the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 04 Sep 2024
Warm proglacial lake temperatures and thermal undercutting drives rapid retreat of an Arctic glacier
Abstract. Determining the characteristics of Arctic proglacial lakes is essential for understanding their current and future influence on glacier mass loss, capacity as a carbon sink and the associated impacts for downstream hydrology and ecology. Here we combine satellite and field observations of Kaskasapakte Glacier (KG) (a lake-terminating glacier in Arctic Sweden) to reveal the interplay between lake parameters and glacier mass loss from 2008–2019. We present the first field evidence of warmer than expected water temperatures (>4 °C at the ice front) at a Scandinavian proglacial lake and illustrate how these drove rapid thermo-erosional undercutting and calving at the terminus, with width averaged retreat rates of up to 23 m per melt year and frontal ablation accounting for 30 % of glacier volume loss between 2015 and 2019. Field observations of how proglacial lake properties influence rates of glacier mass loss remain sparse, yet are increasingly critical for the accurate projection of lake-terminating glacier responses to warming air and lake temperatures, particularly in high-latitude Scandinavia under the influence of Arctic amplification.
- Preprint
(1783 KB) - Metadata XML
-
Supplement
(495 KB) - BibTeX
- EndNote
Status: open (until 16 Nov 2024)
-
RC1: 'Comment on egusphere-2024-2510', Anonymous Referee #1, 27 Sep 2024
reply
Review of Dye et al., Warm proglacial lake temperatures and thermal undercutting drives rapid retreat of an Arctic glacier.
This submission presents a collection of field- and satellite-derived datasets to characterise terminus fluctuations, surface elevation changes, and calving patterns/mechanisms at a proglacial Arctic lake. The primary conclusion is that frontal ablation is a key component of glacier mass loss, and that undercutting from warm lake temperatures is a key driver of rapid terminus retreat. These are logical interpretations to make, and are largely uncontroversial, but they are only weakly substantiated by the evidence that is presented, leaving the reader to take a large amount on trust. This is probably the most significant of a number of major issues with the manuscript in its current form. These are noted below, and given I believe there will have to be quite a bit of re-writing before this work could be published, I have stopped short of highlighting typographic and editorial errors at this stage.
- There is, at present, no clear overriding research question that this study seems to focus on addressing. A well-designed study identifies a gap in knowledge, formulates a data collection strategy to shed light on that gap, collects those data, and then analyses the results to provide an advance in knowledge. The data presented here seek to address four key objectives that are loosely connected but that each demand a major effort by themselves to reach some substantial and novel conclusions (i.e. they could and maybe should each be a study in their own right). As a result, the effort is spread too thinly, and each of the datasets are deficient in some way, undermining the interpretation – for example, if lake thermal characteristics are to be revealed then it really needs more than a single sensor to say something robust, and similarly if calving rates are to be defined then 2D camera imagery acquired over six weeks of the melt season is not really sufficient. This leaves the authors with a rather speculative discussion about what may or may not be driving mass loss at this site, and making some quite significant leaps between observations and interpretations, which leaves the reader unconvinced that they are at all robust.
- Perhaps as a consequence of the above, there is no common thread that can be navigated through the different sections. Different datasets are introduced at different times, describing different aspects of analysis over different timescales. It is very difficult to keep track of what is going on at each step, and it requires a lot of checking backwards and forwards to remind oneself what has already been introduced and what its purpose was. In a similar way the discussion jumps from one aspect to another, with many elements benefitting from only a single paragraph comprising several sentences on the findings of other studies rather than making deep and insightful interpretations of the results that have been presented here. The manuscript would benefit from a re-design to tell a coherent story about the specific problem or question to be solved; at present it unfortunately fails to do this in any meaningful way.
- The presentation of the methods is currently quite difficult to navigate. I have read this section several times and still don’t follow certain elements (in particular what metres of recession per melt year is – lines 87-92 are really confusing – and how subaqueous mass loss was calculated without a corresponding 2015 bathymetric dataset?). I also find several aspects to be missing – for example different lake temperature measurements at different depths pop up in the manuscript discussion and in the SI that are not mentioned at all within the methods section.
- A major deficiency is the almost complete absence of any uncertainty estimates – uncertainty on the DEM calculations (associated with any offsets or biases between the two datasets, which come from very different sources/methods – were these co-registered in any way? And what sort of values were acquired in stable off-ice areas), uncertainty on the sonar depths, uncertainty on the measurement of the ice-cliff positions and the height of the ice face (that is then used in calculations). These need calculating, and adding, to every presented quantitative value.
- The calculation of mass loss is fundamentally flawed by the absence of any ice dynamics information. As the authors note themselves, terminus positions are a composite effect of the forward motion of the glacier and the removal of mass by melt and calving processes. And surface elevation changes are a composite effect of any vertical component of the ice velocity (emergence in this case), dynamic effects, and surface melt/sublimation. To dismiss ice velocity as being negligible because other glaciers in the area are slow flowing is not acceptable. Velocity data are now widely available, for example the velocities from Kaskasapakte Glacier are readily available from Millan et al., 2022, and a quick look suggests these are not negligible, as stated in the manuscript (Ref: https://doi.org/10.1038/s41561-021-00885-z). It would not be a big step to incorporate these measurements into the calculations and make them a lot more robust. An additional observation here though is that the % contributions of each form of mass loss are highly dependent on the areal extent of the surface elevation change analysis – this seems to be an arbitrary distance from the terminus at present, whereas to be able to talk about mass loss from the system this needs to be integrated across the entire glacier. Otherwise, the information collected is simply surface elevation change, not mass loss.
- A final more general point is that to be able to convince the reader that one thing is driving another, it is necessary to show that the effect you have observed has happened because of some behavioural aspect of the control. Here, the lake has been in existence for multiple decades, maybe even a century, and will have warmed up every summer and cooled down every winter, albeit with some warming over the long-term. So what is it that has changed recently to cause the rapid terminus recession? There is a hint in the discussion that calving may have increased, but without evidence. There is also some suggestion that recent heatwaves may have contributed, but again there is no long-term weather station data presented to show this. Is it not much more likely that the glacier is responding to a negative climate forcing, and the ice flux has reduced, and that has caused the rapid recession? Unless some change in the forcing can be shown, and/or all other possibilities can be discounted, the conclusion of a single or key driver being responsible for the changes is highly suspect, especially when this is the same driver that has been around for many years or decades beforehand.
I do think there are some valuable observations within the data that are presented here, and with some careful thinking about (and reformulation of) the manuscript structure they will be worthy of publication. Unfortunately I do not support the publication of this submission in its current form.
Citation: https://doi.org/10.5194/egusphere-2024-2510-RC1
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 233 | 44 | 71 | 348 | 13 | 2 | 2 |
- HTML: 233
- PDF: 44
- XML: 71
- Total: 348
- Supplement: 13
- BibTeX: 2
- EndNote: 2
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1