the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 03 Apr 2024
Monitoring snow depth variations in an avalanche release area using low cost LiDAR and optical sensors
Abstract. Snow avalanches threaten people and infrastructure in mountainous areas. For the assessment of temporal protection measures of infrastructure in dangerous situations, local and up to date information is crucial. One factor influencing the avalanche situation is wind drifted snow, which causes variations in snow depth across a slope, but this data is rarely available. We present a monitoring system using low cost LiDAR and optical sensors, which we use to monitor snow depth variations in an avalanche release area. The system is operational since November 2023, autonomously measuring and providing data from our study area close to Davos in Switzerland, with high spatiotemporal resolution. In the first three operating months we gained experiences and made a preliminary assessment of the system performance. An analysis of the changes in spatial coverage shows the limitations and potentials of the system under different weather conditions. A comparison of the surface models derived from the LiDAR data and a photogrammetric drone shows a good agreement, achieving a mean of 0.005 m and standard deviation of 0.15 m. Two case studies, including an avalanche event and a period of snowfall with strong winds, show the potential of the proposed system to detect changes in the snow depth distribution on a low decimeter level, or better. In addition, we record meteorological parameters which we will use in future, together with the newly established snow depth database, for attempts to refine and further develop models for wind-induced snow redistribution. The near real time information of the snow depth distribution in avalanche prone slopes will be provided to experts, so it can aid their decisions on avalanche safety measures.
- Preprint
(17672 KB) - Metadata XML
- BibTeX
- EndNote
Status: open (until 15 May 2024)
-
CC1: 'Comment on egusphere-2024-744', Thomas Gölles, 18 Apr 2024
reply
Dear Authors,
The state of the art section needs to be updated. We have been working on low cost lidar applications in geosciences and specifically for snow and avalanche applications for years. The paper describing our MOLISENS system already mentioned snow and avalanche applications in 2022. Also, more specifically the paper “Automated snow avalanche monitoring for Austria: State of the art and roadmap for future work” has a section about low cost lidar.
During the “5. Lawinensymposium Graz 2023“ we also presented first results using low cost lidar for avalanches.
For data processing we have our own open source Python package called pointcloudset, which would be useful for your future work and we could collaborate on adapting/testing it for you’re the Livox Avia lidar.
Required citations:
Goelles, T., Hammer, T., Muckenhuber, S., Schlager, B., Abermann, J., Bauer, C., Expósito Jiménez, V. J., Schöner, W., Schratter, M., Schrei, B., and Senger, K.: MOLISENS: MObile LIdar SENsor System to exploit the potential of small industrial lidar devices for geoscientific applications, Geosci. Instrum. Method. Data Syst., 11, 247–261, https://doi.org/10.5194/gi-11-247-2022, 2022.
Kapper KL, Goelles T, Muckenhuber S, Trügler A, Abermann J, Schlager B, Gaisberger C, Eckerstorfer M, Grahn J, Malnes E, Prokop A and Schöner W (2023) Automated snow avalanche monitoring for Austria: State of the art and roadmap for future work. Front. Remote Sens. 4:1156519. doi: 10.3389/frsen.2023.1156519
RSnowAUT-Konsortium (2023). Das Potential von Automobilsensoren für die lokale Detektion von Lawinen im Rahmen des FFG-Projekts RSnowAUT, 5. Lawinensymposium Graz 2023,https://lawinensymposium.naturfreunde.at/files/uploads/2023/11/Tagungsband_2023_gesamt.pdf
Optional citation about pointcloudset:
Goelles, T., Schlager, B., Muckenhuber, S., Haas, S., & Hammer, T. (2021). `pointcloudset`: Efficient Analysis of Large Datasets of Point Clouds Recorded Over Time. Journal of Open Source Software, 6(65), 3471. https://doi.org/10.21105/joss.03471
best regards,
Thomas Gölles
Citation: https://doi.org/10.5194/egusphere-2024-744-CC1 -
RC1: 'Referee Comment on egusphere-2024-744', Katreen Wikstrom Jones, 18 Apr 2024
reply
GENERAL COMMENTS
The paper titled “Monitoring snow depth variations in an avalanche release area using low cost LiDAR and optical sensors” presents a compelling project where high-end technology from the automotive industry is explored and adopted to enhance our understanding of snow processes for avalanche hazard applications, with the long-term goal to improve road safety. The paper is focused on describing the instrumentation, and lessons learned from the first three months of operation. I truly enjoyed reading the manuscript. The scientific significance, the scientific quality and the presentation quality are all ranging from good to excellent. Terrestrial lidar is typically cost-prohibitive and avalanche safety programs rarely have the funds to invest in monitoring networks that can provide near real-time information about the snowpack in the release areas. Here, the authors have designed a low-cost autonomous terrestrial lidar/meteorological station network that shows promising potential in providing high spatiotemporal resolution of snow distribution. I look forward to following this project which, by no doubt, will be of high interest by the global avalanche community. The goal of the project is clear, the background (literature review) is quite thorough, however, as pointed out by the community reviewer Thomas Gölles, it should be updated to include recent implementations of low-cost lidar systems for geoscience applications in Austria. While the purpose of the study is clear, I think the value of the study could be even more emphasized -- there is only one sentence in the introduction that touches on why snow depth variability in the release area matters for avalanches – please elaborate on this a bit further. The selected methodology is well motivated and considered (please see request below for additional brand/sensor details in Table). I commend the authors for their skill in keeping the chapters concise, yet adequately detailed. Chapter 2 Monitoring provides all the details necessary to understand the motivation behind selecting the suitable sensors, where and how to install the stations, and how the setup is configured. Anyone interested in implementing a similar system will find this section of the methodology very satisfying! I would like to point out that it would be important to include a brief paragraph that describes the UAV system that was used for generating photogrammetry-derived DSMs and how that data was processed. If these data were not acquired and processed by the author team, then please provide a reference to the data release/publication. For the most part, it is easy to understand the lidar Data processing (Chapter 3), however some clarification regarding some of the processing steps would be needed prior to publication (see specific comment and line comments below). I also suggest making the Data Processing (Chapter 3) a sub-section of chapter 2. Naturally, given the early stage of the project, the authors are cautious with drawing too many conclusions based on their results, nevertheless they provide some interesting discussion points which deserve further exploring. In addition, they have attempted to assess the system performance and they openly state some challenges and limitations that they have encountered this winter. I appreciate the authors determination in making the effort to assess the system performance thoroughly, to [especially more so in the future] analyze the derived products, and comparing the lidar results with other methods available (photogrammetry-derived DSMs using UAVs for example). The text generally flows well and the language in the paper is mostly good, though there are a few sections with a bit clunky or convoluted sentence structure, and where I would suggest rephrasing or removing unnecessary words (see line comments below). There are also a few typos that need correction, some abbreviations that need to be spelled out the first time they are being used, issues with hyphenation, and some consistency issues using certain terminology (e.g. drone vs UAV). These language issues are relatively minor and should not be a problem for the authors to tackle prior to publication. My suggested rewording or rephrasing aim at improving the flow of the text, to make the content clearer and ultimately strengthen the paper. If you have any questions regarding any of my comments or suggested edits, please feel free to contact me at katreen.wikstromjones@alaska.gov.SPECIFIC COMMENTS
Georeferencing the point cloud and aligning different models) Please elaborate a bit on what was involved in the georeferencing step using the prisms. Can you also please explain how you “projected the 3D models in the vertical direction”? Did you calculate horizontal and vertical offsets and applied shifts to the point cloud or models? It does not appear that co-registration of the lidar model to the photogrammetric model was made. I apologize if I misunderstood any of your processing steps; it would be important to co-register the datasets using stable features, such as the prisms, prior to differencing, especially when comparing a lidar-derived DSM to a photogrammetry-derived DSM (based on Line 269 in the Discussion, it suggests that a co-registration was not done yet). Photogrammetric models tend to be less accurate horizontally, unless they are well georeferenced, which for this type of steep terrain may result in large vertical offsets when compared to another model.
Interpolation of voids in the lidar-derived DSM) Did you evaluate other interpolation methods before choosing Nearest Neighbor? If yes, how did they perform?
Observed diurnal patterns in spatial coverage (by lidar)) This is an interesting observation! Line 209-215 in Chapter 4 Results, where you discuss potential explanations to this diurnal pattern, should be moved to Chapter 5 Discussion, and I would encourage you to elaborate on this a bit. I think you are on the right track – research has shown that the lidar sensor can be saturated from scattered sunlight. Even though this is a topic that you may investigate more in the future, I suggest you include a couple of these previous investigations to strengthen the outlook/discussion (e.g. Wu et al, 2011, https://doi.org/10.5194/acp-11-2641-2011). In the spring and summertime, I am curious to see if you’ll continue to see a diurnal pattern as less dense air caused by warmer air temperatures could scatter the laser pulses and cause a reduction in reflectance.
Terrain impacts for wind-induced redistribution of snow) In “4.4 Case study II: Snowfall event with wind” -- I suggest adding a sentence at the end of this section that emphasizes your results, something like "The clear pattern of eroded and deposited snow agrees with the recorded wind direction at nearby weather stations on this day." Additionally, in this section, I encourage you to analyze the interaction between topography and snow distribution a bit further and add a sentence or two that describe how this redistribution (erosion and deposition) was influenced by small terrain features within the ROI, such as scouring on taller small sub-ridges and deposition in lee-ward depressions/small gullies. In the discussion you could, for example, make a point that your results support the idea that it is important for avalanche safety experts to study the avalanche terrain in the summer to enhance their understanding of snow distribution in the winter (i.e. where the deep pockets of snow are vs. shallower potential trigger points).
Discrepancies vs. potential of the system) In Chapter 6 Discussion you state “While the discrepancies between the UAV-based acquisitions and the early processing results from our system are larger than the expected potential of our system…”, what do you mean by this? Can you please explain what you mean by the “expected potential”?
TECHNICAL CORRECTIONS
Note: Please see attached PDF for mark ups and comments (as they are referred to by line number below)!
Title -- Suggest changing “LiDAR” to “lidar” and doing so throughout the paper. Reference: https://lidarmag.com/wp-content/uploads/PDF/LiDARNewsMagazine_DeeringStoker-CasingOfLiDAR_Vol4No6.pdf
Line 2 “when the avalanche danger is high”, “up-to-date”
Line 3 “wind-drifted”
Line 4 suggest replacing “which we use to” with "to measure snow depth variations in an avalanche release area at a very high spatiotemporal scale."
Line 5 “terrestrial monitoring system”, clarify & add detail about the system "The system, which consists of two stations at xx m a sl and xx m a sl, is operational since Nov 2023", “low-cost”
Line 6 Add the resolution in m in paranthesis
Line 7 replace “changes” with “differences”
Line 8 replace “different” with “varying”
Line 9 “surface elevation models” (to clarify for the non-accustomed reader), "photogrammetry-derived model using a drone" (or UAV – choose one or the other and be consistent!)
Line 10 a mean of what? mean vertical/height difference?
Line 12-13 Suggest switching order to improve sentence structure "To refine and further develop models for wind-induced snow redistribution, we also record meteorological parameters in-situ at the two stations which we will analyze in the future together with a newly established snow depth database for the area."
Line 13 “real-time”
Line 14 "this" or "the avalanche-prone slope of interest"
Line 18 delete “Among other measures” and “in dangerous situations”.
Line 25 "methods to measure"
Line 41 suggest wording "Due to high costs, one disadvantage of airborne systems is the limited temporal resolution.", "ground-based"
Line 49 suggest replacing “features” with “textures” and “structures” with “features”
Line 55 suggest replacing “measure” with “operate”
Line 57 provide detail "near infrared region of the electromagnetic spectrum (900 ~ 1550 nm)"
Line 68 “it operates”, ", is" (comma needed)
Line 75 suggest replace with "present" since it's already installed, “low-cost”
Line 77 suggest replace “present” with "describe"
Chapter 2 Monitoring – General comment regarding order of sections: As a reader I would like to know right away what the study site looks like and learn about the problem (avalanches – the characteristics of this particular avalanche track) because ultimately the avalanche problem is the motivation behind this project. Study site/problem description first – it will motivate why you installed these lidar/meteorological monitoring stations and where, and why you selected the specific instrumentation. I suggest switching the order of these section and provide section 2.2 first (as 2.1) and provide a map of the area with the locations of the stations, outline of the avalanche track, and insert photos of both stations with labels of all the instrumentation. This is obviously my subjective opinion – I can understand why you want to emphasize the study area less and put the monitoring system up-front first. I would recommend that you task a few people, unfamiliar with the project, to read the paper and share their opinion on this.
Line 81 Provide the full description "single-lens reflex (SLR) camera"
Line 82 elaborate a bit "we installed sensors at the stations to record"
Line 85 add "which allows for remote monitoring and data analysis"
Line 91 what's the unit here? Hz?, "application, the" (comma needed)
Line 92 Not sure what you mean here. Do you mean "the spatial coverage increases as the sensor operates farther away from the target features/surface"?
Line 98 "a low power consumption" (add low)
Table 1 since eye safety, cost, and power consumption were important criteria, can you please add those to this table?
Line 103 suggest “This setup”, “on-site” and “as for the”
Line 111 Having a figure here that shows the location of all these instruments would be valuable to the reader!
Line 117 “real-time”
Line 118 suggest adding "due to occasional high avalanche danger."
Line 121 "avalanche track/path", “high-alpine”
Line 127 add "for the laser beam"
Line 128 add "to ensure that the station would not get buried", point #3 Was this a ground assessment? Maybe specify what you were looking for - stable bedrock?
Line 131 suggest replacing “lower” with “shallower”
Line 135 “lidar point clouds”, “, identifiable” (comma needed)
Line 137 replace with “on”, replace “seen” with “visible”
Line 140 "using the" (space needed)
Figure 3. For better presentation, I would suggest rotating the scene so that you're viewing the avalanche path from bottom up with the release area at the top of the figure (add north arrow to indicate viewing direction). Also add two vicinity maps (1) that shows the location of Wildi near Davos, and (2) location of Davos in Switzerland (this map should be very simplified). Those two vicinity maps can be stacked above the legend to the side of the main map. I also suggest coloring the Wildi outline with a different color. Make sure you use a color-blind color scheme for symbology.
Figure 4. This is a great figure! Please label in the image which station is Braema1 and Braema 2. I think it would also help to have the Wildi release area outlined in the background just to show how well the lidar swaths covers the general release area.
Line 142 “derived results” - vague wording
Line 144 typo "measurements", spell out and provide abbreviations for “real-time kinematic (RTK) – Global Navigation Satellite System (GNSS) unit”
Figure 5. I suggest including this figure in Section 2.1 instead where you describe the sensors.
Line 152 suggest using "scan" instead of "measurement"
Line 152 specify "solid-state drive (SSD)"
Line 154 again suggest replacing “measurement” with “scan”
Line 155 suggest "transfers" instead of “sends”
Line 158 again suggest replacing “measurement” with “scan”
Line 161 Not clear on what “local system” means? Local coordinate system?
Figure 6. This is a clear schematic of all the components and how power supply and data transfers are achieved in the system. In the caption, some suggestions “power-related”, replace “in” with “are in”.
Line 166 I do not understand what is being done in this step. Can you please explain?
Line 168 For an unaccustomed reader it might not be clear that you are talking about surface elevation values here. I suggest saying "The elevation value assigned to each cell is the average elevation value of all points that fall within that cell."
Line 172 Use either 1 m x 1 m or 1 m2. "one" instead of "1". Regarding statement of “points with horizontal coordinates” - Do you have points without any horizontal coordinates?
Line 173 Do you mean percentage of covered cells or percentage of entire grid by covered cells? I think you're trying to say that you calculated the percentage of the grid (the ROI) that was covered by lidar. Please rephrase for clarification!
Figure 7. Typically, darker colors mean "more" in a density map, so I'd suggest you invert the color ramp. Caption: replace "line" with "polygon". Changes in spatial coverage from what date? Maybe delete this part of the sentence.
Line 179 suggest simplifying to "photogrammetry-derived surface from a UAV"
Line 177 specify that you're talking about lidar coverage (not snow coverage)!
4.1 Spatial coverage of ROI -- Again, specify "Lidar spatial coverage of region of interest" (not to confuse with snow coverage!)
Line 187 “short-wave”
Line 197-198 This sentence is a bit clunky. I suggest rephrasing to something like "Due to the high temporal resolution of the lidar system, a short weather window during a snowstorm will allow the system to capture the area and provide current information about the snowpack.”
Line 199 add years “2023” and “2024”
Line 200 "a review of the camera images", replace “at these times” with “on these dates”
Line 203 suggest simplifying to "this coincides with the clearing of the lidar lens".
Line 207-209 Your observation of a difference in lidar reflectance during nighttime vs daytime is very interesting! Please provide some numbers here - these diurnal patterns, how significant are they?
Line 209-215 This section is discussing the results and should be moved to the discussion!
Line 212 replace “relation above” with "relationship"
Line 212-213 I'm also wondering if the less dense air caused by warmer air temperatures (though still subzero) were enough to scatter the laser pulses and cause a reduction in reflectance.
Line 217-219 Move this to a sub-section of Chapter 2 Monitoring where you describe the drone and how the data was processed.
Line 217 How was the comparison done? Did you compare the resulting DSMs or was it a point cloud comparison?, write out MP as "megapixel"
Line 220 Mean difference between the two?
Line 223 typo “artifacts”, "For example, " (comma needed)
Line 231 Maybe to clarify, say "towards the edges of the LiDAR swath at far distances away from the sensor" (an “edge” can be close too), clarify by saying “snow distribution pattern”.
Line 231-231 Here you state that there is disagreement between the models “where the pattern changes from ablation to accumulation, which can be due to errors in the alignment of the systems.” Was no horizontal alignment done between the models?
Line 236 replace with “small (D?) avalanche”
Line 237 be consistent in usage of “release area” vs “release zone”.
Line 240 suggest replacing with "100 m, though the length....", replace “ends” with “extends” or “reaches”
Line 241 suggest replacing “due to the limited field of view” with “due to its limited field of view”, replace “mostly also not” with “scarcely”
Line 241 suggest replacing “high angles of incidence” with “high incidence angles”
Line 243 suggest replacing “becomes visible with” with “is demonstrated by”
Line 245 suggest simplifying/clarifying a bit “This captured avalanche data" (....) "will become useful input data in future avalanche modeling in the Wildi avalanche path."
Figure 9. Add at the end of caption "Ski and snowboard tracks are visible outside of as well as cutting through the monitored area."
Figure 10. suggest putting the letters a, b , c in the upper left corners of each figure and put white circle behind for better visibility, clarify in (c) that it’s “elevation difference”
Line 252 typo “was”, "west wind” (no hyphen)
Line 253 suggest replace “difference of snow depth” to “redistribution in snow depth”
Line 255 You have the opportunity here to analyze the interaction between topography and snow distribution – I suggest adding a sentence that describes how this redistribution (erosion and deposition) also was a result of topography within the ROI, such as erosion from taller small sub-ridges and deposition in depressions/small gullies. Suggest adding a sentence at the end emphasizing your point "The clear westward pattern of eroded and deposited snow agree with the recorded wind direction on this day."
5 Discussion
Suggest starting the Discussion by adding a sentence that states which points you are going to discuss. “Here we will discuss (1) …., (2).. etc.”
Line 257 "state-of-the-art"
Line 259 suggest "lidar technology"
Line 260 suggest adding "that are limited by day light"
Line 261 “significantly”
Line 264 “have given”, delete “proposed”
Line 265 add “lidar” to “system”, please be consistent in using the term “UAV” vs. “drone” (choose one or the other!)
Line 266 What do you mean by “potential”?
Line 269 suggest replacing “results” with “products”
Line 271 “small-scale”
Line 272-273 suggest rephrasing " The high temporal resolution of the lidar system enables us to capture an avalanche event by measuring the snowpack surface shortly before and after an event.", remove “therefore”, switch to “derived immediately”
Line 275 suggest rephrasing "Some meteorological conditions can hamper lidar acquisitions, resulting in a reduction of lidar transmission and reflectance."
Line 277 replace “small weather window” with “short weather windows”, suggest replacing with "which offers a valuable alternative for when other methods (e.g. UAV flights) would not be feasible."
Line 279 Suggest rephrasing "Finding a suitable location for installation of the system might be challenging in some areas."
Line 284 clarify "campaigns" with "snow sampling campaigns"
Line 285 suggest also adding "and offers a great monitoring solution for agencies that have limited monetary resources"
6 Conclusions
I'd suggest starting the conclusions paragraph by restating the motivation and purpose of the project.
Line 286 suggest replacing “proposed” with "presented"
Line 289 replace “from” with “since”, suggest replacing “With the analysis of the data” with "Based on the first three operating months.."
Line 290 remove “limitations of the system” unless you’ll give examples!
Line 291 suggest replacing “safety management at avalanche prone traffic routes” with "avalanche safety decision-making for infrastructure corridors".
Line 295 spell out "decimeter"
Line 301 suggest removing “Nevertheless”. Add a sentence that says something like "The first three months of system operation have also revealed some areas of opportunity for improvement."
Line 306 re-wording - nearby? adjacent?
Line 308 "small-scale", "wind-drifting"
Line 310 suggest replacing “state-of-the-art” with “current”, add “that are mainly”. Regarding “its benefit”, I’m wondering, would you provide all the data, such as providing real-time access to the monitoring stations, or would you compile a product based on the data that you’ve collected, or maybe both?
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 182 | 48 | 11 | 241 | 7 | 4 |
- HTML: 182
- PDF: 48
- XML: 11
- Total: 241
- BibTeX: 7
- EndNote: 4
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1