the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 02 Jan 2024
Channel evolution processes in a diamictic glacier foreland. Implications on downstream sediment supply: case study Pasterze/Austria
Abstract. Global warming and glacier retreat are affecting the morphodynamics of proglacial rivers. In response to changing hydrology, the altered hydraulics will significantly impact future glacifluvial erosion and proglacial channel development. This study analyses a proglacial channel evolution process at the foreland of Austria’s biggest glacier Pasterze, by predicted runoff until 2050 based on a glacio-hydrological model. A high-resolution digital elevation model was created by an unmanned aerial vehicle, sediment was sampled, a one-dimensional hydrodynamic-numerical model was generated, and bedload transport formulas were used to calculate the predicted transport capacity of the proglacial river. Due to the fine sediment composition near the glacier terminus (d50< 79 mm), the calculation results underline the process of headward erosion in the still unaffected, recently deglaciated river section. In contrast, an armor layer is already partly established by the coarse grain size distribution in the already incised river section (d50> 179 mm). Furthermore, already reoccurring exposed non-fluvial grain sizes combined with decreasing flow competence in the long term indicate erosion-resistant pavement layer formation disconnecting the subsurface sediments for glacifluvial reworking (vertical landform decoupling). The presented study shows that subsystems exhibiting pavement layer formation by grains exceeding the predicted transport capacity supported by non-fluvial sediments are found at the investigated glacier foreland. Thus, an extension accompanied by a refinement of the fluvial system in the sediment cascade approach was developed as a central result.
- Preprint
(1949 KB) - Metadata XML
- BibTeX
- EndNote
Status: final response (author comments only)
-
RC1: 'Comment on hess-2023-267', Anonymous Referee #1, 17 Jan 2024
Overall, the manuscript: “Channel evolution processes in a diamictic glacier foreland. Implications on downstream sediment supply: case study Pasterze / Austria” is well prepared and addresses relevant scientific questions on the future channel evolution processes in such systems. The submission is well structured and the language is fluent and precise. However, when reviewing the manuscript some questions arose (general comments), which need to be addressed/considered during the review of the manuscript. My main concern is the assumption of a static system to predict future morphological processes in a fully dynamic system. I agree with the authors that different processes can be seen and explained, but interpretations and conclusions are in my opinion associated with uncertainties.
General comments:
- The authors obtained a DEM in 2018 and used predicted runoffs for steady-state 1D simulations (2018, 2035, and 2050), which means the system (topography, roughness) was static over time. However, from the figures, it becomes evident, that the system is fully dynamic and the canyon part changed drastically within the last three years. Hence, I am wondering how reliable these predictions are, as there are no morphological changes considered.
- What was the reason that the authors have chosen runoff in 2035, whereas the maximum mean monthly meltwater runoff is predicted for 2034?
- The authors write about the torrential flow characteristics, which may lead to morphological changes, but use mean monthly values for their steady-state simulations. Hence, I am wondering, if such mean values can replicate the morphological system, or if important runoff peaks, leading to a morphological development of the system, are missing. The authors write: “Landform and subsystem connectivity is highly dynamic (Lane et al., 2017), and changes are often triggered by high-magnitude/low-frequency events.” Especially short-time events may lead to a break-up of the bed armoring layer.
- The authors write about bed armoring and that within the canyon an armor layer has already developed. Have the authors considered the status of the armor layer and if it was fully developed. A fully developed bed armor layer can be calculated depending on the grain size distribution, e.g. given by Günter (1971).
- Even though the authors show many references, I am wondering how transferrable the results are to other glacier regions in Europe or even worldwide. The authors write that there are many boundaries involved, such as slope, bed material compositions, but a statement on that should be given.
- It was not clear to me if newly generated sediments, as a result of glacier melt, are considered in future predictions. These fine sediments may alter the morphology. Here a statement given by the authors: “Combined with the high sediment supply by glacifluvial erosion of glacial diamictic till,..” Here another question arose, how is the suspended sediment transport and the interaction with the bed considered in the study?
- The chapter on the hydrodynamic model needs more details, such as information on the chosen roughness or if a calibration was performed.
- I recommend having a separate chapter on the Glacier Evolution Runoff Model (GERM), and more details on the calibration of the model (bias-corrected data, regional climate model) and the output. Here my question is why daily values (“One of the model output data is runoff in daily resolution”) are used and not e.g. values with an hourly resolution, or maybe three hours maximum.
- Page 10, line 214: “The sediment analysis shows a downstream coarsening”. I think this is misleading, as in the delta the same grain sizes are visible as in the headwater. In general, I think we see here typical morphological patterns, where the different grain sizes depend on the boundaries, such as slope. As the canyon has the highest slope it is evident, that the coarsest sediments can be found there. I think here it is necessary to dig into the data of the HEC-RAS model to get more insight into the hydraulics of the system.
- Page 10, line 242: “the maximum mean monthly runoff.” As mentioned before, a mean value may not be representative for predicting morphology and channel evolution, as the “Glacifluvial sediment reworking is strongly coupled to runoff characteristics.”
- Page 10, lines 233-235: The paragraph needs some modification and may be re-written. The authors introduce three discharges: Qm.melt.max.2018= 4.9 m³s-1; Qm.melt.max.2034= 9.6 m³s-1 >> Qm.melt.max.2050= 3.5 m³s-1) and use Q2034 for comparisons. I think here only comparisons between the sampled sediments and the Q2018 can be made, as it is unsure what the channel and the grain size distribution will look like in 2034 or 2050. Hence, I think the statement “In the transition section (CS 622 m – CS 552 m) with a slightly increased channel gradient (Sm= 2.4 %), a much bigger characteristic grain size was calculated then measured (d50.c:LS.3= 170 mm > d50.m:LS.3= 79 mm; Fig. 5a)” is not valid, as no measurements are available yet. Why is 2050 not included in the figure? Can the authors maybe draw lines of a fully developed bed armor layer? Here I think it needs to be statistically proven that the modified approach by Chiara and Rickenmann led to better results.
- Page 12, line 267: “The analysis of the sediment composition and the hydrodynamic-numerical model results tend to the potential for riverbed incision in the headwater and for pavement layer formation (channel bed stabilization) in the canyon.” This estimate is in my opinion only valid for a static system.
- Page 14, line 317: “However, rivers in proglacial areas with an established pavement layer still enable lateral sediment supply, often triggered by high-magnitude/low-frequency events.” This is evidence that the system will change over time.
- Page 14, line 326- : The authors write about shortcomings and uncertainties, e.g., in geometry and calibration data acquisition as well as in sediment sampling, but do not quantify it.
- Page 16, line 369: “This paper predicts the future flow competence (the largest particle a flow can move) of the proglacial part of the river Möll according to the glacio-hydrological model GERM (glacier runoff evolution model) of the Pasterze Glacier by 2050.” I think this sentence is misleading, as there were many uncertainties not considered. Hence, I would rather say an estimate.
- Page 15, line 340: “The importance of considering energy losses by macro-roughness elements to achieve more plausible results was pointed out in various studies (see literature in Chiari and Rickenmann, 2011) for bedload transport calculations, especially in steep mountain streams (S> 4-6 %; Badoux and Rickenmann, 2008).” But when looking at Figure 5, I have the impression that the approach given by Rickenmann (1990) fits better than the approach with the reduced energy gradient (Fig. 5b) in the canyon. See my previous comment.
Specific comments:
Page 1, line 17: I would replace “sediment was sampled by sediment properties were obtained”
Page 1, line 18: I would not talk here about bedload transport formulas, as it was the one from Rickenmann used and an approach considering the energy line
Page 1, line 18: “Due to the fine sediment composition near the glacier terminus (d50< 79 mm).” I would not call these sediments fine sediment composition.
Page 4, line 116: “and strong seasonal and diurnal fluctuations.” Can the authors give an approximate?
Page 9, line 190: The authors only used the A1B scenario according to IPCC in their study. This needs to be justified.
Page 9, line 191: (v) the glacier edge of 2003 and 2012. What was the glacier edge in 2018 when the field survey was conducted?
Citation: https://doi.org/10.5194/hess-2023-267-RC1 -
RC2: 'Comment on hess-2023-267', Anonymous Referee #2, 21 Apr 2024
I am afraid that this paper is either fundamentally flawed, or there are critical steps in the explanation of what is done that are not explained. This needs to be addressed before a full review is possible.
The work in the paper is based upon the Rickenmann (1990) proposal of a formula for estimating the critical (specific) discharge required for sediment transport (qc). This was modified in Rickenmann (2006) and vulgarized more widely in Chiari and Rickenmann (2007, Table 1, Equation 7, modified as the authors note) to treat the effects of a reduced energy slope given macroroughness elements in streams. The critical (specific) discharge is commonly combined with an actual discharge (q) in a threshold-based sediment transport equation (Chiari and Rickenmann (2007), Table 1, Equation 6) to estimate time-varying sediment transport capacity (i.e. the transport rate is a f (q-qc)).
The authors have taken some future predictions of discharge (q) and want to see how the sediment transport competence changes as the glacier shrinks and goes through “peak water”. This is all a very legitimate thing to do, well justified by the literature which the authors know and use well. The idea that a river erodes and sorts its bed as a glacier retreats and so progressively stabilizes is a good working hypothesis for known river response following glacier retreat; although the authors don't quite pick up on wider knowledge regarding the temporality of this process as argued in the work of Marren and others (i.e. a river erodes close to the glacier but then deposits eroded sediment further downstream causing an erosion-aggradation response).
But, this is where I get lost and this is not helped by Section 3.5 which is poorly explained.
First, when they present their results, they compare the “characteristic grain sizes” with those measured (e.g. Figure 5). Of course, if you know qc and its changes through time, you can invert the Rickenmann type equations to get a characteristic grain size. But you have to know qc and you don't know it or how it will evolve. My only explanation of what they have done here is that they have completely misunderstood the associated equations. In the paper, at L199, the authors define qc as the specific discharge whereas qc is actually the **critical** specific discharge, that which the q must exceed for transport to occur. I fear that they have taken their future discharge scenarios (some measures of q) as qc and then estimated the associated grain-size. That is, in their Eq. (1) they have used q and not qc. This is completely flawed. Now I may have mistaken something here, but to make it clear, if q is being used instead of qc, this is a very basic mistake that makes the analysis meaningless. If I am mistaken it is because how they go from q to qc to a characteristic grain-size is not explained in the paper sufficiently.
Second, Eq. (1) shows that the reduced energy slope (Ir )and the grain-size (D50) drive qc. Erosion/deposition will lead to the sediment sorting that drives changes in Ir and D50 and hence qc. This is not addressed in the modelling as far as I can see. To address this you would need a time-dependent sediment sorting treatment that also took into account changes in subglacial sediment export. Sediment sorting is at its most intense when capacity is greater than supply (subglacial sediment export, sediment supply from banks) and so how the critical discharge and the competence will evolve has to take into account supply as well. The authors recognize this in the introduction and the discussion but none of their analysis actually simulates this process, as far as I can see.
Third, I should add that there is likely one other fundamental flaw. The analysis is done for the maximum mean monthly runoff and the hydrological model is daily; but the actual maximum daily discharge will be substantially greater than this due to diurnal discharge variation. Any analysis of this kind would need to work with downscaled daily discharge data as and used the mean maximum monthly runoff calculated from an hourly timescale. This diurnal discharge variation is also likely to change significantly as the glacier declines in size.
The discussion is well-situated in the literature but deviates substantially from the results that are provided. Indeed, very few results are provided in the paper.
The authors should attend to the following more minor issues if they can resolve the above flaws, redo the analysis and then resubmit the paper.
L29 “by” should be “with”
L35 “exceeds the geological norm”; not clear
L47 “last” in what sense?
L50 what is “triggered”?
L56 “to” should be “from”
L58 “parallel to …” should be “as glaciers retreat”
L64 “blankets” is a poor term
L73 “by …” poorly phrased
L74 but it depends on the glacier
L84-5 but have you measured bedload sediment?
L107-9 does not make sense as written
L131 “kettle-holes”
L138 “on” should be “for”
L172-3 be clearer here that you measured the grain-size off digital images where the canyon was inaccessible. You also need to explain how you guaranteed the equivalence of grain-sizes from the line-sampling I the field which measured b axes and the line-sampling of the imagery which measures surficial exposure of grain-sizes. See also L208-9 – the b axis measured on an image is not the same as the true b axis – there is a bias – and one that increases as a function of the level of sediment reworking
L182 and onwards – the topography you use here will not be the river bed – how did you deal with this in your cross-sections? The same issue also applies to the digital grain-size survey; how do you get the grain-sizes for underwater zones?
Citation: https://doi.org/10.5194/hess-2023-267-RC2
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 189 | 34 | 18 | 241 | 14 | 12 |
- HTML: 189
- PDF: 34
- XML: 18
- Total: 241
- BibTeX: 14
- EndNote: 12
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1