SUMMARY

Recommendation - accept with minor changes, subject to attention to the figures

(with scope for more substantial revision/improvement at editorial/authorial discretion)

This paper addresses an apparently trivial landform feature, a very small and unimpressive RSF, which (most unusually, in the British mountains and generally) happens to be discernible from a major road (M6).  It analyses the slope stresses acting on such a slipped mass by an engineering geology technique Swedge, with and without support from glacier ice.  It dates the outer face and the scar above, and from their significantly older and younger ages, it states that this corroborates a contention that the slipped mass could only have been emplaced thus with the support of a waning valley glacier.

The paper is in fact of great interest, and a novel contribution to geomorphology, for four main reasons:

1. case studies of individual RSFs in the British mountains are uncommon, and those analysing their engineering geology are vanishingly rare (and essentially confined to a couple of PhD theses on Scottish Highlands sites, now decades old).
2. few RSFs in the Lake District (two) and Highlands (a score or so) have been dated - all of them fully disintegrated rock avalanches, whereas this is an arrested translational slide. Even globally, quasi-intact rockslides have seldom been dated due to obvious sampling difficulties, with the emphasis on ‘antiscarp’ trench faces rather than outward-facing slopes as here.
3. glacier support for RSF emplacement is not an entirely original notion, but to the writer’s incomplete knowledge this may be a pioneering study of a possible demonstration example, in Britain at least.
4. morphologically, RSFs in cirques form a small minority, with some of those on outer flanks rather than headwalls; and the distance travelled by the putative slipmass while remaining quasi-intact - 110 m - is remarkable, especially for such a small feature.

(the authors are perhaps too modest here, and could increase the impact of the paper by placing its originality more firmly within the British and global montane RSF literature).

Although the site is trivial in size - at 0.03 km² in extent by a standard method it is only just above the threshold of 0.01 km² for designation as RSF (see ‘References’) - this proves not to be a demerit, for it is in effect a scale model or field-laboratory experiment.  Although this site is unusual in some ways, it is not sui generis, and the findings are relevant to the wider study of RSF behaviour.  Again, these virtues could be made clearer.

While the paper is eminently publishable essentially as it stands, it relies on a number of assumptions, stated and unstated, which bear closer scrutiny.  The paper could be greatly improved by adumbrating these assumptions;  alternatively, the editors and authors may be well content with ‘putting it out there’ as an aunt sally to provoke and stimulate further debate and work.  If the latter course is followed, it is recommended that at least the ‘corroborated hypothesis’ be expressed more circumspectly[1], with due acknowledgement that it does rest on a set of convenient assumptions, as befits a controlled experiment.

These assumptions are explored in the COMMENTARY attached

Review of Ms “Ice buttressing-controlled rock slope failure on a cirque headwall, English Lake District” by Carling et al.

Ms # esurf-2023-14

Introduction This Ms proposes that a displaced but largely intact rock mass in Great Coum, Cumbria reached its present position by long-term, slow translation following failure, its motion being controlled by gradual debuttressing during post-LGM cirque glacier retreat. Two surface exposure dates, one (ca 18 ka) from the surface of the displaced mass and one (ca 12 ka) from the failure surface exposed by its motion, are claimed to constrain the timing of this event to ca 18 ka, which would be in line with established local deglaciation chronologies. The implication is that many of the undated intact RSF deposits in Britain are likely to have been retarded in their motion by gradual debuttressing by retreating ice.

This is a careful and detailed work that, irrespective of whether it is correct, would greatly benefit the community if published. The following comments suggest some further considerations that, in the opinion of the reviewer, would enhance the credibility of the work if addressed.

Review The critical evidence that supports the assumed gradual displacement of the rock mass is the fact that is substantially intact following a vertical displacement of ca 110 m on a slope of slightly greater than 30° (Fig. 4A, lines 644-5); however, there is stated to be an area of disintegrated rubble below the intact mass (lines 400-1) making up about 3% of the mass, so some degree of disintegration may well have occurred during emplacement.

If this slow displacement is accepted, then a cause can be provided in the form of downwasting of a cirque glacier assumed to have been present in Great Coum during the Last Glacial Maximum, so that the lateral support provided by the ice to the outer surface of the detached rock mass prevented it from displacing rapidly. This assumption is confirmed by detailed stability calculations for the in-situ rock mass, assumed to be detached from the parent mass by a planar failure surface in the form of a mapped fault; these show that the factor of safety F for the detached rock mass is in the region of 0.74-0.94 without ice support, whereas with ice support it is greater than 1.0.

The cosmogenic date of 18 ka for the outer surface of the displaced rock mass is asserted (lines 693-4) to be “compatible with the RSF movement during the final deglaciation around 19.2 to 16.6 ka”. This is undoubtedly true; however, approximately the same date would have been found if the RSF had NOT moved at that time. There is no doubt that the rock mass moved – but there is no direct evidence that it moved at ca 18 ka. However, if ice buttressing is proven to have been necessary to explain proven slow RSF motion, then uncovering of the RSF surface would have occurred during deglaciation of the cirque as hypothesised.

The hypothesis of the Ms therefore rests on (i) the assumption of slow displacement of the RSF, (ii) the interpretation of cosmogenic dating, and (iii) the stability analysis. These are now critically examined.

Slow displacement of the RSF In lines 404-406 Carling et al. indicate (correctly) that the H/L value for the event of 0.6 indicates no excessive runout, presumably to infer that this was a slow event; the largely intact nature of the RSF deposit leads to the same conclusion. But it is well-known that rapid blockslides can occur and remain intact at much lower values of H/L (e.g. Davies et al., 2006), meaning that the fact that the deposit is intact is not a reliable indication that the movement was slow. It is therefore quite possible that the RSF was emplaced rapidly (meaning at sliding speed commensurate with the marginal slope gradient – perhaps on the order of cm/sec to m/sec) and therefore it could have occurred after retreat of ice from its outer face – perhaps many thousands of years after.

Cosmogenic dating As outlined in lines 614-6, the exposure date from the exposed headscarp (12 ka ± 0.8 ka) significantly postdates that from the outer face of the deposit (18.0 ± 1.2 ka). This requires that the upper headscarp became exposed about 6000 y after the disappearance of ice from the upper, outer face of the RSF. Carling et al. state (lines 614-620) that this is expected “… due the basal failure plane being progressively exposed after the upper portion of the RSF (where the sample OSF occurs) was clear of ice cover and the RSF began to move downslope” (note, however, that they later state (lines 695-7) “We interpret the much younger exposure age on the fault plane as the result of postglacial weathering and erosion.”, suggesting some lack of confidence in this issue. The Supplementary Material indeed states that these factors have been considered in arriving at the quoted age). This means that the RSF had only displaced by about 20 – 30 m in the 6000 years following lowering of the cirque ice surface to expose the RSF outer surface. This requires extraordinarily slow downwasting of the cirque ice – is it compatible with what is known about post-LGM and pre-YD deglaciation rates? Line 732 states that “…the Lake District was essentially ice-free by 14.7 ka…”, and Great Coum is very low at 262-468 masl, so expecting its cirque glacier to have continued to support the RSF until later than 12 ka seems to be stretching the point too far, especially given that the Younger Dryas stadial onset was 12.9 ka (line 62). From this perspective the younger cosmogenic age is difficult to reconcile with the Ms hypothesis.

Slope stability analysis This appears to have been done with admirable attention to detail, and Carling et al. correctly emphasise the uncertainties involved (line 508). Nevertheless, in such work, particularly in the context of natural slopes where the subsurface composition is unknown and has to be inferred from surface exposures, it is necessary to include an error analysis to demonstrate the possible imprecision of the resulting factors of safety. For example, in section 5.3, F values are given to two decimal places with no indication of possible error. Further, in lines 518-9 reference is made to “marginal” F values of 1.07 to 1.22, suggesting that F = 1.22 is so close to 1.0 that it should be considered marginal. If this is the case (and it seems reasonable to me that there might be ~20% error in F values), then F = 0.78 should also be treated as marginal. In this perspective the factor of safety analysis becomes much less convincing in indicating that the RSF would have been unconditionally unstable in the absence of ice buttressing. Other inherent uncertainties in the stability analysis include

• The assumption that the rear boundary of the RSF is a (wavy) planar fault with specified cohesion and friction coefficient, with no asperities or rock bridges. Particularly at first failure it is quite likely that this was not the case, meaning that the estimated F values for the non-buttressed case could be too low.
• Pore water pressure is treated rather offhandedly in the analysis, presumably because of lack of information. Assuming zero pwp would be the conservative assumption, given the stated fractured nature of the rock and the proximity of the failure surface to the exposed RSF face.
• It has been assumed that the ice surface of the cirque glacier is solid and horizontal throughout. This seems unlikely – my recollection of cirque ice is that it tends to slope very steeply up to a deep bergschrund where it contacts the rock. The presence of other crevasses is also likely to complicate the transmission of ice pressure to the RSF outer face.

Thus while the factors of safety appear to conclusively support the Ms hypothesis at first sight, the lack of information on likely imprecisions means that the outcome is unconvincing.

Summary The above considerations cast some doubt on the certainty of the Ms hypothesis. For example, a plausible alternative hypothesis is that the RSF, having emerged from its ice cover at 18 ka, remained stable (with F > 1 but decreasing due to weathering) until it failed about 12 ka and translated while restrained only by basal and lateral friction, remaining largely intact as it did so.

Recommendation

Revise Ms to address these comments (or rebut them) and resubmit.

Reference

Davies, T.R.H., McSaveney, M.J. and Beetham, R. D. (2006). Rapid block glides – slide-surface fragmentation in New Zealand’s Waikaremoana landslide. Quarterly Journal of Engineering Geology and Hydrogeology, 39: 115–129.

Further comments (refer to annotated Ms)

• Abstract lines 29-31: Stability analysis considers only the situation up to failure – it tells us nothing about whether post-failure motion was “catastrophic” or not
• Line 310: omit “and”
• Line 404: the word “rapid” states that all events with low H/L are rapid, which is probably true; however the implied converse, that all events with high H/L are slow, only applies when “slow” is means of the order of cm/minute to cm/second. It does not imply “extremely slow” as is hypothesised for Great Coum
• Line 508: yes there is parameter uncertanty – it needs to be quantified.
• Lines 556-8: Incorrect use of tonnes as unit of force
• Lines 600-605:Confusion in reference to “cirque erosion”
• Lines 660-1: variation of ice strength with temperature is probably negligible in the context of the real precision of the present analysis
• Lines 705-707: Not all glacial ice carries erratics