Comments on Mayer et al.

The most necessary but lacking studies on laboratory frost weathering is to detect and visualize crack growth in low-porosity high-strength rocks that compose high mountains. Mayer et al. challenged this difficult subject using AE monitoring and CT scan techniques, and presented novel, valuable results from laboratory freeze-thaw experiments. I was particularly impressed with the great efforts to quantify the internal crack growth and its association with theoretical models. In this context, Meyer et al. provides a timely and valuable contribution to periglacial geomorphology and engineering geology. However, the paper has a number of problems that should be solved or improved before acceptation. The major problems are inappropriate laboratory settings, limited number of rock samples, visualization of crack growth, and the gap between the experiments and natural conditions, as listed below in detail. The title is also misleading that the paper presents results of field monitoring in high mountains, but actually only the rock samples were derived from a mountain.

General comments

1. Problems of laboratory settings: The temperature at the sample bottom was used in the analysis, but this ignores the thermal gradient in the sample. If crack growth occurs near the bottom, the analysis is acceptable. When the sample bottom dropped below the freezing point, perhaps most of the scanned area has not yet been frozen, as the side sensor located just above the scanned area never showed subzero temperatures. Thus, the bottom temperature may be much lower than the temperature at which cracking is most active. Another problem is the lack of an external water reservoir, which, if present, might supply water continuously and cause progressive crack growth in a sustained freezing condition.
2. Limited number of experimental runs: Only one sample was used for each moisture and freeze-thaw regime (6 samples in total), despite significant variability of crack conditions. In pore-supported rocks, samples are assumed to be relatively homogeneous, but in crack-supported rocks, samples should be inhomogeneous as indicated in the CT images (Fig. 5) and crack volume profiles (Fig. 6). For such inhomogeneous rocks, multiple samples are required to guarantee the reproducibility and to evaluate the effect of moisture levels. In fact, Figure 6 seems to suggest that crack growth is almost independent of saturation but dependent mainly on the combination of the initial crack volume and saturation. I am also concerned about the accuracy of saturation level, because for the low porosity rocks difference in the weight of water is very small between the three saturation levels. Multiple experimental runs are required also to overcome the inaccuracy of the saturation levels.
3. Visualization of crack growth: Crack growth is not visible in CT images (Fig. 5), although significant volume increase (30-50 %) is computed from the scanning results. Do you have more clear CT images showing crack extension/widening, or can you explain the reason for the invisibility?
4. The gap between the experiments and natural conditions: Instant cooling to -10 degrees used in this study is unusual in natural rockwall conditions. AE events during such a rapid cooling (and warming) are certainly attributable to thermal stress, but is it applicable to weathering in natural conditions? Furthermore, the experimental condition lacks external water supply (i.e., moisture source is confined within the rock sample), while in natural conditions water can be supplied from more distant areas.

In summary, the experiment itself includes very new and interesting results, but as it stands the experiment provides very limited information for understanding weathering in high mountain rockwalls. Further experiments under more suitable conditions and with raised reproducibility, or more reliable discussion to field problems are necessary.

Specific comments

Section 1

Figure 2a: Are the minor grids on the time axis shown at an interval of 12 days? 

Figure 2, caption: Are these curves drawn on the basis of subdaily-scale temperatures or daily mean values? Please add the recording intervals.

L101-102: …laboratory freeze-thaw tests never demonstrated… →Duca et al. (2014) showed crack formation in granite, although they simulated not freeze-thaw but sustained freezing. Wang et al. (2020) also showed F-T induced crack extension in fractured granite with AEs and CT scan.

L105: while monitoring… →with monitoring…?

Section 2

L113: 10 X 5 cm →10 cm in length and 5 cm in diameter?

L135-136: What is the accuracy of the weight? In 0.1 % porosity and 9.2x2.6 cm sample, my simple estimation indicates that weight of water is 0.13 g, 0.09 g and 0.04 g, respectively for 100, 70 and 30 % sample. How did you control such a small difference in water content?

L138: What kind of material does compose the insulating holder?

L140-141: The definition of ‘open system’: Water supply is confined within the rock sample (lacking external water supply), while in natural conditions water can be supplied from more distant areas? In this respect, this experiment might simulate a ‘closed system’ condition.

L142-143: Regarding the temperature cycle used in FT-1: 1) Did rock samples thaw completely during such a very short period? Do you have any evidence? 2) Can the results from such an extreme temperature condition be applied to natural rockwalls?

L201: in between →between?

L215, Eq. (2): Please define n and t in the text.

L216: The crack… propagates crack growth V →The crack… propagates at a growth rate V?

L224: 37.1 m s^-1: According to Table 1, gamma is dimensionless.

Section 3

L242: The upper temperature sensor →‘The top temperature sensor’ may be better.

L243: 100% sample →100% saturated sample

L244-248: To identify AE events due to thermal stress, it may be important to distinguish AEs during rapid temperature change and those during slow change.

L250-252: If the potential AE events during the data gap period is taken into account, the highest AE activity could have occurred between scan 2 and 3 in the 70 % sample?

L255 (Figure 4, caption): c-d) Please explain dots (each event?) and lines (cumulative?).

L258 (Figure 4, caption): Blue rectangles →’Blue backgrounds’ or ‘colored backgrounds’ may be more appropriate?

L262: Do you calculate ‘thermal- and ice-induced stresses’ at the bottom (highest stress?) or at the scanned depth (lower stress)? Is the ice stress given by which equation?

L273-274: Use consistent expression as ‘100 % saturated, 70% saturated...’ instead of ‘fully saturated, partly saturated...’.

L282: You wrote ‘volume changes occur in form of crack growth (Fig. 5)’, but crack growth is not visible in Fig. 5. Do you have more clear CT images?

L283-284: The phrase ‘crack growth revealed...’ seems inappropriate. Do you mean "crack growth positively correlated with initial crack volume"?

L286: Insert ‘and’ between the two adverbs (i.e., uniformly and independently)?

L289: Regarding ‘the amount of crack volume growth per scan varies with saturation’, Figure 6 implies that crack growth is almost independent of saturation, but dependent mainly on the combination of the initial crack volume and saturation.

L290-292: 30-50 % crack growth is significant, but why is such a crack growth unclear in Figure 5?

Comment on Section 3.2.1: You may also refer to the acceleration of crack growth for the 70 % sample in the later period and deceleration for the 100 % sample, which seem to be consistent with AE activity?

Figure 5: If crack extension/widening is visible, please indicate it with an arrow.

L314-315: This result suggests that, also in FT-2, the initial crack volume is a more important factor than the saturation?

Section 4

L324-326: The sentence ‘This pattern…’ is redundant. Perhaps it can be rewritten more concisely.

L326-327: Again, Figure 5 does not show crack growth clearly.

L331: Who are ‘the authors’: the authors of this paper, all authors cited in the preceding sentences, or only De Kock et al.? You also used ‘the authors’ in L389, L444 and L455 probably for those of the reference just cited, while ‘the authors’ in Acknowledgements seem to indicate those of this paper. Please keep consistency.

L351: Correct grammar such that ‘Prick’s empirical findings suggested that…’ or ‘Prick empirically found that…’.

L354: ‘low-porosity’ instead of ‘low porous’?

L359-360: But when the sample bottom dropped below the freezing point, perhaps most of the scanned area has not yet been frozen? Thus, a possibility that volumetric expansion in the scanned occurs later cannot be ruled out?

L362-364: This interpretation is acceptable for the present experiment, but does this result based on ‘instant cooling (favorable for thermal cracking)’ is unlikely applicable to natural rockwall conditions?

L366-368: But the location of crack growth may not be at the bottom of rock sample where the temperature is recorded. Since the side sensor located just above the scanned area never showed subzero temperatures (Fig. 5), most of the scanned (cracked) area never reached below -6 degrees?

L369: For the above reason, the result presented here cannot be compared with Hallet et al. (1990)?

L397-398: This sentence seems misleading. If you compare the results per F-T cycle (i.e., summed AE hits or crack volume fraction divided by the number of F-T cycles) between the two conditions, sustained freezing seems to have higher efficacy. Furthermore, F-T cycles with instant cooling to -10 degrees may be unusual in natural conditions. Freeze-thaw tests with milder cooling bay be preferable to apply to natural conditions.  

L405: You may add Duca et al. (2014) and Wang et al. (2020) for the examples of laboratory frost cracking.

L409-410: I cannot understand the sentence ‘This pattern…’.

L413-414: I think that ‘eight times higher ice stresses in the sustained freezing‘ seems consistent with higher cracking activity for one F-T cycle (see the above comment on L397-398).

L421-422: I support this interpretation. If you provide water from the top during freezing, ice segregation may occur more effectively (cf. Duca et al., 2014)?

Figure 8, label of the horizontal axis: What is ‘Summed Ice Thermal’? Do you mean ‘Summed Thermal Stress’?

L445: The phrase ‘an increase of mean annual rock temperatures…’ may lead to misunderstanding that warming progresses year by year. 'Higher temperature' may be more appropriate.

L465-468: Time scale for warming should also be considered: e.g., can 1-2 degrees of warming lead to significant changes in snow cover and number of freeze-thaw cycles above 3000m?

L470: Do you mean 'increased debris-flow activity due to increased sediment availability at high elevation sites'?

References

Duca, S. et al., 2014. Feasibility of ice segregation location by acoustic emission detection: A laboratory test in gneiss. Permafrost and Periglacial Processes, 25, 208–219, https://doi.org/10.1002/ppp.1814.

Wang, Y. et al., 2020. Rock bridge fracturing characteristics in granite induced by freeze-thaw and uniaxial deformation revealed by AE monitoring and post-test CT scanning. Cold Regions Science and Technology, 177, 103115. https://doi.org/10.1016/j.coldregions.2020.103115.

Dear Editor,

I have reviewed the paper "Quantifying frost weathering induced rock damage in high alpine rockwalls" by Mayer et al. with great interest. The authors address a highly relevant research gap: frost damage in low-porosity and crack-dominated rocks. They present a first attempt to investigate frost damage in low-porosity, high-strength rocks in the laboratory, combining AE monitoring and micro-CT scans. A frost-cracking model complements the data acquired in the laboratory.

The authors present a novel and valuable study, and the manuscript contains interesting and relevant findings. However, the manuscript still has fundamental weaknesses that must be addressed before it can be accepted (selected major concerns are listed below). Further, the results are over-interpreted, and the manuscript is generally not written precisely and self-critically enough. An example is already the title: (1) it is a first attempt to quantify and not a general quantification (too strict), (2) low-porosity limestone was investigated (not generally frost weathering induced rock damage), and (3) it is laboratory experiment with samples from the field (no field experiment, only the implication is for high alpine rock walls; whereby this transfer, in particular, has a weak link to the results).

Major concerns

1) I have a few concerns regarding the setup and representativeness of laboratory experiments.

• First, Figures 1 and 2 are misleading because the authors show selected alpine rock walls and related temperature time series. Still, the thermal conditions in the laboratory experiments do not correspond to reality (neither the gradients nor the duration, which are, besides the water availability, crucial for ice segregation). Further, the timely changing thermal conditions (i.e., periods with non-linear temperature gradient in the sample) driven by the cooling plate at the bottom are not adequately considered (no uniform temperature gradient from bottom to top and lateral temperature gradient from inside to outside are ignored).
• Second, the number of samples is minimal (only 1 sample/experiment for each condition), does not allow such firm conclusions, and certainly not for a direct implication to the real world.
• Third, the coupling between the AE sensor and the rock is crucial, but there is no proof of how good it is and if it is comparable between the experiment's start and end. For example, how do you explain the change in the yellow slope in Figure 4c? Further, a direct comparison between two AE time series must be interpreted carefully, and the threshold level might need to be adjusted. For example, if you normalized the summed AE hits in Figure 7, I would expect that they all have a similar pattern (repeating in FT-1, comparable even to FT-2). Therefore, at the moment, the AE results are over-interpreted, and a more in-depth evaluation is required.
• Fourth, I'm very fascinated by the micro-CT results. I wonder if you saw similar patterns at other locations (e.g., vertically above). Nevertheless, the comparison between the scans seems to be not very sophisticated/quantitative (there are many approaches from photogrammetry for quantifying changes …).
• Fives, how do you assess the scalability and transferability to real-world conditions? There is no critical discussion on this with consideration.

2) The frost-cracking model strongly depends on parametrization and assumptions. Please make a sensitivity analysis, allowing you to visualize the output probabilistically. Publishing the code would certainly give more confidence in the model – I strongly recommend it!

3) The readability of figures is limited due to resolution, font size, line width, and color scheme. Further, I strongly recommend highlighting the key information in the graphs or mentioning it at least in the caption.

4) The manuscript is rather lengthy, especially the introduction. Please reorganize, restructure, and shorten the introduction to avoid repetitions. Section '3 Results' is rather hard to read – please add more explanation and interpretation.