the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 21 Apr 2023
Rockfall monitoring with a Doppler radar on an active rock slide complex in Brienz/Brinzauls (Switzerland)
Abstract. We present and analyze a rockfall catalog from an active landslide complex in Brienz/Brinzauls of the Swiss Alps, collected with a new Doppler radar system. This radar system provides a complete and continuous time-series of rockfall events with volumes of 1m3 and bigger since 2018 and serves as automatic traffic control for an important cantonal road. In the period between January 2018 and September 2022, 6743 events were detected, which is two orders of magnitude higher activity than in stable continental cliffs. A few percent of all rockfall events reached the shadow zone, which hosts an important road and agricultural area. The Doppler radar data set allows us to investigate the triggering factors quantitatively. We found that the background rockfall activity is controlled by seasonal climatic triggers. In winter, more rockfalls are observed during thawing periods, whereas in summer the rockfall activity increases with hourly rainfall intensity. We also found that due to the geological setting in an active landslide complex, increased rockfall activity occurs clustered in space and time, triggered by local displacement hotspots. Thus, monitoring spatial and temporal variations of slope displacement velocity is crucial for detailed rockfall hazard assessment in similar geological settings.
-
Notice on discussion status
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
-
Preprint
(42079 KB)
-
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(42079 KB) - Metadata XML
- BibTeX
- EndNote
- Final revised paper
Journal article(s) based on this preprint
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2023-555', Anonymous Referee #1, 16 May 2023
Referee’s comments on the manuscript “Rockfall monitoring with a Doppler radar on an active rock slide complex in Brienz/Brinzauls (Switzerland)”
This paper addresses the assessment of a complete rockfall inventory based on the Doppler radar technology resulting in an unprecedented temporal and spatial resolution. Thanks to the coupling with meteorological and InSAR measurements, the authors showed that the movement of (single compartments) of the rock slide complex is the main preparatory causal factor for the total of 5667 observed rockfall events (2018-2021). Rain and temperature play a minor, but not neglectable role in triggering rockfalls in the given study area Brienz/Brienzauls (Switzerland). Thanks to the rockfall inventory's high temporal resolution, the data's low daily correlation could have been compared to the hourly meteorological evolution. This thorough analysis showed that during the winter months, the daily maximum temperature and in the summer months, the rain intensity correlated with the rockfall activity.
While the data is worth publishing, this current manuscript needs a revision aiming for more consistent terminology (e.g., shadow zone, danger zone hazard zone) and concise writing with an enhanced reading flow (e.g., without “on the other hand”). As a non-native English speaker, I did not notice many linguistic errors. Instead, minor repetitions or the lack of an introduction (e.g., the activity parameter) should be tackled.
Specific comments:
l. 3 (and throughout the whole manuscript) :
Spaces must be included between number and unit, without italics: 1 m3, not 1m3 (https://www.natural-hazards-and-earth-system-sciences.net/submission.html#assets)
l. 17-20:
Although the reference is given to Evans and Hungr (1993), and the shadow zone of the rockfall process is described in detail with different wording, it might make sense to reference Figure 7, introducing a visual explanation of the shadow zone.
l. 47:
Consider using Geopraevent AG or Geopraevent Ltd. to clarify their requirements as a private sector company (see methods).
l. 49:
The reference is missing any DOI/internet address.
Figure 1:
The layout of Fig. 1 could be enhanced: The inset map of Switzerland has much-unused space, but the legend and the main figure are hardly readable (in original zoom size). The information content of the figure can be increased if the weather-stations of Tab. 1 are additionally shown (certainly those of Brienz in the main map, but also the others in a larger section of the map or a second inset map of the region/canton). Also, displaying the longitudinal profile of Fig. 7, including the APEX point and the (additional) compartment names, would increase the information content. The designation of the danger zone needs to be clarified: Does it refer to the shadow zone, or is it based on the danger map of the canton?
Figure 2:
Although this figure and its caption give a good overview of the study area, some details might be enhanced or clarified: 2.a) Consider including the Photo location in Fig. 1. Caption 2.b): In the text is also the road closing infrastructure mentioned (l. 74). The red light visible in Fig. 2.b) is thus worth mentioning. Additionally, Fig. 2.b, c) would benefit from an overlay with the geological formations and/or the landslide compartments. Fig. 2.e) Why is there a shift between the mapped rock outlines and the orthophoto? Fig. 2.f) Mark the dams better (The lowest, far western dam directly on the road is barely visible). Again, both terms are mentioned (which might be right): Shadow and danger zone. Make it more clear by mapping the shadow zone.
Table 1:
Only here you make use of parentheses. In the other figures’ labels, you use brackets around the units. To enhance consistency, I would use only parentheses (as in most NHESS publications) or adapt at least here the parentheses to brackets.
l. 74-76:
The observed velocity changes of the reflectors between the seasons belong (besides a good explanation) to the result section.
l. 80:
Again a new term: “hazard zone”. Its definition is also not concise. Instead, rephrase it: “We define this grassland including the cantonal road as a danger zone (Fig. 1), due to the increased damage potential (traffic).” This is necessary, as the grassland itself barely accounts for a higher damage potential.
l.81:
Source of the large volume events?
l.85:
“…and a/the local station…”
l. 89-90:
Which temperature sensor is used, and which accuracy according to the manufacturer? Is the housing ventilated? What is the ground material (indirect radiation)?
l. 91-95:
For clarity, I propose to rewrite the section. The most obvious choice to use the temperature data is the local station, as it is the closest one. So I would argue the other way around. E.g., “The proximity to the process area makes it suitable to use the temperature data set from the local station for further analysis. A high overall correlation (0.93) has been observed between the temperature of Savogingn and the local station (BRINZ). The slight differences might be due to topographic temperature effects, as the station in Savognin measures lower mean temperatures in winter (2 °C), but higher temperatures in summer (1 °C). ”
l. 96-100:
Please rewrite for clarity: The sentence “Secondly, freeze-thaw (FT) cycles can be divided into three phases.” has the potential for confusion. Consider rewriting: “Secondly, freeze-thaw (FT) cycles, which we divide into three phases”. Thanks to the proposed phrasing, it is clear, that only two metrics (l. 96) are meant and not three. Also, equations 6-8 could be moved directly after l. 100 to enhance clarity.
l 113.:
How is the temperature corrected?
l. 119:
Is the snow height also elevation-corrected?
l. 127-132:
Shorten the section: Slightly repetition within
l. 137:
Reconsider: “The device has a 90° horizontal field of view, from 302° - 32° azimuth.”
l.140-142:
Unclear relationship of the volume and distance: “Within a distance of 100 m the radar can detect moving masses larger than 0.1m3. Increasing the distance to 1km, a minimum volume of 1m3 is necessary for detection.”
l.142-143:
The minimal velocity is not mentioned in the source (Gassner et al., 2022)
Figure 4:
Consider writing “true positive wrong extent (TPWE)” instead of “TPWE”, as this abbreviation is not common and in the figure is (in the current typesetting) before the mention within the text
l. 146- 150:
Also, the signal-to-noise ratio is not explained in detail in the source (Gassner et al., 2022), However, such deeper insights are very welcome. Especially as the further “advanced algorithms” are described rather mysteriously. What do they do? And how? If these algorithms are developed by the company Geopraevent AG and are not meant to be disclosed, that should be mentioned clearly.
l. 155:
The Nr. of detected rockfalls belongs to the results section. However, another total amount of detected rockfalls (5667, l. 199) is given for a slightly different period. Clarify which period is more meaningful.
l. 156:
Activity maps: provided by whom?
l 178:
Specify how you include the caution in treating the eastern starting point or move this sentence to the discussion.
l 183:
Typo: “in the danger”?
l 192:
Which is the minimum rock size?
Figure 5:
Fig. 5.a and Fig. 5.d look very alike. Therefore, I propose that the subplot titles are descriptive and highlight the difference: “All rockfall event time series…” vs. “Days with high rockfall activity (>10 events d-1)”.
l. 204,
To maintain a concise terminology, you might change grassland to shadow- or danger zones.
l. 209,
The Doppler radar only observed rare events released from colluvial deposits.
l. 211,
The small second peak in the N-S-histogram of Figure 6 may also result from the different orientations of the map and the radar/mountain exposition. Try to rotate the map such that the histograms get even more meaningful.
l. 212:
If the daily rockfall rate strongly fluctuates, besides the mean, the standard deviation would also be of interest.
l. 213, 231, 251:
“Significantly” in terms of a statistical test? Which one? Rephrase otherwise.
l. 219:
To reference all subplots separately, write ”..hereafter considered as days with high rockfall activity (Fig. 5d).”
l. 249:
Consistency: Grassland or shadow: If you use it as a synonym, clarify that earlier and use just one term here. Otherwise, focus on shadow.
l.285:
To prepare the reader that the division between summer and winter also resulted in different subplots, you might add the corresponding references: “…we divide the data set into summer (April to September, Fig 10.b-c) and winter (October to March, Fig. 10.d-e) months.”
Table2:
After looking at the data plotted in Fig. 10 b and d, I would assume a negative coefficient with mean hourly summer rain (highest rockfall activity during the night, while rainfall events have their peak in the afternoon) and a positive coefficient for mean winter temperature (rockfall and temperature have their peak in the afternoon). Explain in the table caption (and text) more proactive the differences compared to Fig. 10 (daily/hourly). Even better: integrate the hourly correlation coefficients as well into the table.
Figure 10:
Adapt the y-axis label of Fig.10.b and d with the corresponding season for clarity. Although the caption correctly describes the summer and winter months, the bar graphs (b, d) are confusing because they show different data, despite identical axis labels: E.g., “Tot. events summer (h-1)”
l. 307-310.:
This topic has already been introduced in the methods section. Shorten this repetition.
l. 316-320:
hard-to-read, long introduction sentence, followed by two very short ones. Rephrase to avoid imbalance and enhance readability.
l. 344-345:
After an interesting thought and a good argumentation(342-344), mental agility is required by the reader to understand also the third sentence, which begins with a typical sentence intro (However, therefore, on the other hand). The wrong references (Fig.3 instead of 2 and Fig. 5 instead of 4?) and the prior missing link between the two figures are additionally unhelpful. Try to combine these figures into one and/or add the Pearson’s correlation coefficients in Tab. 2. Then, the results section will describe the missing linear correlation, and here, the reader would not be surprised. Additionally, a reference to Fig. 8 could underline the argumentation and the importance of InSAR observations.
“Therefore, local rock slope displacement and acceleration can be seen as major rockfall preparatory factors (Fig. 8). This insight is only possible thanks to the spatial InSAR observation, as the single, nearby monitoring points (Fig. 3 or new combined fig.) and rockfall frequency (Fig. 5 or new combined fig.) have no linear correlation. “
l. 346-349:
Elaborate the activity parameter A and its calculation base in more detail. Why is the assumption of a few m3 per event necessary, but appears not in the unit of A? What is the unit of A anyhow: per year per hour per square meter? As your findings are due to the observed high rate very relevant, they should be better introduced.
l. 383-384:
Can this assumption (“is likely related”) be underlined with data? Different DEM over time?
l. 385-386:
Inconsistency in the terminology: Does your mapped “Danger zone” (Fig. 1) contain the here mentioned dam? Even if so, the second sentence does not make sense: reaching the dam would then mean: reaching the danger zone.
l. 387:
If a literature study is mentioned, it would be helpful to provide the sources.
l. 402-405:
Twice “the road is automatically reopened after a few minutes”. Rewrite for reading flow.
l. 400-410:
Own data? Otherwise: Sources?
l 420-430:
Mentioning the activity parameter and the dam filling in the conclusion stresses the importance of the comments l. 346-349 and l. 383-384.
l. 436:
Consider also including the limitations of the previously widely used monitoring system into the conclusion: “than weather data or single monitoring points.”
Citation: https://doi.org/10.5194/egusphere-2023-555-RC1 -
AC1: 'Reply on RC1', Simon Loew, 22 May 2023
Dear Anonymous Reviewer #1,
The authors would like to express their gratitude for your comprehensive comments. We acknowledge the feedback regarding writing style, table formatting, and figure layout, which will be applied after the comment of Anonymous Reviewer #2. We would like to provide a brief response to specific content-related comments:
- Terminology: Danger Zone and Shadow Zone: We agree that the usage of these terms lacks clear definitions and coherence. Therefore, we also find it necessary to revise the use of these terms and provide a clear definition.
- The local weather station used (l. 89-90) is a relatively cheep Holfuy-Station installed 2m above ground. The temperature senor has a radiation shield. Details are given in https://holfuy.com/de.
- Temperature correction (l. 113): We apologize for the inclusion of this text sequence from an earlier draft. No temperature correction was applied to the local weather data.
- Snow depth elevation correction (l. 119): The data utilized in this study originated from an unpublished study conducted by the SLF. The snow depth measurements were obtained through extensive LIDAR measurements using helicopter surveys, where in the end an accumulation field could be calculated. Hence, a very detailed actual snow depth distribution was received for the used time period.
- Methods by Geopraevent AG (l. 146-150): The methods employed by Geopraevent AG are not intended for public disclosure and are subject to confidentiality agreements. As a result, we are not authorized to publish sensitive descriptions. We will provide a more detailed explanation of the public disclosure in the text to ensure complete transparency.
- Activity parameter A (l. 346-349): The frequency analyses can display the cumulative or the non-cumulative distribution of the rockfall volumes. These distributions are usually fitted by a power law for the volume range where the inventory is exhaustive. Then the spatio-temporal frequency F of rockfalls bigger than a volume V can be expressed as: F=A(V0) (V/V0)-B. Where A is the frequency of rockfalls with a volume bigger than V0 (an activity parameter) and B is a uniformity coefficient, which reflects the decrease of the frequency when the volume increases. V0 is the minimal value of the considered volume range or a minimal volume of interest, which depends on the context of the analysis. See Loew et al. 2021 for more details.
Thank you for your valuable feedback, which will significantly contribute to improving the clarity and accuracy of our manuscript.
Sincerely,
The Authors
-
AC1: 'Reply on RC1', Simon Loew, 22 May 2023
-
RC2: 'Comment on egusphere-2023-555', Anonymous Referee #2, 19 Jun 2023
Dear Editor,
Please find below my review of the manuscript: egusphere-2023-555 entitled:
Rockfall monitoring with a Doppler radar on an active rock slide complex in Brienz/Brinzauls (Switzerland)
By
Marius Schneider, Nicolas Oestreicher, and Simon Loew
This paper presents an analysis of rockfall occurrences within an active landslide, utilizing both observational data and Doppler radar measurements. The study demonstrates the significance of climatic factors and landslide displacements in facilitating rockfall activity. While the title implies a focus on Doppler monitoring, the main objective of the research is to investigate rockfall triggering mechanisms and evaluate potential rock arrest locations using Doppler Radar.
This review provides general comments on the paper's content, suggests improvements, and highlights specific areas for clarification.
General Comments
The title could be refined to better reflect the focus on analyzing rockfall triggering and evaluating rock arrest locations using Doppler Radar. Additionally, the inclusion of a geological map would greatly enhance the understanding of the study, considering the frequent references to geological aspects throughout the text. Section 3.3 requires a clearer explanation, potentially aided by a flowchart to elucidate the selection process for the displayed source areas in Figure 4. Further clarification is needed regarding the manual mapping of 70% of the 971 events on pictures and the utilization of machine learning, particularly since only 10% are allocated as test sets. Additionally, the authors could consider utilizing cross-correlation analysis to examine the time delay between climate effects and rockfall triggering. Lastly, certain sections, such as lines 400 to 410, seem irrelevant to the paper's purpose.
In terms of references, while D'Amato et al. (2016) is cited for freezing and thawing, it is also relevant to rainfall effects.
The term "Cantonal Road" requires clarification, as it appears to be a Swiss classification that could be replaced with "main road" for better comprehension.
Exploring the influence of temperature alone could provide valuable insights, especially considering the south-facing slope indicated in Figure 10d, where the effect of freezing and thawing cycles appears immediate, though the feasibility of such an analysis may be challenging.
Specific Comments
- Line 37: Consider adding the following reference: Hungr, O., Evans, S.G., and Hazzard, J. 1999. The magnitude and frequency of rock falls and rockslides along the main transportation corridors of southwestern British Columbia. Canadian Geotechnical Journal, 36: 224–238. doi:10.1139/t98-106.
- Fig. 1: It is recommended to include a geological map alongside the existing figures.
- Figure 3: Clarify the meaning of "TPS."
- Lines 203 and 2008: Specify whether 60 m3 and 100 m3 refer to maximum block volume, and source volume respectively.
- Line 202: Provide information on the minimum and maximum per day values.
- Figure 6: Improve the legend's clarity by explaining the histogram and consider using lines instead of transparent histograms.
- Consider employing the term "shadow area" or "rockfall shadow" instead of "shadow zone," as previously used by Evans and Hungr.
- In conclusion, this scientific paper review acknowledges the research's contributions in analyzing rockfall triggering in an active landslide context
Citation: https://doi.org/10.5194/egusphere-2023-555-RC2 -
AC2: 'Reply on RC2', Simon Loew, 28 Jun 2023
Dear Anonymous Reviewer #2,
Thank you for your positive feedback and valuable comments, which have greatly contributed to improving the quality of our publication. We appreciate the opportunity to address the points you raised.
We agree that incorporating a geological map in Figure 1 will enhance the clarity and overall understanding of the study. We will include the geological map in the revised version of this figure. Furthermore, we acknowledge the importance of providing a clear data processing procedure. To address this, we will incorporate a flowchart in Section 3.3 adressing the ML part, as suggested. An additional figure could be added to the supplementary materials explaining in detail how we have identified the event starting points (release points of all detected rock fall events). We have already prepared a draft of this flowchart and figure, which are added to this response in ZIP file.
We would like to provide some additional clarification regarding the manual mapping process for the ML training dataset. During weather transitions, such as from clear weather to thunderstorms, there were instances where the spatial data points for rockfall events appeared unreasonable. In such cases, we introduced the term "true positive wrong extent" (TPWE). In such events, data points are located at coordinates that are not visible to the radar (behind a ridge), hence being unusable for the event starting point analysis. Fortunately, TPWE events exhibited distinct characteristics that allowed us to differentiate them from valid, undisturbed spatial data points. The manual classification of random 1000 events as valid or TPWE was facilitated by these distinctive characteristics on plots. We acknowledge the potential overrepresentation of valid events in the manual catalog, which could bias the classification towards valid events. To mitigate this, we increased the threshold probability for classifying events as TPWE from the default 0.5 to 0.8. It is important to note that the allocated 10% as the test set serves only for model validation, while the training set, consisting of 700 images, was used for model training.
Regarding the title change suggestion, we understand your perspective. As this publication primarily focuses on 1) demonstrating the possibilities and utility of the Doppler radar data for rockfall monitoring (this is the first time ever that a Doppler radar has been set-up to monitor rockfall events), and 2) studying properties of rockfall events from an active rockslide, we are hesitant to modify the title. However, we will carefully consider this suggestion and engage in discussions within the editor team to determine the most appropriate course of action.
In our early study phase, we focused on investigating the cross-correlation between environmental factors and rockfall events at a daily scale. We observed an instantaneous reaction without any time lag between precipitation and rockfall. However, we are also aware that some compartments of the landslide in Brienz may respond with a time lag, resulting in subsequent rockfall. This hypothesis requires further investigation in subsequent studies and is beyond the scope of this particular publication.
All other detailed comments will be considered for the final revision of the manuscript.
The authors sincerely appreciate your feedback and comments, and we are grateful for the opportunity to address them. Your valuable input has significantly improved the clarity and depth of our work.
Sincerely,
The Authors
-
AC3: 'Comment on egusphere-2023-555', Simon Loew, 24 Jul 2023
Thank you for Reviewer Comments 1 and 2 which we have carefully studied and discussed. We have not received any other interactive community comments and have started to revise text and figures following the reviewer comments. We would like to stress that Doppler Radar has never been used before to monitor rock fall and therefore the contribution is also important from a methodological point of view.
Since submission of the manuscript, the Brienz/Brinzauls "Insel" compartment has failed catastrophically. The event was discussed in an Guest Blog of the AGU Landslide Blog one week later (Loew et al. 2023). As discussed in this EGU manuscript the Insel compartment was an important rock fall source, and the the models of rock fall mechanism and triggers in an active landslide compartment were confirmed by the events which happend since manuscript submission. This paper is now even more topical than before.
Best Regards
The Authors
Citation: https://doi.org/10.5194/egusphere-2023-555-AC3
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2023-555', Anonymous Referee #1, 16 May 2023
Referee’s comments on the manuscript “Rockfall monitoring with a Doppler radar on an active rock slide complex in Brienz/Brinzauls (Switzerland)”
This paper addresses the assessment of a complete rockfall inventory based on the Doppler radar technology resulting in an unprecedented temporal and spatial resolution. Thanks to the coupling with meteorological and InSAR measurements, the authors showed that the movement of (single compartments) of the rock slide complex is the main preparatory causal factor for the total of 5667 observed rockfall events (2018-2021). Rain and temperature play a minor, but not neglectable role in triggering rockfalls in the given study area Brienz/Brienzauls (Switzerland). Thanks to the rockfall inventory's high temporal resolution, the data's low daily correlation could have been compared to the hourly meteorological evolution. This thorough analysis showed that during the winter months, the daily maximum temperature and in the summer months, the rain intensity correlated with the rockfall activity.
While the data is worth publishing, this current manuscript needs a revision aiming for more consistent terminology (e.g., shadow zone, danger zone hazard zone) and concise writing with an enhanced reading flow (e.g., without “on the other hand”). As a non-native English speaker, I did not notice many linguistic errors. Instead, minor repetitions or the lack of an introduction (e.g., the activity parameter) should be tackled.
Specific comments:
l. 3 (and throughout the whole manuscript) :
Spaces must be included between number and unit, without italics: 1 m3, not 1m3 (https://www.natural-hazards-and-earth-system-sciences.net/submission.html#assets)
l. 17-20:
Although the reference is given to Evans and Hungr (1993), and the shadow zone of the rockfall process is described in detail with different wording, it might make sense to reference Figure 7, introducing a visual explanation of the shadow zone.
l. 47:
Consider using Geopraevent AG or Geopraevent Ltd. to clarify their requirements as a private sector company (see methods).
l. 49:
The reference is missing any DOI/internet address.
Figure 1:
The layout of Fig. 1 could be enhanced: The inset map of Switzerland has much-unused space, but the legend and the main figure are hardly readable (in original zoom size). The information content of the figure can be increased if the weather-stations of Tab. 1 are additionally shown (certainly those of Brienz in the main map, but also the others in a larger section of the map or a second inset map of the region/canton). Also, displaying the longitudinal profile of Fig. 7, including the APEX point and the (additional) compartment names, would increase the information content. The designation of the danger zone needs to be clarified: Does it refer to the shadow zone, or is it based on the danger map of the canton?
Figure 2:
Although this figure and its caption give a good overview of the study area, some details might be enhanced or clarified: 2.a) Consider including the Photo location in Fig. 1. Caption 2.b): In the text is also the road closing infrastructure mentioned (l. 74). The red light visible in Fig. 2.b) is thus worth mentioning. Additionally, Fig. 2.b, c) would benefit from an overlay with the geological formations and/or the landslide compartments. Fig. 2.e) Why is there a shift between the mapped rock outlines and the orthophoto? Fig. 2.f) Mark the dams better (The lowest, far western dam directly on the road is barely visible). Again, both terms are mentioned (which might be right): Shadow and danger zone. Make it more clear by mapping the shadow zone.
Table 1:
Only here you make use of parentheses. In the other figures’ labels, you use brackets around the units. To enhance consistency, I would use only parentheses (as in most NHESS publications) or adapt at least here the parentheses to brackets.
l. 74-76:
The observed velocity changes of the reflectors between the seasons belong (besides a good explanation) to the result section.
l. 80:
Again a new term: “hazard zone”. Its definition is also not concise. Instead, rephrase it: “We define this grassland including the cantonal road as a danger zone (Fig. 1), due to the increased damage potential (traffic).” This is necessary, as the grassland itself barely accounts for a higher damage potential.
l.81:
Source of the large volume events?
l.85:
“…and a/the local station…”
l. 89-90:
Which temperature sensor is used, and which accuracy according to the manufacturer? Is the housing ventilated? What is the ground material (indirect radiation)?
l. 91-95:
For clarity, I propose to rewrite the section. The most obvious choice to use the temperature data is the local station, as it is the closest one. So I would argue the other way around. E.g., “The proximity to the process area makes it suitable to use the temperature data set from the local station for further analysis. A high overall correlation (0.93) has been observed between the temperature of Savogingn and the local station (BRINZ). The slight differences might be due to topographic temperature effects, as the station in Savognin measures lower mean temperatures in winter (2 °C), but higher temperatures in summer (1 °C). ”
l. 96-100:
Please rewrite for clarity: The sentence “Secondly, freeze-thaw (FT) cycles can be divided into three phases.” has the potential for confusion. Consider rewriting: “Secondly, freeze-thaw (FT) cycles, which we divide into three phases”. Thanks to the proposed phrasing, it is clear, that only two metrics (l. 96) are meant and not three. Also, equations 6-8 could be moved directly after l. 100 to enhance clarity.
l 113.:
How is the temperature corrected?
l. 119:
Is the snow height also elevation-corrected?
l. 127-132:
Shorten the section: Slightly repetition within
l. 137:
Reconsider: “The device has a 90° horizontal field of view, from 302° - 32° azimuth.”
l.140-142:
Unclear relationship of the volume and distance: “Within a distance of 100 m the radar can detect moving masses larger than 0.1m3. Increasing the distance to 1km, a minimum volume of 1m3 is necessary for detection.”
l.142-143:
The minimal velocity is not mentioned in the source (Gassner et al., 2022)
Figure 4:
Consider writing “true positive wrong extent (TPWE)” instead of “TPWE”, as this abbreviation is not common and in the figure is (in the current typesetting) before the mention within the text
l. 146- 150:
Also, the signal-to-noise ratio is not explained in detail in the source (Gassner et al., 2022), However, such deeper insights are very welcome. Especially as the further “advanced algorithms” are described rather mysteriously. What do they do? And how? If these algorithms are developed by the company Geopraevent AG and are not meant to be disclosed, that should be mentioned clearly.
l. 155:
The Nr. of detected rockfalls belongs to the results section. However, another total amount of detected rockfalls (5667, l. 199) is given for a slightly different period. Clarify which period is more meaningful.
l. 156:
Activity maps: provided by whom?
l 178:
Specify how you include the caution in treating the eastern starting point or move this sentence to the discussion.
l 183:
Typo: “in the danger”?
l 192:
Which is the minimum rock size?
Figure 5:
Fig. 5.a and Fig. 5.d look very alike. Therefore, I propose that the subplot titles are descriptive and highlight the difference: “All rockfall event time series…” vs. “Days with high rockfall activity (>10 events d-1)”.
l. 204,
To maintain a concise terminology, you might change grassland to shadow- or danger zones.
l. 209,
The Doppler radar only observed rare events released from colluvial deposits.
l. 211,
The small second peak in the N-S-histogram of Figure 6 may also result from the different orientations of the map and the radar/mountain exposition. Try to rotate the map such that the histograms get even more meaningful.
l. 212:
If the daily rockfall rate strongly fluctuates, besides the mean, the standard deviation would also be of interest.
l. 213, 231, 251:
“Significantly” in terms of a statistical test? Which one? Rephrase otherwise.
l. 219:
To reference all subplots separately, write ”..hereafter considered as days with high rockfall activity (Fig. 5d).”
l. 249:
Consistency: Grassland or shadow: If you use it as a synonym, clarify that earlier and use just one term here. Otherwise, focus on shadow.
l.285:
To prepare the reader that the division between summer and winter also resulted in different subplots, you might add the corresponding references: “…we divide the data set into summer (April to September, Fig 10.b-c) and winter (October to March, Fig. 10.d-e) months.”
Table2:
After looking at the data plotted in Fig. 10 b and d, I would assume a negative coefficient with mean hourly summer rain (highest rockfall activity during the night, while rainfall events have their peak in the afternoon) and a positive coefficient for mean winter temperature (rockfall and temperature have their peak in the afternoon). Explain in the table caption (and text) more proactive the differences compared to Fig. 10 (daily/hourly). Even better: integrate the hourly correlation coefficients as well into the table.
Figure 10:
Adapt the y-axis label of Fig.10.b and d with the corresponding season for clarity. Although the caption correctly describes the summer and winter months, the bar graphs (b, d) are confusing because they show different data, despite identical axis labels: E.g., “Tot. events summer (h-1)”
l. 307-310.:
This topic has already been introduced in the methods section. Shorten this repetition.
l. 316-320:
hard-to-read, long introduction sentence, followed by two very short ones. Rephrase to avoid imbalance and enhance readability.
l. 344-345:
After an interesting thought and a good argumentation(342-344), mental agility is required by the reader to understand also the third sentence, which begins with a typical sentence intro (However, therefore, on the other hand). The wrong references (Fig.3 instead of 2 and Fig. 5 instead of 4?) and the prior missing link between the two figures are additionally unhelpful. Try to combine these figures into one and/or add the Pearson’s correlation coefficients in Tab. 2. Then, the results section will describe the missing linear correlation, and here, the reader would not be surprised. Additionally, a reference to Fig. 8 could underline the argumentation and the importance of InSAR observations.
“Therefore, local rock slope displacement and acceleration can be seen as major rockfall preparatory factors (Fig. 8). This insight is only possible thanks to the spatial InSAR observation, as the single, nearby monitoring points (Fig. 3 or new combined fig.) and rockfall frequency (Fig. 5 or new combined fig.) have no linear correlation. “
l. 346-349:
Elaborate the activity parameter A and its calculation base in more detail. Why is the assumption of a few m3 per event necessary, but appears not in the unit of A? What is the unit of A anyhow: per year per hour per square meter? As your findings are due to the observed high rate very relevant, they should be better introduced.
l. 383-384:
Can this assumption (“is likely related”) be underlined with data? Different DEM over time?
l. 385-386:
Inconsistency in the terminology: Does your mapped “Danger zone” (Fig. 1) contain the here mentioned dam? Even if so, the second sentence does not make sense: reaching the dam would then mean: reaching the danger zone.
l. 387:
If a literature study is mentioned, it would be helpful to provide the sources.
l. 402-405:
Twice “the road is automatically reopened after a few minutes”. Rewrite for reading flow.
l. 400-410:
Own data? Otherwise: Sources?
l 420-430:
Mentioning the activity parameter and the dam filling in the conclusion stresses the importance of the comments l. 346-349 and l. 383-384.
l. 436:
Consider also including the limitations of the previously widely used monitoring system into the conclusion: “than weather data or single monitoring points.”
Citation: https://doi.org/10.5194/egusphere-2023-555-RC1 -
AC1: 'Reply on RC1', Simon Loew, 22 May 2023
Dear Anonymous Reviewer #1,
The authors would like to express their gratitude for your comprehensive comments. We acknowledge the feedback regarding writing style, table formatting, and figure layout, which will be applied after the comment of Anonymous Reviewer #2. We would like to provide a brief response to specific content-related comments:
- Terminology: Danger Zone and Shadow Zone: We agree that the usage of these terms lacks clear definitions and coherence. Therefore, we also find it necessary to revise the use of these terms and provide a clear definition.
- The local weather station used (l. 89-90) is a relatively cheep Holfuy-Station installed 2m above ground. The temperature senor has a radiation shield. Details are given in https://holfuy.com/de.
- Temperature correction (l. 113): We apologize for the inclusion of this text sequence from an earlier draft. No temperature correction was applied to the local weather data.
- Snow depth elevation correction (l. 119): The data utilized in this study originated from an unpublished study conducted by the SLF. The snow depth measurements were obtained through extensive LIDAR measurements using helicopter surveys, where in the end an accumulation field could be calculated. Hence, a very detailed actual snow depth distribution was received for the used time period.
- Methods by Geopraevent AG (l. 146-150): The methods employed by Geopraevent AG are not intended for public disclosure and are subject to confidentiality agreements. As a result, we are not authorized to publish sensitive descriptions. We will provide a more detailed explanation of the public disclosure in the text to ensure complete transparency.
- Activity parameter A (l. 346-349): The frequency analyses can display the cumulative or the non-cumulative distribution of the rockfall volumes. These distributions are usually fitted by a power law for the volume range where the inventory is exhaustive. Then the spatio-temporal frequency F of rockfalls bigger than a volume V can be expressed as: F=A(V0) (V/V0)-B. Where A is the frequency of rockfalls with a volume bigger than V0 (an activity parameter) and B is a uniformity coefficient, which reflects the decrease of the frequency when the volume increases. V0 is the minimal value of the considered volume range or a minimal volume of interest, which depends on the context of the analysis. See Loew et al. 2021 for more details.
Thank you for your valuable feedback, which will significantly contribute to improving the clarity and accuracy of our manuscript.
Sincerely,
The Authors
-
AC1: 'Reply on RC1', Simon Loew, 22 May 2023
-
RC2: 'Comment on egusphere-2023-555', Anonymous Referee #2, 19 Jun 2023
Dear Editor,
Please find below my review of the manuscript: egusphere-2023-555 entitled:
Rockfall monitoring with a Doppler radar on an active rock slide complex in Brienz/Brinzauls (Switzerland)
By
Marius Schneider, Nicolas Oestreicher, and Simon Loew
This paper presents an analysis of rockfall occurrences within an active landslide, utilizing both observational data and Doppler radar measurements. The study demonstrates the significance of climatic factors and landslide displacements in facilitating rockfall activity. While the title implies a focus on Doppler monitoring, the main objective of the research is to investigate rockfall triggering mechanisms and evaluate potential rock arrest locations using Doppler Radar.
This review provides general comments on the paper's content, suggests improvements, and highlights specific areas for clarification.
General Comments
The title could be refined to better reflect the focus on analyzing rockfall triggering and evaluating rock arrest locations using Doppler Radar. Additionally, the inclusion of a geological map would greatly enhance the understanding of the study, considering the frequent references to geological aspects throughout the text. Section 3.3 requires a clearer explanation, potentially aided by a flowchart to elucidate the selection process for the displayed source areas in Figure 4. Further clarification is needed regarding the manual mapping of 70% of the 971 events on pictures and the utilization of machine learning, particularly since only 10% are allocated as test sets. Additionally, the authors could consider utilizing cross-correlation analysis to examine the time delay between climate effects and rockfall triggering. Lastly, certain sections, such as lines 400 to 410, seem irrelevant to the paper's purpose.
In terms of references, while D'Amato et al. (2016) is cited for freezing and thawing, it is also relevant to rainfall effects.
The term "Cantonal Road" requires clarification, as it appears to be a Swiss classification that could be replaced with "main road" for better comprehension.
Exploring the influence of temperature alone could provide valuable insights, especially considering the south-facing slope indicated in Figure 10d, where the effect of freezing and thawing cycles appears immediate, though the feasibility of such an analysis may be challenging.
Specific Comments
- Line 37: Consider adding the following reference: Hungr, O., Evans, S.G., and Hazzard, J. 1999. The magnitude and frequency of rock falls and rockslides along the main transportation corridors of southwestern British Columbia. Canadian Geotechnical Journal, 36: 224–238. doi:10.1139/t98-106.
- Fig. 1: It is recommended to include a geological map alongside the existing figures.
- Figure 3: Clarify the meaning of "TPS."
- Lines 203 and 2008: Specify whether 60 m3 and 100 m3 refer to maximum block volume, and source volume respectively.
- Line 202: Provide information on the minimum and maximum per day values.
- Figure 6: Improve the legend's clarity by explaining the histogram and consider using lines instead of transparent histograms.
- Consider employing the term "shadow area" or "rockfall shadow" instead of "shadow zone," as previously used by Evans and Hungr.
- In conclusion, this scientific paper review acknowledges the research's contributions in analyzing rockfall triggering in an active landslide context
Citation: https://doi.org/10.5194/egusphere-2023-555-RC2 -
AC2: 'Reply on RC2', Simon Loew, 28 Jun 2023
Dear Anonymous Reviewer #2,
Thank you for your positive feedback and valuable comments, which have greatly contributed to improving the quality of our publication. We appreciate the opportunity to address the points you raised.
We agree that incorporating a geological map in Figure 1 will enhance the clarity and overall understanding of the study. We will include the geological map in the revised version of this figure. Furthermore, we acknowledge the importance of providing a clear data processing procedure. To address this, we will incorporate a flowchart in Section 3.3 adressing the ML part, as suggested. An additional figure could be added to the supplementary materials explaining in detail how we have identified the event starting points (release points of all detected rock fall events). We have already prepared a draft of this flowchart and figure, which are added to this response in ZIP file.
We would like to provide some additional clarification regarding the manual mapping process for the ML training dataset. During weather transitions, such as from clear weather to thunderstorms, there were instances where the spatial data points for rockfall events appeared unreasonable. In such cases, we introduced the term "true positive wrong extent" (TPWE). In such events, data points are located at coordinates that are not visible to the radar (behind a ridge), hence being unusable for the event starting point analysis. Fortunately, TPWE events exhibited distinct characteristics that allowed us to differentiate them from valid, undisturbed spatial data points. The manual classification of random 1000 events as valid or TPWE was facilitated by these distinctive characteristics on plots. We acknowledge the potential overrepresentation of valid events in the manual catalog, which could bias the classification towards valid events. To mitigate this, we increased the threshold probability for classifying events as TPWE from the default 0.5 to 0.8. It is important to note that the allocated 10% as the test set serves only for model validation, while the training set, consisting of 700 images, was used for model training.
Regarding the title change suggestion, we understand your perspective. As this publication primarily focuses on 1) demonstrating the possibilities and utility of the Doppler radar data for rockfall monitoring (this is the first time ever that a Doppler radar has been set-up to monitor rockfall events), and 2) studying properties of rockfall events from an active rockslide, we are hesitant to modify the title. However, we will carefully consider this suggestion and engage in discussions within the editor team to determine the most appropriate course of action.
In our early study phase, we focused on investigating the cross-correlation between environmental factors and rockfall events at a daily scale. We observed an instantaneous reaction without any time lag between precipitation and rockfall. However, we are also aware that some compartments of the landslide in Brienz may respond with a time lag, resulting in subsequent rockfall. This hypothesis requires further investigation in subsequent studies and is beyond the scope of this particular publication.
All other detailed comments will be considered for the final revision of the manuscript.
The authors sincerely appreciate your feedback and comments, and we are grateful for the opportunity to address them. Your valuable input has significantly improved the clarity and depth of our work.
Sincerely,
The Authors
-
AC3: 'Comment on egusphere-2023-555', Simon Loew, 24 Jul 2023
Thank you for Reviewer Comments 1 and 2 which we have carefully studied and discussed. We have not received any other interactive community comments and have started to revise text and figures following the reviewer comments. We would like to stress that Doppler Radar has never been used before to monitor rock fall and therefore the contribution is also important from a methodological point of view.
Since submission of the manuscript, the Brienz/Brinzauls "Insel" compartment has failed catastrophically. The event was discussed in an Guest Blog of the AGU Landslide Blog one week later (Loew et al. 2023). As discussed in this EGU manuscript the Insel compartment was an important rock fall source, and the the models of rock fall mechanism and triggers in an active landslide compartment were confirmed by the events which happend since manuscript submission. This paper is now even more topical than before.
Best Regards
The Authors
Citation: https://doi.org/10.5194/egusphere-2023-555-AC3
Peer review completion
Journal article(s) based on this preprint
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 402 | 186 | 24 | 612 | 12 | 9 |
- HTML: 402
- PDF: 186
- XML: 24
- Total: 612
- BibTeX: 12
- EndNote: 9
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
Cited
2 citations as recorded by crossref.
Marius Schneider
Nicolas Oestreicher
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(42079 KB) - Metadata XML