on a wrong formula which is not considering the doble orbit configuration and the direction cosines.
The results section could be interesting but it should be reorganized and rethinked. For example, the role of the 2nd and 3rd acquifers is central and critical. This should be the focus.
The dynamic analysis requires to be detailed. The source of the "numbers" used must be explained and motivated.
The description of the two statistical approaches used to verify the contribution of the subsidence factors must be improved.
There is not a real discussion of the results obtained.
Reviewer 2 Report
Good paper and suitable for publication
- What is the main question addressed by the research? Is it relevant and interesting?
Reply. The authors use two techniques for evaluating the contribution rate of the influencing factors to land subsidence: the Spearman's rank correlation coefficient and Extremely Randomized Trees (ERT) machine learning approaches. The focus of the paper is relevant, within the scope of the journal and very interesting.
- How original is the topic? What does it add to the subject area compared with other published material?
Reply. Although the Land subsidence is well known and broadly discussed in the literature, the proposed approaches especially the ERT machine learning model is new and not reported itself, in addition the authors employed a series of ENVISAT ASAR (EA) images and RADARSAT-2 (R2) images, and validate with leveling benchmark monitoring results, which is very interesting and help for providing a robust conclusions.
- Is the paper well written? Is the text clear and easy to read?
Reply. The paper is well written and well structured, clear and easy to read.
- Are the conclusions consistent with the evidence and arguments presented? Do they address the main question posed?
Reply. The conclusion is consistent and summarize what is really obtained from the investigation.
Reviewer 3 Report
The authors presented a deformation study of the Easter Beijing Plain.
I thank the authors for their submission, I think the manuscript has a scientific value and potentially a broad readership. However, some changes would improve it. The use of English is occasionally poor. Some of the methods need a better description.
I invite the authors to consider the comments bellow and provide a revised version of their work.
-Line 42-43: “especially for uneven land subsidence” you mean “especially in the case of spatially uneven land subsidence”?. You could consider also using the term “differential settlement” in the manuscrip.
-Line 53: Delete the term “Unscientific”. You can just use: “underground water overexploitation” or “irrational exploitation of the aquifer “ or “The lack of a water management plan” or something similar.
-Line 53:Better adopt: “..has led to a dramatic ground water level decrease..”
-Line 61: “As a new ground observation technique, The InSAR technique can acquire land subsidence information with millimeter accuracy”.
There can be a millimetre accuracy because it is a new technique? Please rephrase. Also correct the capital T at “The InSAR”.
-Line 65: “by recognize pixels” correct “by identifying pixels”
-Line 66: “stable scattering characteristics” change to “stable scattering characteristics over time”
-Line 70: BP? Better declare the abbreviation in a parenthesis like: Beijing plane (BP) in also your maintext. While you mention abbreviations on the Abstract, I would advice you to mention them also in the maintext the first time you use it. I believe this would make it easier for the reader to follow you.
-Line 79-80: “However, quantitative analysis on contribution rate of its influencing factors to uneven land subsidence have been rarely reported”. Please rephrase, use better syntax and grammar. Consider having a detailed revision of the use of English of your manuscript.
-Line 127: “was obtained based on EA data and R2 data”. Again for Envisat-Asar and Radarsat, provide explanation for the abbreviations in the maintext. Some readers might not be InSAR experts, or they might not be familiar at all with the various satellite missions.
-Line 129: “proposed by Boussinesq”. Provide a reference for the formula proposed by Boussinesq.
-Section 3.1. This is kind of a short description of the method used. More information on the basis of this study, which is the processing of the SAR data, would be needed. Please provide information about processing parameters, unwrapping method and parameters, unwrapping threshold. Did you adopted the same parameters for both the ENVISAT and Radarsat datasets?
-Provide the SBAS SLC connection graph in the maintext.
-There is no data Table. Provide two tables (one for Envisat and one for Radarsat) showing the data images that you used. You can put them in the maintext or in the supplementary material.
-Line 247: “pixel/km2”. Correct the exponent notation.
-Line 252: Avoid expressions like “extremely obvious”
-Line 255-256: “…land subsidence has experienced a rapid expansion period from 2004 to 2007.” What would you like to express. Please rephrase.
-Line 261-262: “Land subsidence developed rapidly from 2008 to 2010 with the maximum land subsidence rate and area with land subsidence rate greater than 100 mm/year increased…”. Please rephrase and improve the expression.
-Figure 6: Is the InSAR rate in LOS or the vertical? If it’s in LOS, did you make any projections of the leveling results? If it is not in LOS, and it is the vertical component please mention it on the graph.
