the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
Seasonal and diurnal freeze-thaw dynamics of a rock glacier and their impacts on mixing and solute transport
Abstract. Rock glaciers play a vital role in the hydrological functioning of many alpine catchments. Here, we investigate seasonal and daily freeze-thaw cycles of the previously undocumented Canfinal rock glacier (RG) located in the Val d'Ursé catchment (Bernina Range, Switzerland) and the RG's influence on the dynamics of the hydrogeological system. We combine digital image correlation techniques, geochemical and isotopic analyses, time-series analysis, and hydrological monitoring to understand the functioning of the hydrological system. An acceleration of RG creep since 1990 has occurred, with the most active regions exhibiting horizontal velocities of ~1 m/yr. Distinct geochemical signatures of springs influenced by RG discharge reflect contrasting and temporally-variable groundwater mixing ratios. A novel application of frequency-domain analysis to time-series of air temperature and spring electrical conductivity enables a quantitative understanding of the RG thaw and subsurface flow dynamics. A gradual decrease in time-lag between air temperature maximum and spring EC minimum, caused by dilution from RG ice melt, is observed over the snow-free period, implying progressively shorter residence times. Through our multi-method approach, we develop conceptual models for RG-influenced alpine hydrogeological systems on daily and seasonal time-scales.
- Preprint
(14298 KB) - Metadata XML
-
Supplement
(1741 KB) - BibTeX
- EndNote
Status: final response (author comments only)
-
CC1: 'Comment on egusphere-2024-927', Giacomo Medici, 23 May 2024
General comments
Novel research in the field of hydrology. The manuscript needs some minor corrections that should improve the final version of the manuscript. See below the specific comments.
Specific comments
Lines 14-68. The link between creep and groundwater flow is an unexplored topic. I would emphasize more this point in your introduction/discussion.
Lines 27-28. You mention snowmelt and groundwater flow in the introduction and the conceptual model. Please, expand this point and add recent literature on snowmelt aquifer recharge in mountain ranges that combines isotope analysis and monitoring:
- Lorenzi, V., Banzato, F., Barberio, M. D., Goeppert, N., Goldscheider, N., Gori, F., Lacchini A., Manetta M., Medici G., Rusi S., Petitta, M. 2024. Tracking flowpaths in a complex karst system through tracer test and hydrogeochemical monitoring: Implications for groundwater protection (Gran Sasso, Italy). Heliyon, 10(2).
- Stevenazzi, S., Zuffetti, C., Camera, C. A., Lucchelli, A., Beretta, G. P., Bersezio, R., & Masetti, M. (2023). Hydrogeological characteristics and water availability in the mountainous aquifer systems of Italian Central Alps: A regional scale approach. Journal of Environmental Management, 340, 117958.
Line 68. Disclose the specific objectives of your research by using numbers (e.g., i, ii and iii) at the end of your introduction.
Line 73. “Mostly”. Please, specify the other lithologies. Alternatively, you can also fix the issue by deleting the vague term “mostly”.
Line 77. “fractured aquifer”. Insert more detail on the nature of the tectonic structures and joints. Thrusts and folds? Also normal faults? Unclear the nature of the fault zone in the conceptual model.
Line 280. Specify the area of the French Alps and the lithologies of the fractured bedrock aquifer there. Crystalline basement there?
Lines 397-550. Take into account the literature suggested above.
Figures and tables
Figure 2. Insert an approximate spatial scale.
Figure 6a. Do you need to add an equation and parameters (R2) to the line?
Figure 8. Please, add the intermediate months on the horizontal axis.
Figure 10. Insert the spatial scale and specify if there is vertical exaggeration.
Figure 10. Unclear the nature of the fault zone. Normal fault, or thrust with vertical exaggeration? This point is unclear even by reading the text.
Figure 10 vs. Study Area and instrumentation. You need to provide more detail on the tectonic structures on the paragraph 2 to make clear the final conceptual model.
Citation: https://doi.org/10.5194/egusphere-2024-927-CC1 - AC1: 'Reply on CC1', Landon Halloran, 27 Aug 2024
-
RC1: 'Comment on egusphere-2024-927', Anonymous Referee #1, 13 Jun 2024
General Comments:
The authors present an interesting study of the Canfinal Rock Glacier in the Swiss Alps. They investigate the factors that have contributed to the rock glacier’s flow and how the rock glacier contributes to spring flow during different times of year. The study was interesting, novel and I appreciate the use of multiple methods/lines of evidence to characterize the hydrological dynamics.
My suggestions for improvement mostly focus on increasing context and detail in places. For the hydrochemistry results, more graphical representation of the hydrochemical signatures of different sample types and locations is needed. They are currently presented in PCA form in Figure 6 but the different waters (springs, streams, etc.) are not differentiated. Were any samples taken from the well? That would be interesting to see in comparison to the other samples. Consider adding an additional panel to Figure 6 to show EC versus some other ion with the sample types coloured or shaped by water type (spring, groundwater, stream, etc.) or location. Or else in panel B, you could make the points different shapes for different types of samples (although that may get too busy and be less clear). That way, the reader can easily see how the different samples and presumed end members relate to each other.
In general, some additional explanation of the frequency-domain analysis methods and results would be helpful for readers who are not particularly familiar with these techniques. Around line 133, a conceptual statement about how the time-domain analysis is going to be interpreted would improve clarity. Consider adding a diagram to help clarify the phase shifting described on lines 148-150. In the results, the frequency-domain analysis results are not always intuitive, so some additional contextualization in text or annotation of the plots would likely help readers follow.
Specific Comments:
39: Briefly identify the other sources of water released from rock glaciers other than meltwater.
51: The hypotheses are nicely presented. Perhaps you could clarify that these hypotheses are not mutually exclusive. I.e., “…several hypotheses (which are not mutually exclusive) have been proposed…” if that is the case.
63: Could you add a sentence as to why this is the case?
64: The objectives are stated in the last paragraph of the introduction, but they come after mentioning the methods and the word objective is not used. I suggest stating 2-3 numbered objectives for maximum clarity.
121: Some basic details around sampling/analytical procedure (e.g., bottles, preservation, analytical equipment) would be expected here.
124: At this point, it’s not clear what the correlation analysis is used for. Some explanation of the bigger picture is needed.
Figure 4: Neat figure!
158-169: These data sources and methods should be included in the methods section. Are there any limitations associated with ERA5 performance in mountains that should be acknowledged?
Figure 7: There are blocks in the EC data for April for S1, August for S2, January to March for S3 where EC is jumping between a certain value and 0 many times. Does that represent some kind of sensor error? Or is the spring going dry and reactivating in quick succession? This should be explained in the text.
Figure 10: I suggest adding a legend entry for the light grey geologic material since all others are labeled (seasonally frozen talus hosting perennial ice lenses?). Also, why does the hillslope-scale flow line have such a bend? I found the captions “freezing conditions” and “thawing conditions” a little unclear and suggest simply “winter” and “summer” might be more intuitive.
Technical Corrections:
188-189: The phrasing of this sentence is awkward, consider re-phrasing.
201: Should be “…this sampling campaign…”
Citation: https://doi.org/10.5194/egusphere-2024-927-RC1 - AC2: 'Reply on RC1', Landon Halloran, 27 Aug 2024
-
RC2: 'Comment on egusphere-2024-927', Anonymous Referee #2, 14 Jun 2024
GENERAL COMMENTS
The manuscript by C. Louis, L.J.S. Halloran, and C. Roques provides an interesting and for the rock glacier research community novel hydro-chemical characterization of the previously uninvestigated Canfinal rock glacier and its surrounding springs in the southeastern Swiss Alps. I discuss the manuscript along its three storylines: (1) long-term kinematics and its relations to selected climatic drivers, (2) seasonal hydro-chemical (electrical conductivity EC, stable isotopes, major ions) characterization of several springs below the rock glacier, and (3) diurnal frequency-domain analysis of the EC of the rock glacier outflow. Finally, I have some suggestions for Fig. 10.
First, the kinematic investigations, limited to a multi-year time scale by the available imagery, are interesting and well in line of the observations of the Swiss Permos Monitoring Network. Perhaps sufficient for the hydrological storyline would be the delineation and rough characterization of the rock glacier material (ice content) via the kinematics (L258–265) in support of the morphological evidence of ice-rich permafrost occurrence. Due to the scale mismatch, the relations between multi-year climatic and kinematics trends are hard to connect to the seasonal to daily/hourly hydrological analysis, although links between hydrology and kinematics undoubtedly exist. Still, keep it in the manuscript since it gives clues on the thermal state and provides valuable baseline kinematic observations on a previously uninvestigated, unknown site.
Second, the seasonal hydro-chemical characterization enabled the seasonal differentiation of water sources contributing to the springs throughout the catchment. This was all very convincing and relevant. The spatio-temporal clustering of surface water chemistry (Figs. 6, S2) at small catchment scale reveals the distinct chemistry and “imprint” of the rock glacier compared to the vegetated plain “La Casina”. The Canfinal borehole provides unique insights into the snow and permafrost interactions with groundwater in a thermally sensitive environment close to the lower permafrost limit. Concretely, Fig. 9 shows a link between ground water storage changes (trend of hydraulic head KB4) and EC S1, itself related to precipitation events and time elapsed since snowmelt.
Third, the diurnal frequency-domain analysis of EC is potentially a useful tool in shedding light on the timing of water and heat transfer and applied to the first time in alpine permafrost (to my knowledge). This is an important contribution. I think however that far-reaching interpretations on freeze/thaw cycles based on minuscule 0.5–2 µS/cm oscillations of the EC signal are presented too boldly: Low EC is associated with high discharge (L322) and linked to intense ground ice melt (L324) supposedly driven by diurnal temperature oscillations. Such a behavior might be more typical for (debris-covered) glaciers, but it is not typical for permafrost and rock glaciers. For the lack of independent, local measurements to corroborate these links, this remains a hypothesis and should be framed more cautiously. Please address:
- No discharge observations are presented to corroborate the presumed (admittedly common) negative EC–discharge relation with discharge maxima in the afternoon. I am aware of the difficulties of obtaining a water level–discharge relation in such terrain. Given your repeated field visits: Do you have water level measurements/visual observations that would attest the afternoon high-flow at least qualitatively?
- The assumption of low-EC, “clean” ground ice whose melt dilutes the outflow is untested on the Canfinal rock glacier. I cannot require you to dig a sample, but it should be mentioned that ground ice in rock glaciers was found to have differing solute content. The (few) available measurements of the chemical composition of rock glacier ice (exposures, drillcores) range from low-EC ice (e.g., Murtèl, Haeberli (ed.), 1990) that would result in dilution to “dirty” high-EC ice (e.g., Lazaun, Nickus et al., 2023) that would result in solute enrichment (Brighenti et al., 2021; Bearzot et al., 2023). On top of that, in a degrading rock glacier, two types of ground ice melt in the thaw season: First the ‘young’ ice in the active layer (‘superimposed ice’) and later the ‘old’ permafrost ice (del Siro et al., 2023).
- Please mention that dilution from ice melt is not the only behavior found in the outflow of rock glaciers. EC is also not necessarily a conservative tracer that solely hints at water provenance (ice melt) in periglacial/permafrost environments. Colombo et al. (2018) lists contrasting mechanisms of solute export and EC-discharge relations, some are regular, some are tied to precipitation events or weather spells, or related to weathering (enrichment, dilution, flushing). Briefly discuss which processes you consider likely given the measured regular EC oscillations.
- No ground thermal data is shown to corroborate the diurnal freeze/thaw cycles. Due to the thermal inertia of the active layer, I strongly doubt that temperatures and melt rates at depths typical for ground ice melt in rock glaciers significantly vary on an hourly basis (certainly in the late thaw season when the ground ice table has receded to depth)! The melting ice must be at or near the ground surface, not deeper than a few tenths of centimeters (the penetration length-scale of diurnal oscillations). Could you get an idea of the active layer thickness on Canfinal? This reasoning rather hints at snow or shallow seasonal ice hidden in the rough terrain – on the rock glacier but also on the adjacent talus slopes and headwalls. This explanation would be consistent with seasonally (broadly) decreasing amplitudes of the S1 diurnal cycles (Fig. 8): waning influence of snowmelt. Nonetheless, the pattern of discharge inversely varying with EC and concomitant with peak air temperatures is also reported by Mateo & Daniels (2018).
My point is: Your hypothesis is just one chain of processes out of many thinkable ones! Considering the complexity, it is not possible to make all links robust in the scope of this publication. My suggestion is that you introduce the novel diurnal frequency-domain analysis more cautiously as a tool and frame your ice melt-dilution hypothesis as one example of the chains of processes that can be explored with this tool. EC is a commonly measured variable. Many past & future data sets can be analyzed!
Finally, Figure 10, the conceptual model sketch of the annual freeze-thaw cycles and its implications on groundwater flow. It is the first rock glacier hydrological model that focuses explicitly on their role in the entire catchments and brings up the permafrost interactions with deep groundwater flows. This is an important contribution. With a few modifications, permafrost and thermal aspects can be depicted more accurately, namely:
- The extent of permafrost: The rock glacier, as a permafrost landform, is also in summer frozen (cryogenic, ≤0°C), hence must be enclosed by the 0°C isotherm in both panels. Vice versa for the ‘unfrozen till layer’.
- Time/spatial scales of freeze/thawing: Only the active layer, the uppermost ca. 3–10 m beneath the surface, is subject to annual freezing/thawing. Thermal changes at depth are slow. A pervasive freezing/thawing of the bedrock with large changes at depth as depicted is not possible on a seasonal scale, the sketch rather evokes a long-term (decadal) permafrost degradation. Also, adding a spatial scale would help to grasp the spatio-temporal changes.
- Site specificity: At the relatively low-altitude Canfinal site, available permafrost distribution maps (Map of potential permafrost distribution (Federal Office for the Environment FOEN) and the SLF ‘Permafrost and ground ice map’; https://www.slf.ch/en/services-and-products/permafrost-and-ground-ice-map/) concordantly hint at patchy and likely shallow permafrost in the headwall that is not necessarily connected to the permafrost bound to the rock glacier below.
I suggest that the authors reshape the manuscript and resubmit it. I emphasize that the seasonal hydro-chemical characterization based on your large data set of sampled springs and the connection to the piezometer borehole is convincing. The frequency-domain analysis has its merits as a tool, there is no need to overstretch to explanations that are insufficiently backed up by local measurements. I am looking forward to receiving an updated version of the work!
SPECIFIC COMMENTS
L70ff (study site). Is the catchment currently glacierized or not? What is the mean annual air temperature and annual precipitation?
L101ff (methodology). Snow cover duration: The determination of the snow cover duration, given its important role in the analysis, merits a few sentences in the methods section (currently only mentioned on L165). How reliable is the ERA5 snow cover product for complex terrain? To what extent might long-lasting snow among the coarse blocks on the rock glacier surface contribute to melt (Bearzot et al., 2023)?
L118. When/at which intervals were the five sampling campaigns carried out?
L134. The daily EC amplitudes the frequency analysis is based on are small within 0.5–2 µS/cm (Fig. 9A). No rock glacier study (to my knowledge) has harnessed EC data down to such fine resolution. How does this compare to the precision & resolution of the EC probes? The EC is weakly sensitive to water temperature. How was the measured EC corrected to 25°C? Were the EC loggers fully submerged also at low flow (or shaded/enclosed in a stilling well?) and water temperature reliably measured? Since I am not familiar with this analysis, this is intended as a request for clarification and not a critique.
L136 and Fig. 9. Especially towards the end of the thaw season, the EC signal is quite irregular and far from sinusoidal. How well does the 1-cpd component describe such a signal with a broader frequency band in terms of amplitude and phase? Showing the 1-cpd component would be helpful to grasp the method.
L200. The PCA analysis is very intriguing! Fig. S2 is based on the Oct 2022 samples. How persistent is the found clustering over the season? This is shown in Fig. 6B but could be stated more clearly.
L235, L239, L242. Measured facts (diurnal EC variations) are alongside interpretations (dilution behavior, melt driver). Please move the latter to the discussion Sect. 5.3 to avoid repetition (i.e., “…seasonal trends which, for the snow-free period, can be interpreted as measures of the intensity of dilution from RG melt”, “The ratio of the EC 1 cpd amplitudes to those of Tair normalizes the EC amplitudes by the main driver of daily melt rate variations”, “…indicating a potentially significant contribution from RG meltwater”).
L301–304: “An isolated contribution from the Canfinal RG cannot be detected…” This important finding is furthermore corroborated by Fig. S2 (PCA, spatial coherence): The distinct geochemical signal is lost a few hundred meters downstream of the rock glacier front. Please add Bearzot et al. (2023) at L304 as they also provided an estimate.
L290–318: Very interesting!
L312. “Some suggest that bacterial activity…” Who?
L332. Please write “the active layer thickens” instead of “the ice thickness in the active layer decreases”.
Fig. 5. A neat figure!
Fig. 6. The single most important figure, panel B could be enlarged. Same color coding of the months in panels A and B eases comparison, nice! The ellipses in B) are unnecessary, the different coloring distracts. What do the different circle sizes mean? Could a few key springs (among S1) be marked so that we can follow how their chemistry evolves in the PC plot?
Fig. 7, caption. Should read “July 2022”, not “July 2002”. Just a thought: Flipping the map or the order of the EC panels would place the data next to location in the map, the more so, as the labels of the EC data set are “hidden” in the subscript of the y-axis label (optional).
Fig. 9A. What exactly means ‘filtered’ here (cleaned?) and why is the S1 EC here in the range 100–130 µS/cm whereas it is 50–75 µS/cm in Fig. 7? Am I missing something?
REFERENCES
Bearzot, F., Colombo, N., Cremonese, E., di Cella, U. M., Drigo, E., Caschetto, M., ... & Rossini, M. (2023). Hydrological, thermal and chemical influence of an intact rock glacier discharge on mountain stream water. Science of The Total Environment, 876, 162777.
Brighenti, S., Engel, M., Tolotti, M., Bruno, M. C., Wharton, G., Comiti, F., ... & Bertoldi, W. (2021). Contrasting physical and chemical conditions of two rock glacier springs. Hydrological Processes, 35(4), e14159.
Colombo, N., Gruber, S., Martin, M., Malandrino, M., Magnani, A., Godone, D., ... & Salerno, F. (2018). Rainfall as primary driver of discharge and solute export from rock glaciers: The Col d'Olen Rock Glacier in the NW Italian Alps. Science of the Total Environment, 639, 316-330.
Del Siro, C., Scapozza, C., Perga, M. E., & Lambiel, C. (2023). Investigating the origin of solutes in rock glacier springs in the Swiss Alps: A conceptual model. Frontiers in Earth Science, 11, 1056305.
Haeberli, W., ed.: Pilot analysis of permafrost cores from the active rock glacier Murtèl I, Piz Corvatsch, Eastern Swiss Alps. A workshop report., no. 9 in Arbeitsheft, VAW/ETH Zürich, 1990.
Mateo, E. I., & Daniels, J. M. (2019). Surface hydrological processes of rock glaciated basins in the San Juan Mountains, Colorado. Physical Geography, 40(3), 275-293.
Nickus, U., Thies, H., Krainer, K., Lang, K., Mair, V., & Tonidandel, D. (2023). A multi-millennial record of rock glacier ice chemistry (Lazaun, Italy). Frontiers in Earth Science, 11, 1141379.
Citation: https://doi.org/10.5194/egusphere-2024-927-RC2 - AC3: 'Reply on RC2', Landon Halloran, 27 Aug 2024
Viewed
HTML | XML | Total | Supplement | BibTeX | EndNote | |
---|---|---|---|---|---|---|
380 | 163 | 36 | 579 | 41 | 30 | 27 |
- HTML: 380
- PDF: 163
- XML: 36
- Total: 579
- Supplement: 41
- BibTeX: 30
- EndNote: 27
Viewed (geographical distribution)
Country | # | Views | % |
---|
Total: | 0 |
HTML: | 0 |
PDF: | 0 |
XML: | 0 |
- 1