the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 27 Jun 2025
Geological-geotechnical analysis of a rock-toppling prone canyon in Furnas, Brazil, after a fatal event
Abstract. When a disaster related to a natural phenomenon occurs in areas dependent on Geotourism, restoring tourist confidence can be a challenge. An example is the rock-toppling event that occurred on January 2022 in one of the four canyons located in the Furnas reservoir in Brazil, which caused 10 fatalities. Visitation fear and their temporary closure severely impacted the economy of the surrounding municipalities that rely on tourism. To support a safer operation of the canyons to visitation, our study investigates the factors that can lead to landslide events in the region, based on the combination of field investigations, rock-mass quality evaluation (RMR14) and kinematic analysis. We hypothesize that the assessment of rock-mass quality can successfully identify specific areas in the bedrock that are prone to rockfall and rock toppling, supporting risk management strategies in tourist regions. Our results indicate that the 2022 rock-toppling event occurred due to the combination of different factors, such as rainfall infiltration in the unfavorably-oriented joints of the bedrock and reservoir water-level fluctuations. Moreover, the long-term erosion at the base of the slope, caused by the nearby waterfall flow and water-level variations, weakened rock-mass support. The RMR14 method adapted to open rock slopes successfully supports the estimation of the bedrock’s geomechanical properties, identifying structural zones in the rock mass that are prone to slope failure(s) and, as a consequence, should be monitored. The kinematic analysis further indicates that the four canyons are highly susceptible to planar failures and, less so, to toppling, although specific locations in the slopes show a higher rock-toppling susceptibility, especially where two perpendicularly-oriented fault zones (NW-SE and NE-SW strikes) intersect. The consideration of geomechanical properties in hazard evaluation is recommended as a risk management strategy, supporting the delimitation of regions near the rock slopes that should be restricted and of specific portions in the bedrock that should be retained. Our study was fundamental to establishing visitation procedures in the canyons, so that tourists and workers are more protected and aware of the existing geohazards.
- Preprint
(3256 KB) - Metadata XML
-
Supplement
(2187 KB) - BibTeX
- EndNote
Status: final response (author comments only)
- RC1: 'Comment on egusphere-2025-1248', Anonymous Referee #1, 31 Jul 2025
-
RC2: 'Comment on egusphere-2025-1248', Anonymous Referee #2, 04 Aug 2025
Dear Editor,
Please find the review of the paper egusphere-2025-1248: Geological-geotechnical analysis of a rock-toppling prone canyon in Furnas, Brazil, after a fatal event
By
Cabral et al.
This paper examines a rock slope failure that occurred along an artificial lake on January 8, 2022, resulting in the deaths of ten individuals. The incident involved a rock column falling onto boats carrying tourists, as the lake is a popular destination.
The paper presents a conceptual model describing the rock-toppling event and its contributing factors. Using the RMR14 classification, the authors evaluate the likelihood of slope failure and provide examples of this rating system.
General remarks
Such a paper has a particular status: it is not only an academic work, but also a document that could potentially be used in court if the case goes there. Therefore, the reviewers also act as experts.
In this sense, the paper is more akin to a report than a scientific paper. The technique applied is not innovative, and in my opinion, the quality of the assessment is inconsistent and poor.
First, the analysis of the January 2022 event is weak and lacks detail, because it is important to understand the 2022 event in detail in order to develop correct regional susceptibility mapping.
Readers would like to know if the kinematic criterion for toppling was met because toppling can be caused by the deformation of a weak base of a column, which progressively meets the conditions for toppling. A precise cross-section of the geology of the fallen column must be provided, including the locations of the discontinuities. The triggering and worsening factors must be distinguished, even if they have the same origin. The increase in water pressure is also questionable because no exceptional precipitation and a relatively low reservoir level were measured during the event. These conditions can be considered triggers for failure in highly fractured rock, but this must be argued. Additionally, the topography of the upper part of the cliff must be detailed, as it can concentrate water through overland flow, which can then enter the rock cliff.
A real 3D model of the column and structure must be presented to fully understand the mode of failure, which is important for further analysis. The volume of this column has not even been estimated.
In my opinion, using only an average stereonet for slope orientation is no longer relevant in such a case. This real 3D data must be used for the kinematic test. Additionally, why are overhangs not explicitly studied as seen in Figure 13b?
A susceptibility map must also be produced, as well as an illustration of the identified structures that pose a problem for stability.
Additionally, the term "landslide" can be replaced with "rock instability."
Specific Comments
Line 59: I'm not sure if this is useful. Either way, you said that tourism increased.
Line 66: What do you mean by "textural (2D)"?
Line 89: Why is the county surface area relevant to the paper?
Line 92: Provide numbers.
Line 93: What are the criteria and numbers for overtourism?
Line 108: I guess you mean paragneiss.
Line 109: Quartzite is usually very weak due to weathering and fracturing.
Use "rock instability" instead of "landslide."
Lines 145 and 152 do not have the same definition of JCS.
Line 159: And so what?
Line 187: GNSS RTK with a base?
Figure 4 is too fuzzy.
Lines 266–267 have no arguments.
Line 272: What can be deduced if the aperture can reach 50 cm?
Figure 6: As shown, the rainfall cannot fill the back joint. There must be a way to bring water there: overland flow or the water table level.
Figure 7: Add a stereonet and illustrate the discontinuities that play a role in the picture.
Lines 310-312 are unclear.
Figure 9: The same remarks apply as for Figure 7.
Table 5: How is φb calculated? Not in the formula or graphs. By the way, it is extremely high!
Line 375: Which discontinuity sets have the largest spacing?
Lines 401 and 413: How are these movements detected?
Line 402: This invalidates the approach of the average steronet slope orientation.
Tables 8 and 9: You need to explain the percentages.
Figure 13: The same remarks apply as for Fig. 7. How do you get pressure with an aperture of 50–60 cm?
Lines 473–476 are not demonstrated in the paper.
Citation: https://doi.org/10.5194/egusphere-2025-1248-RC2
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 466 | 60 | 15 | 541 | 45 | 12 | 24 |
- HTML: 466
- PDF: 60
- XML: 15
- Total: 541
- Supplement: 45
- BibTeX: 12
- EndNote: 24
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
The paper presents a comprehensive geological-geotechnical analysis of the rock-toppling event in the Furnas reservoir canyons, Brazil, which resulted in 10 fatalities in 2022. The study combines field investigations, rock-mass quality assessment (RMR14), kinematic analysis, and UAV-based photogrammetry to identify predisposing and triggering factors of slope instability. The research is well-structured, methodologically sound, and provides valuable insights for landslide hazard management in geotourism-dependent regions. The findings are supported by extensive field data, analytical methods, and clear visualizations. The study has practical implications for risk mitigation strategies in similar environments. Some suggestions are recommended for improvement.
Major suggestions:
Minor suggestions: