the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 17 Mar 2022
Quantification of post-glacier erosion in the European Alps using 10Be and OSL exposure dating
Joanne Elkadi
Benjamin Lehmann
Georgina King
Olivia Steinemann
Susan Ivy-Ochs
Marcus Christl
Frederic Herman
Abstract. The retreat of glaciers since the Last Glacial Maximum (LGM) in the European Alps has left an imprint on topography through glacial and non-glacial erosional processes. However, few methods are currently capable of resolving these mechanisms on Lateglacial to Holocene timescales. Quantifying the relative contributions of mountain erosion, during these different climate cycles, is useful for understanding long-term landscape evolution and the links between global climate and erosion. Here, we combine three Optically Stimulated Luminescence (OSL) exposure dating signals with 10Be surface exposure dating to constrain the post-glacier erosion rates of bedrock samples down a vertical transect adjacent to the Gorner glacier in Zermatt, Switzerland. The results reveal erosion rates on the order of 10-2 to 10-1 mm a-1, in general agreement with other studies in the region, as well as a strong negative correlation between erosion rates and elevation. Finally, at present glacial erosion is assumed to have a greater influence on landscapes, yet a global compilation of both glacial and non-glacial erosion rates in deglaciated environments shows that erosion rates during interglacial times could be equally important.
-
Notice on discussion status
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
-
Preprint
(4011 KB)
-
Supplement
(276 KB)
-
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(4011 KB) - Metadata XML
-
Supplement
(276 KB) - BibTeX
- EndNote
- Final revised paper
Journal article(s) based on this preprint
Joanne Elkadi et al.
Interactive discussion
Status: closed
-
CC1: 'Comment on egusphere-2022-43', Nicolas Young, 28 Mar 2022
I enjoyed reading this paper and I found it really helpful to have such a nice summary of glacial vs. non-glaical erosion rates that are found in the literature. In particular, I like your Figure 6 that nicely summarizes the range of erosion rate estimates, and including the method used to derive these rates is a very nice touch. I just wanted to mention that the Young et al datasets shown in Figure 6, were recently expanded upon by Balter-Kennedy et al., 2021. Balter-Kennedy 2021 adds several new data points and also introduces a new method for constraining long-term erosion rates with TCN. Authors could take a look a decide if its appropriate for this figure or discussion.
Anyway, I really enjoyed reading this manuscript.
Balter-Kennedy et al., 2021. Centennial- and orbital-scale erosion beneath the Greenland Ice Sheet near Jakobshavn Isbræ. Journal of Geophysical Research: Earth Surface, 126, e2021JF006429
--
Nicolás E. Young
Lamont-Doherty Earth Observatory
Columbia University
nicolasy@ldeo.columbia.edu
Citation: https://doi.org/10.5194/egusphere-2022-43-CC1 -
AC4: 'Reply on CC1', Joanne Elkadi, 12 Jul 2022
Thank you for your positive feedback regarding our paper and for bringing this new study to our attention. Indeed, it is a useful contribution to the compilation and we have now added it.
Best regards,
Joanne Elkadi, on behalf of all the co-authors.
Citation: https://doi.org/10.5194/egusphere-2022-43-AC4
-
AC4: 'Reply on CC1', Joanne Elkadi, 12 Jul 2022
-
RC1: 'Comment on egusphere-2022-43', Derek Fabel, 16 Apr 2022
General comments
It is good to see development of this novel technique is continuing with the addition of the IRSL50 and post-IR IRSL225 signals from feldspar to the OSL125 signal from quartz. It is encouraging that the three signals combined show similar trends.
The use of local calibration sites to address known variations in the optical electron detrapping parameters due to geographical and mineralogical variations is a good solution to a challenging problem.
Testing the effect of different sample aspect is a very valuable contribution. As is the comparison of the results to the world-wide compilation of glacial and nonglacial erosion rates which supports the idea that nonglacial erosion rates are not significantly different to glacial erosion rates.
Clarification is needed on some aspects of the manuscript.
Specific comments
The anti-correlation between erosion rate and elevation is intriguing and much less strong than in the Mont Blanc study. Please provide a reasoned explanation for the difference.
Although the 10Be data for the lower three samples are compromised by inheritance, could the authors deliberate on the inverted erosion rates for GG02 and GG03. The 10Be derived steady state erosion rates for these two samples are about 4.8E-2 mm a-1 and 5.8E-2 mm a-1. These are roughly half of the rates predicted by the inversion method. The 10Be derived steady-state erosion rates are directly related to the measured 10Be concentration in the samples and represent maximum steady-state erosion rates.
The higher erosion rates calculated using the inversion method used in this study are not compatible with the measured 10Be concentrations It is not possible to get the measured 10Be concentrations with the calculated erosion rates. It is important that the authors state very clearly if the 10Be data was in fact used in the inversion, or did they derive the erosion rates simply from Eq. 1, which does not incorporate the 10Be data.
If the 10Be data was used, please explain how the erosion rates from the inversion method are reconciled with the measured 10Be concentrations. What was the exposure/erosion history of GG02 and GG03, especially given Figure 3b suggests that the inversion modelled erosion rate is invariant for erosion onset times ts >102 a. Is it the case that the OSL signal only records the last few hundred years at the inversion method erosion rate, and prior to that time the samples were eroding at half the rate to accumulate the measured 10Be concentrations? If that is the explanation, what caused the acceleration in the erosion rate?
If the 10Be data was not used, explain why, and revise the title of the paper to reflect that 10Be data was not used to quantify the post-glacier erosion rates discussed in the manuscript.
Specific comments by line number:
104 “…since TCN are formed ~50-60 cm (Lal, 1991) below the rock surface…” is incorrect. TCN are formed at the surface and down to several metres. The ~50-60 cm is the e-folding depth for common rock densities.
117 “…due its…” should be ‘due to its’
303 Table S1 does not show summary for each sample. It shows data for Sample 5 (which I assume is GG05). Table S1 is not referred to in the main text. It is referred to in the Supplement. Table S1 in the main text should be Table S2, or change the labels in the Supp.
306 1.13 x 10-6 is 1.8E-6 in Table S2. Check the data.
307 7.34 x 10-7 is 7.3E-6 in Table S2. Check the data.
337 1.12 x 10-2 is 7.22E-2 in Table 4.
338 Add reference to Table 4 so the sentence ends… 0.16 mm a-1 (Table 4).
484 “…local differences…” This is vague. Please elaborate.
Citation: https://doi.org/10.5194/egusphere-2022-43-RC1 - AC1: 'Reply on RC1', Joanne Elkadi, 12 Jul 2022
-
RC2: 'Comment on egusphere-2022-43', Pierre Valla, 02 May 2022
Dear Authors, dear Editors,
Please find below my evaluation concerning the manuscript by Elkadi and co-authors entitled "Quantification of post-glacier erosion in the European Alps using 10Be and OSL exposure dating" (manuscript egusphere-2022-43).
This manuscript investigates bedrock surface erosion in an alpine environment, focusing on post-glacial surfaces and combining OSL and in situ 10Be surface exposure dating. The authors targeted different samples along a formerly-glaciated topographic profile, and used a multi-signal OSL investigation to constrain bedrock erosion rates and durations. Their results show variable erosion rates between signals and samples, with an elevation relationship that they relate to periglacial erosion mechanisms (e.g. frost-cracking). They compare their results to recent study in a similar environment, and finally propose a compilation of “non-glacial” vs. glacial surface erosion rates (bedrock and boulder) from the literature that they discuss in terms of rates variability and magnitude.
This is an interesting manuscript, well-written and referenced. It follows the recent developments of OSL surface exposure dating and the original approach combining OSL with 10Be data to retrieve local erosion histories. In the present study, the authors used a multi-signal approach for OSL exposure dating, which is a very good illustration of the potential of luminescence techniques for quantifying exposure/erosion histories of bedrock or boulder rock surfaces. They also investigated the usefulness of artificial calibration surfaces, with short exposure time (1 year), to efficiently constrain bleaching parameters, as well as the influence of surface orientation on these parameters. Finally, they placed their results within a large-scale compilation of surface erosion rates from the literature, discussing the relative overlap between non-glacial and glacial surface erosion rates. I thus think that the present manuscript would be a very interesting contribution for Earth Surface Dynamics, bringing new evidence and quantification of bedrock surface erosion rates in post-glacier settings, and nicely complementing recent studies in this topic while raising fruitful discussion in the geomorphology community.
I have outlined below my questions and suggestions in a set of general and specific comments below. Most of my suggestions are concerning the presentation of information related to the OSL exposure dating and calibration/multi-signal investigation. In addition, I would think that more discussion about the actual erosion/weathering processes (physical mechanisms) would help the readers to better appreciate the discussion about compiled glacial/non-glacial erosion rates.
General comments:
1 – In the present study, the authors refer to “post-glacier erosion” for the surface erosion rates they aim to quantify. I agree with the used term, although this is maybe too vague and can be specified already at the beginning of the manuscript. This should go along with a description of the sampled landscape/morphological features and a clear statement of the adopted strategy: why targeting formerly-glaciated bedrock surfaces and not random surfaces in the catchment? how comparable would be then the output surface erosion rates between GG01 and other samples, given that GG01 has never been glaciated? What are the exact erosion mechanisms investigated there? Wind erosion, surface chemical weathering, frost cracking, a mixture of all?
This is not an easy question, but I feel that the readers will better appreciate the approach and outcomes if these are better clarified in the manuscript.
2 – OSL surface exposure dating. The multi-signal approach is really interesting and promising, however the comparison between signals could be extended and complemented in my opinion. First of all, this is not entirely clear to me why bleaching parameters would be similar between different signals, as we know from literature than bleaching of IR signals are more difficult/slower than OSL signal. I would encourage the authors to provide more discussion about this interesting result. Second, the output erosion rates differ between signals, can these be indicative of the uncertainty in erosion quantification? In their output results (Table 4), the authors provide estimated erosion rates but these are not associated to any uncertainty. Would it be possible to estimate some uncertainties from the likelihood results?
More importantly, how can we explain that some bleaching profiles are in steady state for a given signal and in transient state for another signal, within the same sample/core? This is really intriguing but would need I think more discussion.
3 – Compilation of non-glacial/glacial surface erosion rates. This is a nice compilation and this questions the relative idea of efficient subglacial processes in shaping mountainous landscapes. However, I think several clarifications/information are missing to fully appreciate this compilation. First, this is unclear to me what are “non-glacial” erosion rates, since I have the impression that fluvial or landslide rates have not been included. So this is more a comparison between periglacial/hillslope erosion vs. glacial erosion, for the later the geomorphic agent being easily identified (subglacial ice or water). Second, I think that for surface erosion rates the setting/environment is also very important, i.e. one would expect different erosion rates for a bedrock/surface exposed since long time to atmospheric agents than a recently deglaciated surface, no?
Finally, I guess that the measurement time could be also important in the output erosion rate; have the authors tried to confront the compiled erosion rate to the measurement period?
Specific comments, by line number:
- Line 1. “post-glacier erosion...”. Maybe precise in title that this study investigates “bedrock surface” erosion, and is thus focusing rather on local/small-scale erosion and not large-scale landscape evolution (e.g. fluvial erosion...).
- Line 14. “glacial and non-glacial”. Please be more specific there, what is considered as “non-glacial” in the present study. Are these post-glacial evolution of glacial surfaces (by atmospheric erosion/weathering), periglacial processes or more generally fluvial/hillslope erosion? See also my general comment about this.
- Line 19. “in Zermatt, Switzerland”. Maybe precise that this is located in the (central) European Alps.
- Line 24. “...could be equally important.” I would suggest to add a sentence there for the potential implications of such result, this appears not entirely clear as presently phrased.
- Line 37. “...global feedback loop that exists...”. Some references there would be needed to introduce this feedback loop.
- Lines 42-43. “In contrast, studies exploring erosion during interglacial times have mainly investigated at catchment-wide erosion rates”. I don’t entirely agree with this statement, some studies have also investigated more local fluvial erosion (gorge incision etc., e.g. for the European Alps Korup and Schlunegger, 2007; Rolland et al., 2017; van den Berg et al., 2012; Valla et al., 2010) or the spatial distribution within a catchment (e.g. Fox et al., 2015 for the Alps). Maybe rephrase or add more information there.
- Line 42. “glacial erosion, bedrock surface erosion and rockfall”. See my general comment about this, all terms refer to “bedrock surface erosion” but physical processes and scales differ. Please check and rephrase.
- Line 49. Again there, what is “post-glacier erosion”. Hillslope, fluvial, or atmospheric weathering? This needs specification for your study.
- Line 49. “six samples”. Please precise what kind of samples (i assume glacially-polished bedrock or glacial morphologies like roches moutonnées no?). This is important to understand what processes are targeted.
- Line 75. “2 x 10-1”.
- Line 83. “post-glacier erosion rates...”. There is a good reason why targeting formerly-glaciated bedrock surfaces in the present study, but this is not really explicit in the introduction. Please consider adding one or two sentences on the adopted strategy and why targeting post-glacier surfaces rather than other bedrock surfaces randomly in the landscape.
- Line 88. “converted into an exposure age”. Add “apparent” there.
- Line 95. “surface traps”. Unclear whether these relates to traps at the rock surface or energetically for luminescence. Please rephrase. Also, maybe already precise the depth range at which the sun’s energy is sufficient to reset the OSL signal (lines 96-99).
- Line 101. Maybe add “apparent” there too for exposure age.
- Line 105. Maybe also include the recent work of Sellwood et al. 2019 and/or Sellwood and Jain 2022.
- Line 106. “influenced by exposure”. Sunlight exposure? Exposure time? Please specify.
- Line 119. “in the local area”. Not clear, please rephrase.
- Line 121. “six sampling sites down a vertical transect”. Same comment as line 49. The reader is missing a morphological/geomorphological description of the targeted bedrock surfaces (glacially-polished or not, glacial or periglacial features, etc.) and explanations for the adopted sampling strategy. This is really difficult to have a good understanding based on small insets in figure 1. Also, this is important I think to present the surface slopes for the different samples, etc.
- Line 123. “aside from the highest sample”. So this is important to explain that this sample is not reflecting “post-glacier” erosion, but periglacial erosion as this was never ice-covered 8or at least no during the LGM). Also, then what is the bedrock surface morphology for this sample (see my previous comment)?
- Line 136, Figure 1. This is a nice figure, but not totally informative for the setting area. Is it possible to add the LGM ice contours on panel b? and to replicate the ice lines on panel c for clarity (for instance I cannot really tell if the three bottom samples have been lastly exposed in 1973 or 2009 based on panel b)? Pictures as inset in panel c are really small, and scale is missing? What is the source(s) of the photos showed in panel b and c? Another question, what is above the sample GG02, it looks like a small plateau or morainic ridge (Younger Dryas?)? Maybe consider adding also a topographic profile for the transect, on which you can locate the samples and ice thickness/extent from LGM to present-day.
One suggestion would be to add another figure (supp or main text) to show the sampled morphologies and potentially the different lithologies (rock-slice pictures?).
- Line 136. “Sample preparation”. Please specify where sample preparation and chemical Be extraction have been performed.
- Line 165. There is a ) to be removed for the blank value.
- Line 189. “with a DASH head”. I would suggest to describe the different filters listed in Table 2 for non-specialists.
- Lines 192-194. Are the different criteria arbitrary or common for rock-slice luminescence? Maybe refer to technical paper to support these, e.g. Elkadi et al., 2021?
- Line 205. This is unclear and not explained in the main text how equation 1 is treated with respect to the recombination distance r’. For non-specialist readers this will appear relatively obscure, given that athermal detrapping parameters are not presented for these measurements/samples. This is also similar for the dose rate parameters (D0 and Ddot), no information about their values (and how D0 is obtained) is provided, only description in Table 3.
- Line 220. “Previous calibration sources”. Unclear, please rephrase.
- Line 224. “unknown parameter values”. Please specify this parameter for clarity (σφ?).
- Line 225. “the influence of the surface orientation”. Additionally, I think discussion about the outcome results would be interesting for readers if reported in main text, not in supp (at least briefly).
- Line 239. “... using the random parameter values and Equation 1”.
- Lines 250-251. “1.25x10^8 trials”. For each individual sample or in total?
Also, I don’t fully understand how the ranges for the inverted parameters have been defined, especially for exposure time t only between 1 and 200 years but setting information suggest much longer exposure times for high-elevation samples no?
Please clarify on which basis/information the parameter ranges have been defined.
- Lines 258-259. “...simple, step wise erosion history where, at a specific time in the past, the surface goes from experiencing no erosion to an instantaneous onset of fixed rate of erosion”. I am wondering whether this is possible to also have a simpler scenario where you estimate erosion rate since the exposure of the bedrock surface (i.e. ts = t from 10Be data). Have you tested this and if yes is there any compatible scenario(s) with OSL/10Be data?
- Line 259. For the inversion of erosion history, what are the bleaching (σφ and μ) parameters and exposure times used? Best-fitting values for bleaching parameters (Table S1)? Please clarify.
- Line 265. Several questions for Table 1:
Can you add more information for surface orientation? Two values are given, but no unit nor details.
Please also provide 10Be/9Be ratios in the table, so that 10Be concentrations can be recalculated in the future.
Are the uncertainties reported for exposure ages internal or external?
- Line 274. “3. Results and interpretation”. The presented results are already quite interpreted in this section, so I would suggest to rephrase the section label.
- Line 276. “apparent exposure ages”. I would also suggest to add a figure with 10Be apparent exposure ages and topography for illustration.
- Line 277. “The highest elevation sample (GG01) is younger than suggested from ice thickness reconstructions (Bini et al., 2009)”. If this sample has been collected above the LGM ice surface, then it reflects periglacial exposure and its apparent exposure age is not related to LGM glaciation, see for instance results in Gallach et al. 2018; 2020. Please consider rephrasing or clarifying this sentence.
- Line 279. This is a very interesting result as you can reconstruct the YD ice thickness from your 10Be apparent exposure ages, which may be linked to this small plateau/surface just above sample GG02. Please consider expanding this result, this is relatively similar outcomes compared to Lehmann et al. 2020.
- Line 288. Maybe also consider citing the work of Goehring et al. 2011 on the Rhone glacier.
- Lines 297-299. This sentence may be moved to methods.
- Line 303. “results for each sample summarised in Table S1”. I would strongly encourage the authors to present results as figures (like figure 2) for all samples, either in main text or in supplementary. This would be important for the readers to evaluate the noise in data and reproducibility between cores for each sample (old and calibration, and also for different orientations).
- Line 308. Is it possible to present there quickly the results about different orientations? I guess this would be interesting for some readers to have such information, not all in supplementary.
- Line 313. “...mineralogical variations”. Is there a link between μ values and lithology? Can the authors provide some pictures of the rock slices, especially for GG02 which seems different from others?
- Line 315, Figure 2. I would suggest to have at least one figure showing the bleaching profiles of the different signals, at present only IRSL50 signals are shown. Is it possible to provide such information?
On figure 2, inversion outcomes for t, the OSL apparent exposure time, is shown. However, this outcome is not presented in Table S2, nor discussed in the main text. I think this is important to show this, and to clearly present the differences in apparent exposure ages between OSL and 10Be data for all samples.
- Line 318. “inversion outcomes for e and ts”. Please provide the range for these parameters.
- Line 320. “exposure age information from the historical maps and photos were employed”. Where can the reader access the used exposure ages for these samples? Please specify in main text what exposure durations you used.
- Line 323. “transient state”. This is not totally clear what is transient state from looking at figure 3d, please clarify for non-specialists that there is a wide range of e/ts combinations, reflecting non-steady bleaching profile, or something similar.
- Line 324, Figure 3. Please indicate the used exposure time for model without erosion in panels a and c. Concerning panel d, since total exposure time used is historical data (so few tens of years), I don’t understand how ts range can be explored between 0.1 and 10000 years with an output likelihood. If I understand well, ts <= exposure time, so there should be a large white (non-possible) area in panel d no?
Please justify the adopted approach, this is not really clear at present.
What is the red line on panels a and c (model with erosion)? The best-fitting parameter combination (maybe indicate with a star in b and d panels) or the region of high likelihood? Please clarify (same question for figure 4).
- Line 338-339. “When looking at the signals individually, the OSL125 and post-IR IRSL225 results reveal an anti-correlation between post-glacier erosion rates and elevation, whereas no trend is observed in the IRSL50 data (Fig. 5)”. On Figure 5a one cannot differentiate the different signals (same symbols), can the authors change the symbols so that the reader can evaluate the differences?
- Line 341. “Based on this, an average of the three signals was calculated for each site to generate one post-glacier erosion rate value”.
I think this would be first interesting to discuss the different e/ts results between signals, before going to an average calculation. Is there some variability between signals in the output surface erosion rates? Why some signals appear in steady-state while other appear in transient state? I would think this is important for readers to have such information.
In addition, would it be possible to estimate some uncertainties (standard deviation? from likelihood?) and to show these on figure 5 for individual/averaged erosion rates?
- Line 346. “minimum ts”. There is no presentation of these outcomes in the section, I would suggest to provide more details about these and to confront them to total exposure time. For low-elevation samples, ts is close to exposure time, whereas it is really different (much lower) for high-elevation samples. I think this is important for the exposed results on lines 347-351, otherwise the readers could think longer exposure time = more eroded material...
- Lines 368-370. It reads a bit strange to have the presentation of the slope relationship there (discussion), and not in the previous section along with the elevation relationship. Please consider presenting these in results too.
- Line 375. “local variations influencing the dominant post-glacier erosional mechanisms”. Really vague, please specify what are those variations and mechanisms.
Alternatively, have the authors thought about potential correlation between erosion rate and exposure time? For Lehmann et al. (2020), the exposure times vary between ~20 ka and few years, while there the difference in exposure times is much lower. I agree that GG01 is not following this potential relationship, with a young exposure age, but given the different morphology/settings (cliff with periglacial erosion over 10s of ka), this may explain the low erosion rate.
- Lines 379-390. I agree that this is worth noting low bedrock surface erosion rates for such high-elevation environments, but these low erosion rates may also be the result of the sampling strategy, no? The sampling targets are specifically glacially-formed surfaces that are more or less preserved in the landscape, so they do reflect low surface erosion. I think that some further clarification could be given there.
- Line 403. “...bedrock surface erosion rates from surfaces in glaciated environments, not currently subjected to glacial erosion,...”. Reads a bit odd, please rephrase.
- Lines 411-415. Are the referenced studies targeting bedrock/boulder surfaces that have been previously glaciated or not? Maybe this is important to specify. Same question for line 421 (“ In Europe, Andrée (2022b)...”).
- Line 415. “bedrock erosion rates”. I thought Sohbati et al. (2018) only targeted boulders, please check.
- Line 427. “these orders of magnitude are comparable with estimations of sub-glacial erosion rates and a summary of glacial and non-glacial erosion rates worldwide is displayed in Fig. 6”. Have the authors tried to perform a pdf of the glacial and non-glacial erosion estimates. From visual inspection, I have the impression that glacial erosion rates, although they do overlap with non-glacial ones, are higher (and the presented scale is a log one!).
I appreciate this comparison and think that the compilation is interesting to discuss, however, I have a doubt about the actual comparison: “non-glacial rates” are apparently referring to “atmospheric” erosion/weathering and fluvial or landslide/hillslope erosion rates are not included right?
Then, what is really compared between these rates and glacial rates which do involve geomorphic agent as subglacial water/ice? I think this is important to clarify this point and justify why fluvial or landslide erosion rates (which are non-glacial agents) are not considered.
- Line 444. “The large range is due to differences in sample locations...”. How about differences in lithology (e.g. carbonate vs. crystalline bedrocks)?
- Line 462. “A full compilation of glacier erosion rates, calculations and methods can be found in Herman et al. (2021)”. Maybe the authors can provide there the range in compiled glacial erosion rates?
- Line 484. “the dominant post-glacier erosion mechanisms”. Please specify.
I hope these comments and suggestions may be useful for revising the manuscript, and I look forward to seeing it published.
Sincerely,
Pierre Valla
Grenoble, 2 May 2022
Citation: https://doi.org/10.5194/egusphere-2022-43-RC2 - AC2: 'Reply on RC2', Joanne Elkadi, 12 Jul 2022
-
RC3: 'Comment on egusphere-2022-43', Rachel Smedley, 03 May 2022
Quantification of post-glacier erosion in the European Alps using 10Be and OSL exposure dating
Elkadi et al.
Manuscript ID: https://doi.org/10.5194/egusphere-2022-43
Summary
This paper uses innovative 10Be-OSL measurements to derive erosion rates in the Swiss Alps during the Late Glacial and Holocene. It targets a vertical transect to assess the influence of elevation on erosion rates in this setting and shows that a negative correlation exists i.e. as the elevation decreases, the erosion rate increases. This is new and useful information as little is known about the factors that control erosion rates, especially in interglacial times. The erosion rates derived are similar in magnitude to existing studies, which gives confidence in the robustness of this new technique. Finally, the authors apply their data to address the long-standing uncertainties in our understanding of glacial vs non-glacial/interglacial erosion rates. Interestingly, their data suggests that interglacial erosion rates can be equally as important as glacial erosion rates in deglaciated environments, which is a key finding because this has important implications for understanding the drivers of rock erosion rates (e.g. climate) and thus, future rock erosion with anthropogenic climate change.
Overall, this is an excellent study, applying new techniques to a long-standing, challenging research question. The methods applied are robust, well justified and well performed. The study is generally well contextualised within the literature, but the understanding of the factors driving erosion rates could be better explained in the text in places (see specific comments) so it is easier for the reader to follow the authors interpretations. It was a very interesting read and I have some comments and questions below. It will be an excellent contribution to the literature in this area and I hope the comments are constructive.
General comments:
- From my understanding, this is the first study to determine rock surface erosion rates using this technique using both K-feldspar and quartz, which is very important and interesting. The authors may wish emphasise this more in the intro/rationale/abstract, but I leave it to their discretion.
- One of the advantages of the Lehmann et al. (2018) approach is that transient erosion rates can be derived, in addition to steady-state erosion rates. Given that this paper is focussed on interpreting the character and drivers of erosion, I would expect the authors to have more thoughts and interpretation of those samples that determine transient erosion, rather than just dismissing them as is stated in Line 324. For example, do these samples derive transient erosion rates because the technique/analysis is not reliable? Do these samples have different surficial characteristics than other samples? Is there any evidence of transient erosion for these samples (e.g. frost shattering) that is not present for the other samples? What natural processes could have caused transient erosion in this setting? What even is transient erosion? This could be its own discussion point in the discussion before the steady-state erosion is discussed (Sections 4.1, 4.2).
- One of the main findings from this study is that “at present glacial erosion is assumed to have a greater influence on landscapes, yet a global compilation of both glacial and non-glacial erosion rates in deglaciated environments shows that erosion rates during interglacial times could be equally important” (Abstract, Lines 21-24). This is very interesting and is reflected in the data presented in this study. However, the discussion lacks discussion about glacial vs non-glacial or interglacial erosion rates. It could further unpack what natural processes differ between glacial and interglacial conditions that may or may not modulate the rock surface erosion (e.g. climate). Kirkbride and Bell (2010) do this well in the discussion of their study with respect to changing temperature and precipitation in glacial vs interglacial periods. Perhaps the discussion here could provide more insight into this as it is largely unknown due to the difficulty in determining glacial and interglacial erosion rates (i.e. deriving erosion rates on such resolution). The new data presented in this study on timeframes that were previously difficult to measure erosion rates on, therefore offers a great opportunity to explore these themes.
Specific comments:
- Please could the authors explain what they are referring to when they use the term “non-glacial erosion”. Is it referring to the interglacial period (i.e. it has a time dimension) or a deglaciated setting (i.e. it has a space dimension)? It is a minor comment but it would help to clarify this in the introduction before the reader continues on through the paper, perhaps around Line 34 where it is first mentioned.
- Line 42 – here you refer to erosion studies during interglacial times and state that they are mainly limited to catchment-wide erosion rates but you could add 1-2 sentences to highlight that there are a few papers that have quantified interglacial erosion rates (e.g. Kirkbride and Bell, 2010; Sohbati et al. 2018; Lehmann et al. 2019; Smedley et al. 2021), which you will later expand upon in Section 1.1.
- Line 51-54 – it is useful to set up the aim of the study here, but I find it a little confusing that you report the main findings before presenting the data. Perhaps this is a feature of the journal and if so, that is fine as it is. If not, you might want to consider waiting to report the findings later in the paper.
- Line 81 - Smedley et al. (2021) also measured erosion rates over the last 4 ka so derived interglacial erosion rates and suggested that the transient nature of the erosion could have been caused by climate fluctuations over this time period. This is probably worth adding given the scarcity of papers that use TCN and OSL surface exposure methods to derive erosion rates.
- Line 94 – technically Jenkins et al. (2018) performed burial dating, which is quite different from the exposure dating techniques mentioned. Discussing burial dating here is not necessary, but if you wish to demonstrate that it can be used for burial dating, I would be explicit about it and also add a reference to Freiesleben et al. 2015, for example:
- “In recent years, the application of OSL to rock surface dating has proved successful in a variety of settings for exposure dating (e.g. Sohbati et al., 2015; Liu et al., 2019; Lehmann et al., 2018) and burial dating (e.g. Freiesleben et al. 2015; Jenkins et al., 2018).”
- Line 100 – calibration for what? I suggest you add “after calibration to account for the rock-specific light attenuation rates” or something similar.
- Line 105 – add reference to Smedley et al. (2021) as it is possibly the only other reference that has used multiple luminescence signals specifically for deriving rock erosion rates with 10Be and OSL measurements as you are doing in this study.
- Line 181 – please could you add a few words as to why you were sampling areas with minimal lichen cover and red, iron-oxide staining to explain to those who may wish to sample using this approach in the future. Why is it important?
- Line 188 – please could you add a line to explain why the approach of Elkadi et al. (2021) was beneficial for these measurements and so demonstrate the importance to the reader, e.g. does it dramatically improve the measurement reproducibility? Are the measurements more accurate?
- Line 188 – it would be worth stating explicitly here that you will derive three signals per sample for comparison, so OSL signal of quartz, IR50 and pIRIR225 signals of feldspar. It would also be helpful to non-experts/users to explain why analysing multiple signals is useful in this context. It is really unique and interesting so worth emphasising.
- Line 193 – subscript the n in Tn in both occurrences.
- Line 194 – please explain why the slices were excluded from further analysis? Does it mean the results would not be reliable? At present, to a non-expert the sentence makes it sound a little like they are just rejected and could be better explained (although very briefly!) why these criteria are applied.
- Line 219 – here you may wish to also consider the work recently published by Furhmann et al. (2022) on the incidence angle of light given your interest in the orientation of the sample for calibration (https://doi.org/10.1016/j.radmeas.2022.106732).
- Line 222 – here you state that you have provided sample-specific calibration parameters by returning to each site after a year. Presumably this is for all three lithologies, so holnfels, schist and gneiss, AND for all three signals, which would be worth highlighting here for clarity.
- Given the infancy of the technique, the variability in lithology and the fact that you’re using quartz and feldspar, I think this would be of great interest to the community and so would be worth including Table S1 into the main manuscript but this is the authors discretion.
- Lines 312-313 – it is unusual to include some interpretation in the results section but given that the discussion is focussed on the erosion rates rather than the specifics of the luminescence technique, it is reasonable. However, if you are going to offer some discussion of the OSL unknown parameters in Section 3.2, it would be useful to discuss how the quartz and feldspar attenuation rates compared given that no (or few) other examples exist in the literature showing such data and it would be interesting to unpack this unique data, especially relative to the variability in lithologies of the samples.
- Line 356 – “Several factors, often working in combination with each other, modulate bedrock surface erosion rates. These include temperature, elevation and surface slope”. This makes it sound like only three factors modulate erosion rates, which is not the case as explored by Portenga and Bierman (2011) amongst other studies. Presumably temperature, elevation and surface slope are factors you will focus on in this study? If so, either state all the factors that may modulate erosion and then say explicity that you’ll only consider these three, or just re-phrase to “Several factors, often working in combination with each other, modulate bedrock surface erosion rates. These include, but are not limited to, temperature, elevation and surface slope”.
- Line 364 – lithology is known to have a dominant control on rock surface erosion (e.g. Ford and Williams, 1989; Twidale, 1982; Moses et al. 2014), but this is not explicit from this section. It would be worth adding a sentence or two discussing the dominant role lithology has in modulating rock erosion rates, and then perhaps discussing whether you observe this in the erosion rates you measured for hornfels, schist and gneiss, or are they all similarly resistance to weathering and subsequent erosion? Given the metamorphic origin or these rocks, it is possible that they are more resistant than other lithologies (e.g. sandstones, limestones). Either way, it would be interesting having this discussion relative to your measured erosion rates, which are difficult to obtain.
- Lines 379-390 – you state here that the anti-correlation between erosion rate and elevation is likely reflecting the lack of frost crack weathering in this setting, which is very interesting and new information, but where do your samples that derived transient erosion rates fit into this picture? Could these samples be reflecting frost crack weathering given that presumably frost cracking processes would be more stochastic over time and so more likely to be reflected by transient erosion, rather than steady-state. It would be interesting to have a better understanding of what transient erosion rates may be recording from the natural environment in general.
- Lines 391-394 – I find this a little confusing so perhaps you could better explain it for the reader. How do the observed patterns of glacial erosion in a valley due to quarrying and/or abrasion (that occur when the ice is present) control the interglacial erosion rates (when the ice is not present)? Are you suggesting that the rock has been weakened more during the glacial and so the interglacial erosion rates are higher at lower elevations? I think it would help the reader follow your arguments and interpretations better in this section if you provided a little more explanation for this.
- Line 399 – you give an example of frost crack weathering despite stating in Line 387-388 that “frost crack weathering is perhaps not a dominant form of post-glacier erosion in these areas”, and rather “bedrock erosion is most likely occurring through continuous grain-by-grain erosion”. I feel like these two interpretations do not align. Alternatively, have you considered the role of moisture via precipitation in this setting? Do lower elevations receive more rainfall/snowfall and therefore are subject to greater chemical weathering and subsequent erosion? It has long been known that precipitation can be a driver of rock weathering and subsequent erosion (e.g. Hall et al. 2012; Merill, 1906; Moses et al. 2014; Swantesson et al. 1992). Furthermore, in the ‘global’ compilation of rock outcrop erosion rates by Portenga and Bierman (2011), multi-variate statistical analysis showed that 32% of the variation in the global population of outcrop erosion rates could be explained by the five environmental parameters considered (latitude, elevation, relief, mean annual precipitation, mean annual temperature and seismicity), with mean annual precipitation being the most important parameter, accounting for 14% of the variability in this ‘global’ dataset even across many different settings. As such, it might be worth considering precipitation in your discussion. Although palaeo-precipitation records will be almost impossible, perhaps there are at least contemporary observational data of mean annual rainfall and snowfall from an elevation range of the alps for contextualisation?
- Line 409 – Given the scarcity of studies, it is worth adding Smedley et al. 2021 as an OSL application, and then potentially expanding upon the findings of this study in Lines 411-425, given the authors determined interglacial erosion rates. Although the erosion rates derived were transient, it would be worth considering the erosion rates in the range that were lower and could be sustained for longer time intervals as these are more comparable to your steady-state erosion rates, in comparison to the higher erosion rates that can only be sustained over shorter timeframes.
Citation: https://doi.org/10.5194/egusphere-2022-43-RC3 - AC3: 'Reply on RC3', Joanne Elkadi, 12 Jul 2022
Interactive discussion
Status: closed
-
CC1: 'Comment on egusphere-2022-43', Nicolas Young, 28 Mar 2022
I enjoyed reading this paper and I found it really helpful to have such a nice summary of glacial vs. non-glaical erosion rates that are found in the literature. In particular, I like your Figure 6 that nicely summarizes the range of erosion rate estimates, and including the method used to derive these rates is a very nice touch. I just wanted to mention that the Young et al datasets shown in Figure 6, were recently expanded upon by Balter-Kennedy et al., 2021. Balter-Kennedy 2021 adds several new data points and also introduces a new method for constraining long-term erosion rates with TCN. Authors could take a look a decide if its appropriate for this figure or discussion.
Anyway, I really enjoyed reading this manuscript.
Balter-Kennedy et al., 2021. Centennial- and orbital-scale erosion beneath the Greenland Ice Sheet near Jakobshavn Isbræ. Journal of Geophysical Research: Earth Surface, 126, e2021JF006429
--
Nicolás E. Young
Lamont-Doherty Earth Observatory
Columbia University
nicolasy@ldeo.columbia.edu
Citation: https://doi.org/10.5194/egusphere-2022-43-CC1 -
AC4: 'Reply on CC1', Joanne Elkadi, 12 Jul 2022
Thank you for your positive feedback regarding our paper and for bringing this new study to our attention. Indeed, it is a useful contribution to the compilation and we have now added it.
Best regards,
Joanne Elkadi, on behalf of all the co-authors.
Citation: https://doi.org/10.5194/egusphere-2022-43-AC4
-
AC4: 'Reply on CC1', Joanne Elkadi, 12 Jul 2022
-
RC1: 'Comment on egusphere-2022-43', Derek Fabel, 16 Apr 2022
General comments
It is good to see development of this novel technique is continuing with the addition of the IRSL50 and post-IR IRSL225 signals from feldspar to the OSL125 signal from quartz. It is encouraging that the three signals combined show similar trends.
The use of local calibration sites to address known variations in the optical electron detrapping parameters due to geographical and mineralogical variations is a good solution to a challenging problem.
Testing the effect of different sample aspect is a very valuable contribution. As is the comparison of the results to the world-wide compilation of glacial and nonglacial erosion rates which supports the idea that nonglacial erosion rates are not significantly different to glacial erosion rates.
Clarification is needed on some aspects of the manuscript.
Specific comments
The anti-correlation between erosion rate and elevation is intriguing and much less strong than in the Mont Blanc study. Please provide a reasoned explanation for the difference.
Although the 10Be data for the lower three samples are compromised by inheritance, could the authors deliberate on the inverted erosion rates for GG02 and GG03. The 10Be derived steady state erosion rates for these two samples are about 4.8E-2 mm a-1 and 5.8E-2 mm a-1. These are roughly half of the rates predicted by the inversion method. The 10Be derived steady-state erosion rates are directly related to the measured 10Be concentration in the samples and represent maximum steady-state erosion rates.
The higher erosion rates calculated using the inversion method used in this study are not compatible with the measured 10Be concentrations It is not possible to get the measured 10Be concentrations with the calculated erosion rates. It is important that the authors state very clearly if the 10Be data was in fact used in the inversion, or did they derive the erosion rates simply from Eq. 1, which does not incorporate the 10Be data.
If the 10Be data was used, please explain how the erosion rates from the inversion method are reconciled with the measured 10Be concentrations. What was the exposure/erosion history of GG02 and GG03, especially given Figure 3b suggests that the inversion modelled erosion rate is invariant for erosion onset times ts >102 a. Is it the case that the OSL signal only records the last few hundred years at the inversion method erosion rate, and prior to that time the samples were eroding at half the rate to accumulate the measured 10Be concentrations? If that is the explanation, what caused the acceleration in the erosion rate?
If the 10Be data was not used, explain why, and revise the title of the paper to reflect that 10Be data was not used to quantify the post-glacier erosion rates discussed in the manuscript.
Specific comments by line number:
104 “…since TCN are formed ~50-60 cm (Lal, 1991) below the rock surface…” is incorrect. TCN are formed at the surface and down to several metres. The ~50-60 cm is the e-folding depth for common rock densities.
117 “…due its…” should be ‘due to its’
303 Table S1 does not show summary for each sample. It shows data for Sample 5 (which I assume is GG05). Table S1 is not referred to in the main text. It is referred to in the Supplement. Table S1 in the main text should be Table S2, or change the labels in the Supp.
306 1.13 x 10-6 is 1.8E-6 in Table S2. Check the data.
307 7.34 x 10-7 is 7.3E-6 in Table S2. Check the data.
337 1.12 x 10-2 is 7.22E-2 in Table 4.
338 Add reference to Table 4 so the sentence ends… 0.16 mm a-1 (Table 4).
484 “…local differences…” This is vague. Please elaborate.
Citation: https://doi.org/10.5194/egusphere-2022-43-RC1 - AC1: 'Reply on RC1', Joanne Elkadi, 12 Jul 2022
-
RC2: 'Comment on egusphere-2022-43', Pierre Valla, 02 May 2022
Dear Authors, dear Editors,
Please find below my evaluation concerning the manuscript by Elkadi and co-authors entitled "Quantification of post-glacier erosion in the European Alps using 10Be and OSL exposure dating" (manuscript egusphere-2022-43).
This manuscript investigates bedrock surface erosion in an alpine environment, focusing on post-glacial surfaces and combining OSL and in situ 10Be surface exposure dating. The authors targeted different samples along a formerly-glaciated topographic profile, and used a multi-signal OSL investigation to constrain bedrock erosion rates and durations. Their results show variable erosion rates between signals and samples, with an elevation relationship that they relate to periglacial erosion mechanisms (e.g. frost-cracking). They compare their results to recent study in a similar environment, and finally propose a compilation of “non-glacial” vs. glacial surface erosion rates (bedrock and boulder) from the literature that they discuss in terms of rates variability and magnitude.
This is an interesting manuscript, well-written and referenced. It follows the recent developments of OSL surface exposure dating and the original approach combining OSL with 10Be data to retrieve local erosion histories. In the present study, the authors used a multi-signal approach for OSL exposure dating, which is a very good illustration of the potential of luminescence techniques for quantifying exposure/erosion histories of bedrock or boulder rock surfaces. They also investigated the usefulness of artificial calibration surfaces, with short exposure time (1 year), to efficiently constrain bleaching parameters, as well as the influence of surface orientation on these parameters. Finally, they placed their results within a large-scale compilation of surface erosion rates from the literature, discussing the relative overlap between non-glacial and glacial surface erosion rates. I thus think that the present manuscript would be a very interesting contribution for Earth Surface Dynamics, bringing new evidence and quantification of bedrock surface erosion rates in post-glacier settings, and nicely complementing recent studies in this topic while raising fruitful discussion in the geomorphology community.
I have outlined below my questions and suggestions in a set of general and specific comments below. Most of my suggestions are concerning the presentation of information related to the OSL exposure dating and calibration/multi-signal investigation. In addition, I would think that more discussion about the actual erosion/weathering processes (physical mechanisms) would help the readers to better appreciate the discussion about compiled glacial/non-glacial erosion rates.
General comments:
1 – In the present study, the authors refer to “post-glacier erosion” for the surface erosion rates they aim to quantify. I agree with the used term, although this is maybe too vague and can be specified already at the beginning of the manuscript. This should go along with a description of the sampled landscape/morphological features and a clear statement of the adopted strategy: why targeting formerly-glaciated bedrock surfaces and not random surfaces in the catchment? how comparable would be then the output surface erosion rates between GG01 and other samples, given that GG01 has never been glaciated? What are the exact erosion mechanisms investigated there? Wind erosion, surface chemical weathering, frost cracking, a mixture of all?
This is not an easy question, but I feel that the readers will better appreciate the approach and outcomes if these are better clarified in the manuscript.
2 – OSL surface exposure dating. The multi-signal approach is really interesting and promising, however the comparison between signals could be extended and complemented in my opinion. First of all, this is not entirely clear to me why bleaching parameters would be similar between different signals, as we know from literature than bleaching of IR signals are more difficult/slower than OSL signal. I would encourage the authors to provide more discussion about this interesting result. Second, the output erosion rates differ between signals, can these be indicative of the uncertainty in erosion quantification? In their output results (Table 4), the authors provide estimated erosion rates but these are not associated to any uncertainty. Would it be possible to estimate some uncertainties from the likelihood results?
More importantly, how can we explain that some bleaching profiles are in steady state for a given signal and in transient state for another signal, within the same sample/core? This is really intriguing but would need I think more discussion.
3 – Compilation of non-glacial/glacial surface erosion rates. This is a nice compilation and this questions the relative idea of efficient subglacial processes in shaping mountainous landscapes. However, I think several clarifications/information are missing to fully appreciate this compilation. First, this is unclear to me what are “non-glacial” erosion rates, since I have the impression that fluvial or landslide rates have not been included. So this is more a comparison between periglacial/hillslope erosion vs. glacial erosion, for the later the geomorphic agent being easily identified (subglacial ice or water). Second, I think that for surface erosion rates the setting/environment is also very important, i.e. one would expect different erosion rates for a bedrock/surface exposed since long time to atmospheric agents than a recently deglaciated surface, no?
Finally, I guess that the measurement time could be also important in the output erosion rate; have the authors tried to confront the compiled erosion rate to the measurement period?
Specific comments, by line number:
- Line 1. “post-glacier erosion...”. Maybe precise in title that this study investigates “bedrock surface” erosion, and is thus focusing rather on local/small-scale erosion and not large-scale landscape evolution (e.g. fluvial erosion...).
- Line 14. “glacial and non-glacial”. Please be more specific there, what is considered as “non-glacial” in the present study. Are these post-glacial evolution of glacial surfaces (by atmospheric erosion/weathering), periglacial processes or more generally fluvial/hillslope erosion? See also my general comment about this.
- Line 19. “in Zermatt, Switzerland”. Maybe precise that this is located in the (central) European Alps.
- Line 24. “...could be equally important.” I would suggest to add a sentence there for the potential implications of such result, this appears not entirely clear as presently phrased.
- Line 37. “...global feedback loop that exists...”. Some references there would be needed to introduce this feedback loop.
- Lines 42-43. “In contrast, studies exploring erosion during interglacial times have mainly investigated at catchment-wide erosion rates”. I don’t entirely agree with this statement, some studies have also investigated more local fluvial erosion (gorge incision etc., e.g. for the European Alps Korup and Schlunegger, 2007; Rolland et al., 2017; van den Berg et al., 2012; Valla et al., 2010) or the spatial distribution within a catchment (e.g. Fox et al., 2015 for the Alps). Maybe rephrase or add more information there.
- Line 42. “glacial erosion, bedrock surface erosion and rockfall”. See my general comment about this, all terms refer to “bedrock surface erosion” but physical processes and scales differ. Please check and rephrase.
- Line 49. Again there, what is “post-glacier erosion”. Hillslope, fluvial, or atmospheric weathering? This needs specification for your study.
- Line 49. “six samples”. Please precise what kind of samples (i assume glacially-polished bedrock or glacial morphologies like roches moutonnées no?). This is important to understand what processes are targeted.
- Line 75. “2 x 10-1”.
- Line 83. “post-glacier erosion rates...”. There is a good reason why targeting formerly-glaciated bedrock surfaces in the present study, but this is not really explicit in the introduction. Please consider adding one or two sentences on the adopted strategy and why targeting post-glacier surfaces rather than other bedrock surfaces randomly in the landscape.
- Line 88. “converted into an exposure age”. Add “apparent” there.
- Line 95. “surface traps”. Unclear whether these relates to traps at the rock surface or energetically for luminescence. Please rephrase. Also, maybe already precise the depth range at which the sun’s energy is sufficient to reset the OSL signal (lines 96-99).
- Line 101. Maybe add “apparent” there too for exposure age.
- Line 105. Maybe also include the recent work of Sellwood et al. 2019 and/or Sellwood and Jain 2022.
- Line 106. “influenced by exposure”. Sunlight exposure? Exposure time? Please specify.
- Line 119. “in the local area”. Not clear, please rephrase.
- Line 121. “six sampling sites down a vertical transect”. Same comment as line 49. The reader is missing a morphological/geomorphological description of the targeted bedrock surfaces (glacially-polished or not, glacial or periglacial features, etc.) and explanations for the adopted sampling strategy. This is really difficult to have a good understanding based on small insets in figure 1. Also, this is important I think to present the surface slopes for the different samples, etc.
- Line 123. “aside from the highest sample”. So this is important to explain that this sample is not reflecting “post-glacier” erosion, but periglacial erosion as this was never ice-covered 8or at least no during the LGM). Also, then what is the bedrock surface morphology for this sample (see my previous comment)?
- Line 136, Figure 1. This is a nice figure, but not totally informative for the setting area. Is it possible to add the LGM ice contours on panel b? and to replicate the ice lines on panel c for clarity (for instance I cannot really tell if the three bottom samples have been lastly exposed in 1973 or 2009 based on panel b)? Pictures as inset in panel c are really small, and scale is missing? What is the source(s) of the photos showed in panel b and c? Another question, what is above the sample GG02, it looks like a small plateau or morainic ridge (Younger Dryas?)? Maybe consider adding also a topographic profile for the transect, on which you can locate the samples and ice thickness/extent from LGM to present-day.
One suggestion would be to add another figure (supp or main text) to show the sampled morphologies and potentially the different lithologies (rock-slice pictures?).
- Line 136. “Sample preparation”. Please specify where sample preparation and chemical Be extraction have been performed.
- Line 165. There is a ) to be removed for the blank value.
- Line 189. “with a DASH head”. I would suggest to describe the different filters listed in Table 2 for non-specialists.
- Lines 192-194. Are the different criteria arbitrary or common for rock-slice luminescence? Maybe refer to technical paper to support these, e.g. Elkadi et al., 2021?
- Line 205. This is unclear and not explained in the main text how equation 1 is treated with respect to the recombination distance r’. For non-specialist readers this will appear relatively obscure, given that athermal detrapping parameters are not presented for these measurements/samples. This is also similar for the dose rate parameters (D0 and Ddot), no information about their values (and how D0 is obtained) is provided, only description in Table 3.
- Line 220. “Previous calibration sources”. Unclear, please rephrase.
- Line 224. “unknown parameter values”. Please specify this parameter for clarity (σφ?).
- Line 225. “the influence of the surface orientation”. Additionally, I think discussion about the outcome results would be interesting for readers if reported in main text, not in supp (at least briefly).
- Line 239. “... using the random parameter values and Equation 1”.
- Lines 250-251. “1.25x10^8 trials”. For each individual sample or in total?
Also, I don’t fully understand how the ranges for the inverted parameters have been defined, especially for exposure time t only between 1 and 200 years but setting information suggest much longer exposure times for high-elevation samples no?
Please clarify on which basis/information the parameter ranges have been defined.
- Lines 258-259. “...simple, step wise erosion history where, at a specific time in the past, the surface goes from experiencing no erosion to an instantaneous onset of fixed rate of erosion”. I am wondering whether this is possible to also have a simpler scenario where you estimate erosion rate since the exposure of the bedrock surface (i.e. ts = t from 10Be data). Have you tested this and if yes is there any compatible scenario(s) with OSL/10Be data?
- Line 259. For the inversion of erosion history, what are the bleaching (σφ and μ) parameters and exposure times used? Best-fitting values for bleaching parameters (Table S1)? Please clarify.
- Line 265. Several questions for Table 1:
Can you add more information for surface orientation? Two values are given, but no unit nor details.
Please also provide 10Be/9Be ratios in the table, so that 10Be concentrations can be recalculated in the future.
Are the uncertainties reported for exposure ages internal or external?
- Line 274. “3. Results and interpretation”. The presented results are already quite interpreted in this section, so I would suggest to rephrase the section label.
- Line 276. “apparent exposure ages”. I would also suggest to add a figure with 10Be apparent exposure ages and topography for illustration.
- Line 277. “The highest elevation sample (GG01) is younger than suggested from ice thickness reconstructions (Bini et al., 2009)”. If this sample has been collected above the LGM ice surface, then it reflects periglacial exposure and its apparent exposure age is not related to LGM glaciation, see for instance results in Gallach et al. 2018; 2020. Please consider rephrasing or clarifying this sentence.
- Line 279. This is a very interesting result as you can reconstruct the YD ice thickness from your 10Be apparent exposure ages, which may be linked to this small plateau/surface just above sample GG02. Please consider expanding this result, this is relatively similar outcomes compared to Lehmann et al. 2020.
- Line 288. Maybe also consider citing the work of Goehring et al. 2011 on the Rhone glacier.
- Lines 297-299. This sentence may be moved to methods.
- Line 303. “results for each sample summarised in Table S1”. I would strongly encourage the authors to present results as figures (like figure 2) for all samples, either in main text or in supplementary. This would be important for the readers to evaluate the noise in data and reproducibility between cores for each sample (old and calibration, and also for different orientations).
- Line 308. Is it possible to present there quickly the results about different orientations? I guess this would be interesting for some readers to have such information, not all in supplementary.
- Line 313. “...mineralogical variations”. Is there a link between μ values and lithology? Can the authors provide some pictures of the rock slices, especially for GG02 which seems different from others?
- Line 315, Figure 2. I would suggest to have at least one figure showing the bleaching profiles of the different signals, at present only IRSL50 signals are shown. Is it possible to provide such information?
On figure 2, inversion outcomes for t, the OSL apparent exposure time, is shown. However, this outcome is not presented in Table S2, nor discussed in the main text. I think this is important to show this, and to clearly present the differences in apparent exposure ages between OSL and 10Be data for all samples.
- Line 318. “inversion outcomes for e and ts”. Please provide the range for these parameters.
- Line 320. “exposure age information from the historical maps and photos were employed”. Where can the reader access the used exposure ages for these samples? Please specify in main text what exposure durations you used.
- Line 323. “transient state”. This is not totally clear what is transient state from looking at figure 3d, please clarify for non-specialists that there is a wide range of e/ts combinations, reflecting non-steady bleaching profile, or something similar.
- Line 324, Figure 3. Please indicate the used exposure time for model without erosion in panels a and c. Concerning panel d, since total exposure time used is historical data (so few tens of years), I don’t understand how ts range can be explored between 0.1 and 10000 years with an output likelihood. If I understand well, ts <= exposure time, so there should be a large white (non-possible) area in panel d no?
Please justify the adopted approach, this is not really clear at present.
What is the red line on panels a and c (model with erosion)? The best-fitting parameter combination (maybe indicate with a star in b and d panels) or the region of high likelihood? Please clarify (same question for figure 4).
- Line 338-339. “When looking at the signals individually, the OSL125 and post-IR IRSL225 results reveal an anti-correlation between post-glacier erosion rates and elevation, whereas no trend is observed in the IRSL50 data (Fig. 5)”. On Figure 5a one cannot differentiate the different signals (same symbols), can the authors change the symbols so that the reader can evaluate the differences?
- Line 341. “Based on this, an average of the three signals was calculated for each site to generate one post-glacier erosion rate value”.
I think this would be first interesting to discuss the different e/ts results between signals, before going to an average calculation. Is there some variability between signals in the output surface erosion rates? Why some signals appear in steady-state while other appear in transient state? I would think this is important for readers to have such information.
In addition, would it be possible to estimate some uncertainties (standard deviation? from likelihood?) and to show these on figure 5 for individual/averaged erosion rates?
- Line 346. “minimum ts”. There is no presentation of these outcomes in the section, I would suggest to provide more details about these and to confront them to total exposure time. For low-elevation samples, ts is close to exposure time, whereas it is really different (much lower) for high-elevation samples. I think this is important for the exposed results on lines 347-351, otherwise the readers could think longer exposure time = more eroded material...
- Lines 368-370. It reads a bit strange to have the presentation of the slope relationship there (discussion), and not in the previous section along with the elevation relationship. Please consider presenting these in results too.
- Line 375. “local variations influencing the dominant post-glacier erosional mechanisms”. Really vague, please specify what are those variations and mechanisms.
Alternatively, have the authors thought about potential correlation between erosion rate and exposure time? For Lehmann et al. (2020), the exposure times vary between ~20 ka and few years, while there the difference in exposure times is much lower. I agree that GG01 is not following this potential relationship, with a young exposure age, but given the different morphology/settings (cliff with periglacial erosion over 10s of ka), this may explain the low erosion rate.
- Lines 379-390. I agree that this is worth noting low bedrock surface erosion rates for such high-elevation environments, but these low erosion rates may also be the result of the sampling strategy, no? The sampling targets are specifically glacially-formed surfaces that are more or less preserved in the landscape, so they do reflect low surface erosion. I think that some further clarification could be given there.
- Line 403. “...bedrock surface erosion rates from surfaces in glaciated environments, not currently subjected to glacial erosion,...”. Reads a bit odd, please rephrase.
- Lines 411-415. Are the referenced studies targeting bedrock/boulder surfaces that have been previously glaciated or not? Maybe this is important to specify. Same question for line 421 (“ In Europe, Andrée (2022b)...”).
- Line 415. “bedrock erosion rates”. I thought Sohbati et al. (2018) only targeted boulders, please check.
- Line 427. “these orders of magnitude are comparable with estimations of sub-glacial erosion rates and a summary of glacial and non-glacial erosion rates worldwide is displayed in Fig. 6”. Have the authors tried to perform a pdf of the glacial and non-glacial erosion estimates. From visual inspection, I have the impression that glacial erosion rates, although they do overlap with non-glacial ones, are higher (and the presented scale is a log one!).
I appreciate this comparison and think that the compilation is interesting to discuss, however, I have a doubt about the actual comparison: “non-glacial rates” are apparently referring to “atmospheric” erosion/weathering and fluvial or landslide/hillslope erosion rates are not included right?
Then, what is really compared between these rates and glacial rates which do involve geomorphic agent as subglacial water/ice? I think this is important to clarify this point and justify why fluvial or landslide erosion rates (which are non-glacial agents) are not considered.
- Line 444. “The large range is due to differences in sample locations...”. How about differences in lithology (e.g. carbonate vs. crystalline bedrocks)?
- Line 462. “A full compilation of glacier erosion rates, calculations and methods can be found in Herman et al. (2021)”. Maybe the authors can provide there the range in compiled glacial erosion rates?
- Line 484. “the dominant post-glacier erosion mechanisms”. Please specify.
I hope these comments and suggestions may be useful for revising the manuscript, and I look forward to seeing it published.
Sincerely,
Pierre Valla
Grenoble, 2 May 2022
Citation: https://doi.org/10.5194/egusphere-2022-43-RC2 - AC2: 'Reply on RC2', Joanne Elkadi, 12 Jul 2022
-
RC3: 'Comment on egusphere-2022-43', Rachel Smedley, 03 May 2022
Quantification of post-glacier erosion in the European Alps using 10Be and OSL exposure dating
Elkadi et al.
Manuscript ID: https://doi.org/10.5194/egusphere-2022-43
Summary
This paper uses innovative 10Be-OSL measurements to derive erosion rates in the Swiss Alps during the Late Glacial and Holocene. It targets a vertical transect to assess the influence of elevation on erosion rates in this setting and shows that a negative correlation exists i.e. as the elevation decreases, the erosion rate increases. This is new and useful information as little is known about the factors that control erosion rates, especially in interglacial times. The erosion rates derived are similar in magnitude to existing studies, which gives confidence in the robustness of this new technique. Finally, the authors apply their data to address the long-standing uncertainties in our understanding of glacial vs non-glacial/interglacial erosion rates. Interestingly, their data suggests that interglacial erosion rates can be equally as important as glacial erosion rates in deglaciated environments, which is a key finding because this has important implications for understanding the drivers of rock erosion rates (e.g. climate) and thus, future rock erosion with anthropogenic climate change.
Overall, this is an excellent study, applying new techniques to a long-standing, challenging research question. The methods applied are robust, well justified and well performed. The study is generally well contextualised within the literature, but the understanding of the factors driving erosion rates could be better explained in the text in places (see specific comments) so it is easier for the reader to follow the authors interpretations. It was a very interesting read and I have some comments and questions below. It will be an excellent contribution to the literature in this area and I hope the comments are constructive.
General comments:
- From my understanding, this is the first study to determine rock surface erosion rates using this technique using both K-feldspar and quartz, which is very important and interesting. The authors may wish emphasise this more in the intro/rationale/abstract, but I leave it to their discretion.
- One of the advantages of the Lehmann et al. (2018) approach is that transient erosion rates can be derived, in addition to steady-state erosion rates. Given that this paper is focussed on interpreting the character and drivers of erosion, I would expect the authors to have more thoughts and interpretation of those samples that determine transient erosion, rather than just dismissing them as is stated in Line 324. For example, do these samples derive transient erosion rates because the technique/analysis is not reliable? Do these samples have different surficial characteristics than other samples? Is there any evidence of transient erosion for these samples (e.g. frost shattering) that is not present for the other samples? What natural processes could have caused transient erosion in this setting? What even is transient erosion? This could be its own discussion point in the discussion before the steady-state erosion is discussed (Sections 4.1, 4.2).
- One of the main findings from this study is that “at present glacial erosion is assumed to have a greater influence on landscapes, yet a global compilation of both glacial and non-glacial erosion rates in deglaciated environments shows that erosion rates during interglacial times could be equally important” (Abstract, Lines 21-24). This is very interesting and is reflected in the data presented in this study. However, the discussion lacks discussion about glacial vs non-glacial or interglacial erosion rates. It could further unpack what natural processes differ between glacial and interglacial conditions that may or may not modulate the rock surface erosion (e.g. climate). Kirkbride and Bell (2010) do this well in the discussion of their study with respect to changing temperature and precipitation in glacial vs interglacial periods. Perhaps the discussion here could provide more insight into this as it is largely unknown due to the difficulty in determining glacial and interglacial erosion rates (i.e. deriving erosion rates on such resolution). The new data presented in this study on timeframes that were previously difficult to measure erosion rates on, therefore offers a great opportunity to explore these themes.
Specific comments:
- Please could the authors explain what they are referring to when they use the term “non-glacial erosion”. Is it referring to the interglacial period (i.e. it has a time dimension) or a deglaciated setting (i.e. it has a space dimension)? It is a minor comment but it would help to clarify this in the introduction before the reader continues on through the paper, perhaps around Line 34 where it is first mentioned.
- Line 42 – here you refer to erosion studies during interglacial times and state that they are mainly limited to catchment-wide erosion rates but you could add 1-2 sentences to highlight that there are a few papers that have quantified interglacial erosion rates (e.g. Kirkbride and Bell, 2010; Sohbati et al. 2018; Lehmann et al. 2019; Smedley et al. 2021), which you will later expand upon in Section 1.1.
- Line 51-54 – it is useful to set up the aim of the study here, but I find it a little confusing that you report the main findings before presenting the data. Perhaps this is a feature of the journal and if so, that is fine as it is. If not, you might want to consider waiting to report the findings later in the paper.
- Line 81 - Smedley et al. (2021) also measured erosion rates over the last 4 ka so derived interglacial erosion rates and suggested that the transient nature of the erosion could have been caused by climate fluctuations over this time period. This is probably worth adding given the scarcity of papers that use TCN and OSL surface exposure methods to derive erosion rates.
- Line 94 – technically Jenkins et al. (2018) performed burial dating, which is quite different from the exposure dating techniques mentioned. Discussing burial dating here is not necessary, but if you wish to demonstrate that it can be used for burial dating, I would be explicit about it and also add a reference to Freiesleben et al. 2015, for example:
- “In recent years, the application of OSL to rock surface dating has proved successful in a variety of settings for exposure dating (e.g. Sohbati et al., 2015; Liu et al., 2019; Lehmann et al., 2018) and burial dating (e.g. Freiesleben et al. 2015; Jenkins et al., 2018).”
- Line 100 – calibration for what? I suggest you add “after calibration to account for the rock-specific light attenuation rates” or something similar.
- Line 105 – add reference to Smedley et al. (2021) as it is possibly the only other reference that has used multiple luminescence signals specifically for deriving rock erosion rates with 10Be and OSL measurements as you are doing in this study.
- Line 181 – please could you add a few words as to why you were sampling areas with minimal lichen cover and red, iron-oxide staining to explain to those who may wish to sample using this approach in the future. Why is it important?
- Line 188 – please could you add a line to explain why the approach of Elkadi et al. (2021) was beneficial for these measurements and so demonstrate the importance to the reader, e.g. does it dramatically improve the measurement reproducibility? Are the measurements more accurate?
- Line 188 – it would be worth stating explicitly here that you will derive three signals per sample for comparison, so OSL signal of quartz, IR50 and pIRIR225 signals of feldspar. It would also be helpful to non-experts/users to explain why analysing multiple signals is useful in this context. It is really unique and interesting so worth emphasising.
- Line 193 – subscript the n in Tn in both occurrences.
- Line 194 – please explain why the slices were excluded from further analysis? Does it mean the results would not be reliable? At present, to a non-expert the sentence makes it sound a little like they are just rejected and could be better explained (although very briefly!) why these criteria are applied.
- Line 219 – here you may wish to also consider the work recently published by Furhmann et al. (2022) on the incidence angle of light given your interest in the orientation of the sample for calibration (https://doi.org/10.1016/j.radmeas.2022.106732).
- Line 222 – here you state that you have provided sample-specific calibration parameters by returning to each site after a year. Presumably this is for all three lithologies, so holnfels, schist and gneiss, AND for all three signals, which would be worth highlighting here for clarity.
- Given the infancy of the technique, the variability in lithology and the fact that you’re using quartz and feldspar, I think this would be of great interest to the community and so would be worth including Table S1 into the main manuscript but this is the authors discretion.
- Lines 312-313 – it is unusual to include some interpretation in the results section but given that the discussion is focussed on the erosion rates rather than the specifics of the luminescence technique, it is reasonable. However, if you are going to offer some discussion of the OSL unknown parameters in Section 3.2, it would be useful to discuss how the quartz and feldspar attenuation rates compared given that no (or few) other examples exist in the literature showing such data and it would be interesting to unpack this unique data, especially relative to the variability in lithologies of the samples.
- Line 356 – “Several factors, often working in combination with each other, modulate bedrock surface erosion rates. These include temperature, elevation and surface slope”. This makes it sound like only three factors modulate erosion rates, which is not the case as explored by Portenga and Bierman (2011) amongst other studies. Presumably temperature, elevation and surface slope are factors you will focus on in this study? If so, either state all the factors that may modulate erosion and then say explicity that you’ll only consider these three, or just re-phrase to “Several factors, often working in combination with each other, modulate bedrock surface erosion rates. These include, but are not limited to, temperature, elevation and surface slope”.
- Line 364 – lithology is known to have a dominant control on rock surface erosion (e.g. Ford and Williams, 1989; Twidale, 1982; Moses et al. 2014), but this is not explicit from this section. It would be worth adding a sentence or two discussing the dominant role lithology has in modulating rock erosion rates, and then perhaps discussing whether you observe this in the erosion rates you measured for hornfels, schist and gneiss, or are they all similarly resistance to weathering and subsequent erosion? Given the metamorphic origin or these rocks, it is possible that they are more resistant than other lithologies (e.g. sandstones, limestones). Either way, it would be interesting having this discussion relative to your measured erosion rates, which are difficult to obtain.
- Lines 379-390 – you state here that the anti-correlation between erosion rate and elevation is likely reflecting the lack of frost crack weathering in this setting, which is very interesting and new information, but where do your samples that derived transient erosion rates fit into this picture? Could these samples be reflecting frost crack weathering given that presumably frost cracking processes would be more stochastic over time and so more likely to be reflected by transient erosion, rather than steady-state. It would be interesting to have a better understanding of what transient erosion rates may be recording from the natural environment in general.
- Lines 391-394 – I find this a little confusing so perhaps you could better explain it for the reader. How do the observed patterns of glacial erosion in a valley due to quarrying and/or abrasion (that occur when the ice is present) control the interglacial erosion rates (when the ice is not present)? Are you suggesting that the rock has been weakened more during the glacial and so the interglacial erosion rates are higher at lower elevations? I think it would help the reader follow your arguments and interpretations better in this section if you provided a little more explanation for this.
- Line 399 – you give an example of frost crack weathering despite stating in Line 387-388 that “frost crack weathering is perhaps not a dominant form of post-glacier erosion in these areas”, and rather “bedrock erosion is most likely occurring through continuous grain-by-grain erosion”. I feel like these two interpretations do not align. Alternatively, have you considered the role of moisture via precipitation in this setting? Do lower elevations receive more rainfall/snowfall and therefore are subject to greater chemical weathering and subsequent erosion? It has long been known that precipitation can be a driver of rock weathering and subsequent erosion (e.g. Hall et al. 2012; Merill, 1906; Moses et al. 2014; Swantesson et al. 1992). Furthermore, in the ‘global’ compilation of rock outcrop erosion rates by Portenga and Bierman (2011), multi-variate statistical analysis showed that 32% of the variation in the global population of outcrop erosion rates could be explained by the five environmental parameters considered (latitude, elevation, relief, mean annual precipitation, mean annual temperature and seismicity), with mean annual precipitation being the most important parameter, accounting for 14% of the variability in this ‘global’ dataset even across many different settings. As such, it might be worth considering precipitation in your discussion. Although palaeo-precipitation records will be almost impossible, perhaps there are at least contemporary observational data of mean annual rainfall and snowfall from an elevation range of the alps for contextualisation?
- Line 409 – Given the scarcity of studies, it is worth adding Smedley et al. 2021 as an OSL application, and then potentially expanding upon the findings of this study in Lines 411-425, given the authors determined interglacial erosion rates. Although the erosion rates derived were transient, it would be worth considering the erosion rates in the range that were lower and could be sustained for longer time intervals as these are more comparable to your steady-state erosion rates, in comparison to the higher erosion rates that can only be sustained over shorter timeframes.
Citation: https://doi.org/10.5194/egusphere-2022-43-RC3 - AC3: 'Reply on RC3', Joanne Elkadi, 12 Jul 2022
Peer review completion
Journal article(s) based on this preprint
Joanne Elkadi et al.
Joanne Elkadi et al.
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 499 | 190 | 29 | 718 | 36 | 8 | 8 |
- HTML: 499
- PDF: 190
- XML: 29
- Total: 718
- Supplement: 36
- BibTeX: 8
- EndNote: 8
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(4011 KB) - Metadata XML
-
Supplement
(276 KB) - BibTeX
- EndNote
- Final revised paper