Pre-Event Deformation and Failure Mechanism Analysis of the Pusa Landslide, China with Multi-Sensor SAR Imagery

Round 1

Reviewer 1 Report

The Authors by using a multi-sensor SAR imagery analyzed the triggering factors and the failure mechanism of the Pusa landslide (China). Following some of my concerns about the manuscript:

1. Introduction section and in other parts of the text body: the landslide is defined as "karst landslide". Karst-landslide deformation generally manifests itself due to karst processes. As reported by Parise (2008) “rock failures in karst environments present peculiar features related to the typical setting of karst, characterized by presence of caves created by chemical solution of soluble rocks, direct connection between the surface and the subsurface, and an overall high fragility of the environment”. Due the complexity of the Pusa landslide and considering that the main triggering factor is the mining activity due to the land mismanagement I suggest to classify the landslide using the referring classifications. In this framework, I suggest the classification proposed by Fan et al. (2018) who classifies the Pusa landslide as "rock avalanche".

- Parise, M. (2008). Rock failures in karst. Landslides and engineered slopes, 1, 275-280.

2. Geological setting: the geological setting of the Pusa landslide is not clear or hard to understand. No geological map is presented and the longitudinal geological section (Fig. 3) has no legend. The acronyms used in the figure are explained in the text body. Being the section similar to that presented by Fan et al. (2018), symbols, colours and geological fill patterns have to be explained in the figure caption.

3. The novelty of the manuscript compared to Fan et al. (2018) is the checking of pre- and post-deformations by SBAS-InSAR technique. The discussion section should be used to describe the reliability of the detected cumulative deformation instead of the already well-known failure mechanisms and triggering factors.

Concluding, the manuscript describes the well-known triggering factors and failure mechanisms of the Pusa landslide. The opinion of the Reviewer is that the manuscript needs to be improved: geological properties has to be better explained and the discussion section has to be focused on the reliability of deformations obtained by SBAS-InSAR technique instead of the well-known triggering factors and failure mechanisms of the landslide.

 

Author Response

Comments and Suggestions for Authors:

The Authors by using a multi-sensor SAR imagery analyzed the triggering factors and the failure mechanism of the Pusa landslide (China). Following some of my concerns about the manuscript:

 

1. Introduction section and in other parts of the text body: the landslide is defined as "karst landslide". Karst-landslide deformation generally manifests itself due to karst processes. As reported by Parise (2008) “rock failures in karst environments present peculiar features related to the typical setting of karst, characterized by presence of caves created by chemical solution of soluble rocks, direct connection between the surface and the subsurface, and an overall high fragility of the environment”. Due the complexity of the Pusa landslide and considering that the main triggering factor is the mining activity due to the land mismanagement I suggest to classify the landslide using the referring classifications. In this framework, I suggest the classification proposed by Fan et al. (2018) who classifies the Pusa landslide as "rock avalanche".

- Parise, M. (2008). Rock failures in karst. Landslides and engineered slopes, 1, 275-280.

>> Many thanks for your suggestions. Pusa landslide is located in the bare karst region in China composed of carbonate rocks, while this landslide is commonly classified as rock avalanche due to its long runout characteristics or extremely rapid landslide based on the classification by Cruden and Varnes. Accordingly, we revised it as rock avalanche through the manuscript.

 

2. Geological setting: the geological setting of the Pusa landslide is not clear or hard to understand. No geological map is presented and the longitudinal geological section (Figure 3) has no legend. The acronyms used in the figure are explained in the text body. Being the section similar to that presented by Fan et al. (2018), symbols, colours and geological fill patterns have to be explained in the figure caption.

>> Thanks for your suggestions, we have revised the figure to add legend and the explanation in the caption. Please see Figure 3 and the caption as follows:

“The geological profile of the Pusa landslide along line L-L’ in Figure 2a. 1 Quaternary deposits; 21-Sandstone (1); 3 Mudstone; 4 Silty_mudstone; 5 Limestone; 6 Coal seam; 7 Fault; 8 Avalanche accumulation; 9 Source area; 10 Yelang Formation of the Lower Triassic; 11 Changxing-Dalong Formation of the upper Permian; 12 Longtan Formation of the upper Permian; 13 Coal mine tunnel.”

3. The novelty of the manuscript compared to Fan et al. (2018) is the checking of pre- and post-deformations by SBAS-InSAR technique. The discussion section should be used to describe the reliability of the detected cumulative deformation instead of the already well-known failure mechanisms and triggering factors.

>> Thanks for your suggestions and comments. We have revised the discussion section to emphasize the reliability of the InSAR results. Besides, we want to keep the discussion on mechanism and triggering factors, as it can be manifested by the InSAR measurements. Please see lines 321-330

“As no in-situ measurements one year before the failure of Pusa rock avalanche, pre-event InSAR deformation time series can hardly assess quantitatively with external measurements. Even so, the InSAR deformation field can be verified by the stacking interferograms with different modes of ALOS/PALSAR-2 images shown in Figure 7, where the deformed regions are consistent. Moreover, the deformation rate can be visually compared with ascending and descending Sentinel-1A images shown in Figure 8, where, green points indicate stable region in ascending and descending results. And deformed regions can be verified by Google earth image shown in Figure 2(a) before the event. Further, the deformation time series can be validated by ascending and descending results at P2 shown in Figure 10 (b) and (d), where the main deformation trends are consistent and the difference is resulted from different SAR geometry.”

Concluding, the manuscript describes the well-known triggering factors and failure mechanisms of the Pusa landslide. The opinion of the Reviewer is that the manuscript needs to be improved: geological properties has to be better explained and the discussion section has to be focused on the reliability of deformations obtained by SBAS-InSAR technique instead of the well-known triggering factors and failure mechanisms of the landslide.

>> Many thanks for your comments.

Reviewer 2 Report

The paper aims to develop a methodological approach for studying precursory signals and sliding mechanisms of a large slope instability through the use of high-resoluted temporal SAR data.

The paper is very interesting and well presented. The methodological approach is reliable and could allow to be a good tool also for early-warning strategy in similar contexts. Instead, some aspects have to be clarified better.

Suggested revisions follow:

• Clarifying the meaning of karst landslides
• Referrring only to works on SAR application for landslides studies in Chinese area is too limited. There is a huge amount of works in European, American and different Asian contexts. Please, improve these references
• Authors have to improved significantly the description of the slope instability, in particular with a more detailed description of: i) affected soil and bedrock materials; ii) depth of the sliding surface; landslides typology, according to Cruden and Varnes’ classification; iv) geomorphological features of the slope affected by the landslide
• Add a table with the features of all the satellite images used for SAR analysis
• A flow-chart showing the different phases of the methodological approach could help in the comprehension of the paper
• To verify if V_LOS is reliable in the estimation of the real kinematic behavior of the landslide, add a map of visibility of satellites (e.g C-Index), in order to calculate the percentage of the movement which could be detected by the satellites, in both the geometries of acquisition
• Justify the choice of the three selected points as representative of the different sliding mechanisms across the slope instability
• The relation between displacements and rainfall amounts seems clear. Maybe, it could be also possible to calculate the cumulated rainfall amounts in days (e.g. 3-days, 5-days, 10-days?) that fit the best with the measured rate of displacement. This is fundamental also for early warning strategy in similar settings
• Figure 11c: the legend of the colors of the circe is necessary
• Discussions should be improved, focusing on: i) similarities and differences of the sliding mechanism with slope instabilities occuured for similar predisposing and triggering factors; ii) potential application of the methodological approach for land planning and/or early warning strategies

Author Response

Please see the enclosed file.

Author Response File: Author Response.docx

Reviewer 3 Report

Overview: The authors utilize various aspects of InSAR (coherence, interferograms, and SBAS time series) to calculate and monitor various aspects of a catastrophic Pusa landslide that occurred in China on 28 August 2017. Coherence maps are used to detect changes in surface reflectors (vegetation vs bare soil, before and after the event). Interferograms are used to delineate landslide boundary. SBAS is used for pre-landslide deformation detection and assignment of possible causes as triggering mechanisms (rainfall and mining). The authors make a good case for the feasibility of this approach for landslide detection.

 

General Comments:

I selected "Extensive editing of English language and style are required" because there are many sentences that are worded awkwardly or are not complete sentences. For example, these sentences in lines 41-44: "The missing of early identification, long-term deformation monitoring and accurate prediction of landslides results in the unavoidable catastrophic disaster. Therefore, the diagnosis of the key triggering factors of karst landslides plays a key role in reducing the loss of such landslide effectively." In the first sentence, the beginning "The missing..." and ending "...the unavoidable catastrophic disaster" both sound odd. In the second sentence, the final word "effectively" is out of place within the sentence. There are many examples of grammatical oddities throughout the publication and the authors should comb through the text and rework most of the sentences.

Figure 1. What is the inset within the inset (the small rectangle of lines within Fig1b)? Also, the text for the word "Yuunan(?)" is difficult to read in grey.

Conclusions. I would be interested if the authors elaborated on their final (one-sentence) paragraph with additional avenues for research, whether at this site or elsewhere using SBAS InSAR, coherence, etc. for landslide prediction/detection prior to failure. It feels like the authors are heading in this direction without ever saying it.

Line-by-Line Comments:

Lines 19, 280, and 333: Change "researches" to "research"

Lines 22-23: Please define SAR (synthetic aperture radar) and LOS (line-of-sight) for the readers.

Lines 121-122: What does this sentence mean? "Now, the coherence magnitude (gamma) is the most common measure of coherence with the range of [0,1]." 

Lines 122-123: The sentence "It [coherence] is close to 1 in bare surface area, but close to 0 in serious vegetation area" is not necessarily true. If the bare surface area has a geometry where all radar signal reflects away from the satellite, then there will not be any signal and the coherence will be near 0. A coherence near 0 in "serious vegetation" (dense or thick vegetation like forests, I presume) is true, there are many variables that play a role here that could be alluded to. I understand what the authors are trying to say here, but readers unfamiliar with InSAR may be mislead or confused. 

Line 185: Two capital letters begin the sentence in this line.

References:

The reference list is lacking. There are many key InSAR studies from around the world that are never mentioned -- particularly those that explain how InSAR work. Thirty-seven publications have been cited by the eight authors of this paper. Of the 37 citations, nine are self-citations (24% self-citation rate). I advise the authors to expand their citation view.

An example of how these references are lacking. The authors cite a two-page document written in 2019 (in Chinese) when showing the equation for coherence (lines 126-130). There are many other better sources (in the sense of completeness and date of publication) that could be used here. Yet the authors choose a relatively recent and diminutive source. 

Author Response

Please see the enclosed file.

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

The authors responded to my comments and suggestions risen in the review round 1. The manuscript is now improved and more readable than the previous version.