the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 22 Dec 2025
Analysis of factors influencing rockfall activity with a new method to estimate the release frequency in case of scarce data
Abstract. Understanding the relationship between rockfall release frequency and volume is essential for quantitative hazard and risk analyses. The volume–frequency relationship is usually modelled with a power law distribution determined by an exponent and an activity parameter that can be estimated based on volume distributions from rockfall event inventories. Here, a new method is presented, which allows for the estimation of a confidence interval for the activity parameter, even if the inventory contains a small number of events. It was applied to estimate the spatio-temporal frequency of failures larger than 1 m3 for 22 natural mountain rock walls located in temperate climate zones and below the permafrost area. The obtained frequencies were similar to those given by power law fitting of the inventories. The meta-analysis showed that these frequencies vary by at least three orders of magnitude from 0.01 to 20 yr−1.hm−2, depending mainly on the geomorphological context of the cliff, the homogeneity of the cliff and the joint spacing. The exponent of the power law varies between 0.3 and 1.0 and tends to be lower for massive rock than for bedded rock. A primary classification is proposed that enables an estimation of the volume–frequency relationship parameters when no inventory is available. More inventories are needed to enhance this classification.
- Preprint
(1184 KB) - Metadata XML
-
Supplement
(10351 KB) - BibTeX
- EndNote
Status: open (until 13 Apr 2026)
- RC1: 'Comment on egusphere-2025-6247', Anonymous Referee #1, 09 Feb 2026 reply
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 272 | 163 | 26 | 461 | 98 | 50 | 47 |
- HTML: 272
- PDF: 163
- XML: 26
- Total: 461
- Supplement: 98
- BibTeX: 50
- EndNote: 47
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
This scientific article presents several interesting and relevant ideas for analyzing rockfall activity. In particular, it offers an improved estimation of failure probabilities for low-frequency events, as well as insight into the relationship between joint spacing and the A‑factor of power‑law distributions. However, in its current form, the proposed approach is very difficult for the reader to follow and understand. Certain explanatory elements seem to be missing or are not presented with sufficient clarity. As a result, it is difficult for readers to assess whether the interpretations are solidly supported by the data.
Considering the potential originality of this contribution, the paper is fully aligned with international standards and with the scope of NHESS. Nevertheless, its structure and presentation require substantial revision. In its current form, the text appears to be composed of several components whose interconnections are difficult to discern. The authors may consider cutting non‑essential elements in order to better highlight the key points.
MAJOR CORRECTIONS
In the text the term “spatio-temporal frequency “ points sometimes to Fst and sometimes to Ast. Is there a way to better distinguish these two parameters to help the reader?
Paragraph 2.3 describes a new way to estimate the mean failure frequencies. But it is not clear if and how it is used later in the paper. Is it something completely disconnected from the rest of the paper or used in Ast calculations?
From line 205: It is not clear how the different Ast (obs, min, max, mod) are calculated. For instance, I tried to calculate Ast-obs based on table 2 and 3 for the two first lines, dividing the number of blocks >1m3 by the sampling extent of table 2. But the results do not fit the AST-obs of table 3. All the calculation of Ast parameters must be clearly explained (obs, min, max, mod); presently I am not able to understand where these numbers are coming from.
L210-215: We don’t have access in the paper or supplementary material to the data from the literature used in the compilation of tables 2 to 4. As one major limitation of this contribution is the joint spacing assessment, why use the ISRM81 (pre-lidar) intervals? original data may be better than this rough classification? In particular since the use of point clouds. Please discuss.
Figure 5: this figure looks completely useless. Generating data by Monte Carlo simulations to extract a linear regression does not really make sense. In addition the joint spacing is a discrete variable (2 classes), which makes a very weird distribution of points. That would be much more informative to have the mean value with error bars for each dataset. And the final regression coefficient should not be very different than eq10 anyway.
L480-482: I do not understand how you reach this statement based on table 8. Please develop.
L499-502: a new idea (pseudo-RQD) is introduced here but not really explained, and it is not explained how it is used. Then you can properly develop it, or remove it and keep it for a future paper (it looks like a good idea, I am looking forward to see real results when you get them).
MINOR CORRECTIONS
L78-81: “garbage” text to be removed
L85-86: fuzzy statement (“appropriate”, “sufficient”) that does not provide any information – rephrase or remove this sentence and correct the typo “volumethat”
L102 -Eq1: “Poisson(P(…))” Check the equation: is the “P” needed compared to eq(2) ?
L110 – Eq3: explain the factor 1 at the end of the equation. The rate ?
L113: in Erlang function, when the rate = 1, then mean= standard deviation. Is there anything to say about this property in the context of rockfalls?
L126: “smaller” than what ? complete the sentence
L132-133: “the mean frequency …” Not clear. Better explain why that can be of interest.
Table 1: Is there a reason why the number of significant figures is not consistent in the table?
Tables 2 to 4: explain the meaning of colors in the legend of table 2 (types of rocks) - if any colors are possible for tables (?)
Figure3: how is the 95% confidence interval calculated ? based on what ?
Line 300: intervals of what? complete sentence
Line 379: explain “growth disturbances”
L544: why uppercase letters ?
L605: extra-line to remove