Author Responses to RC1

 RC1: 'Comment on cp-2024-21', Kathy Licht, 30 Apr 2024  reply 

This paper represents an interesting contribution with new cosmogenic surface exposure ages that complement previous work by the authors.  It addresses the challenging problem of identifying and evaluating long-term terrestrial record of Antarctic ice sheet history.  Such records are essential to pair with offshore and far field data to gain a holistic picture of past ice sheet configurations.  This paper provides useful data documenting past ice sheet stability in the central Transantarctic Mountains and will motivate additional work in the Antarctic interior, thus I recommend it for publication.

I found the manuscript well organized and well written.  In places the text would benefit from more nuance and/or details.  These recommendations and comments are provided below:

This paper is missing reference to and discussion of relevant long-term terrestrial archives that extend through at least one glacial-interglacial cycle in the TAM. These would be valuable to add to the Introduction and/or Discussion section (e.g., Bader et al., 2017; Bergelin et al., 2022; Bibby et al., 2016; Kaplan et al., 2017).  See links to references below.  I recommend also stretching the disciplinary boundaries by briefly considering the implications of your work for life in extreme environments (e.g., Dragone et al., 2021 & 2022).

• The reviewer highlights several recent studies that are pertinent to the glacial-geologic story of the central Transantarctic Mountains, yet which were not included in our original manuscript. Collectively, these papers provide more nuanced context for (a) the timing of past expansions of the EAIS in this region (e.g., in Ong Valley: Bergelin et al., 2022), (b) the relatively minor fluctuations in EAIS surface elevation here on various relevant timescales (e.g., Kaplan et al., 2017; Bader et al., 2017; Bergelin et al., 2022), and (c) the general persistence of cold polar conditions in the central TAM even during the middle Pliocene. We agree with the reviewer that these – and related (e.g., Bibby et al., 2016) – studies would strengthen our revised manuscript and thus will incorporate this material into the Introduction (specifically, to the discussion of cold-based deposits: between current Lines 86 and 98) and also into the Discussion, where we will highlight the strong sedimentary and stratigraphic similarities between our results and deposits at Ong Valley (Bergelin et al., 2022) and Mount Achernar (Hagan, 1995; Bader et al., 2017; Kaplan et al., 2017). The Discussion will also be adjusted to underscore the absence of clear glaciologic evidence from the central TAM for any interruption of cold polar conditions since the early-mid Pliocene (Bergelin et al., 2022) and reiterate the current methodological limitations when assessing the impact of relatively short-lived climate events on the Antarctic interior.
• Recognising that the high elevation and interior position of Otway Massif makes it likely that our study area contains microbe-deficient soils, and that these ecological conditions have probably persisted for millions of years, we are concerned that including such allusions in the revised manuscript would be overly speculative, particularly as the subject is beyond the expertise of the authors. Nonetheless, the documentation of such life-limiting soils elsewhere in the central TAM (Dragone et al., 2021) is an intriguing benchmark of Earth’s habitability and so we will endeavour to include this descriptor as a characteristic of modern surficial geomorphology in our general field area.

Line 83-84.  As written, this implies that the Sirius Group was deposited at the same everywhere, which may not be the case.  Refine the text here.

• We agree with the reviewer and propose to modify the text here to highlight (a) the potentially broad temporal distribution of the wet-based tills associated with the Sirius Group and (b) the fact that existing minimum-limiting ages cannot be assumed to apply uniformly to the Sirius Group.

Lines 86-90.  This section, in particular, requires more nuance in the discussion of polar conditions and cold-based ice.  I read this as referring to most of East Antarctica or at least the TAM, not just your field site, and in that context, the text is misleading. It isn’t true that cold polar conditions at the surface always produces cold-based ice. We know that much of the EAIS has water at its bed (e.g., Siegert et al., 2007; Wright et al., 2017) and at moraines adjacent to your field site, there is geochemical evidence of warm-based basal entrainment (Graly et al., 2018). This doesn’t mean I think the deposits at Otway are warm-based, just that the text about this is not clear about local/thin ice vs the ice sheet more broadly.  Avoid overgeneralizations.

• Again, we agree with the reviewer here and propose to adjust the language of our revised manuscript to reflect the inherently patchy nature of cold-based ice throughout East Antarctica and the adjacent TAM, and to avoid over-generalising the glaciological conditions here.

Lines 94-96.  As presented, this sets up a false dichotomy as regions being compared have several major differences including elevation, distance to the coast, uplift history.

• In this section, we do not seek to restrict discussion to the TAM; the physical nature of East Antarctica (i.e., ice covered) means observations are few and far between, so our discussion here has to extend beyond the TAM. Nonetheless, we appreciate the reviewer’s concern and will adjust the wording of this statement to clarify that (a) environmental conditions vary considerably among EAIS sites and (b) we are not suggesting an ‘either/or’ scenario when it comes to conditions during the Pliocene.

Line 120-121.  What is the adjacent ice thickness?  Can you provide a bed/ice surface topographic profile and comment on what conditions leads to the outstanding preservation at this location?  This would add value in identifying other sites that may provide complementary long-term records.

• While we can provide accurate surface elevations for the Mill Stream Glacier immediately adjacent Otway Massif, there are not currently basal topographic data with which to construct bed profiles. Based on our topographical observations from Otway Massif, and assuming past ice surface profiles were broadly similar to today for the duration of our record, we can postulate that the EAIS margin at Otway Massif was relatively thin, and therefore minimally erosive, due to the low-gradient landscape over which it fluctuated. As at Roberts Massif (Balter-Kennedy et al., 2020), Dominion Range (Ackert and Kurz, 2004), and parts of Reedy Glacier mentioned in the manuscript (Bromley et al., 2010), thin ice margins fluctuating over broad, low-angled terrains are likelier to result in deposition – and preservation – of discrete moraine ridges than are sites where thick ice abuts steeper mountainsides. As suggested, we will add this material to our revised manuscript to (a) enhance the geographical context and (b) contribute criteria for identifying further sites of interest.

Line 310.  “any exposure age on a moraine that is older (younger) than ALL exposure ages on a stratigraphically older (younger) landform”  Have you statistically evaluated what the minimum/ideal value of ‘All’ must be?  I presume one isn’t enough – is two?  Three?  Clarify this point.

• Any statistical estimate of how many samples are enough relies on one’s assumptions about the distribution of inheritance and occurrence of exhumation, which are themselves speculative. To estimate moraine/deposit age, we measured at least three (and typically > 5) samples in order to perform statistics on each population. While this statistical assumption could be viewed as arbitrary, we believe it is nonetheless justifiable and in stating our procedure clearly are enabling readers to fairly evaluate the result. Building upon the reviewer’s comment, we will clarify these points in this paragraph of the revised manuscript.

Paragraph beginning line 322 and related to paragraph beginning Line 385.  Three sandstones were analyzed with paired nuclides and all three showed a complex exposure history and/or erosion. I don’t see three double-dated samples on Fig. 4 – only 2 pairs of pink/blue dots are visible.  With these data in hand, it seems reasonable that ALL samples analyzed have a complex exposure history and/or erosion.  Why would only sandstones suffer this fate?  If erosion is the culprit, should the takeaway be that in ‘old’ glacial deposits, sandstones should not be used for surface exposure dating in these kinds of settings? Are they always unreliable beyond X exposure time?  This wasn’t the finding in Balter-Kennedy, so why might these two adjacent settings produce difference results?  Alternatively, if the culprit is complex burial history, then why would the sandstones be the only rocks affected?  This seems to indicate that all the rocks analyzed should be interpreted as having a complex burial history. This requires more explanation/discussion.

• The reviewer identified an omission from Figure 4, which we shall rectify for the revised manuscript.
• The reviewer raises valid points for discussion concerning the role of complex exposure and/or erosion in our age distributions, and we propose to revise the text in this paragraph accordingly. First, we will clarify that, while erosion is always a possibility for explaining non-normal distributions in Antarctic cosmogenic datasets, the physical appearance of the sandstone boulders from the MOG moraine does not support this mechanism; in contrast to sandstone erratics on older deposits elsewhere in the central TAM (e.g., Balter-Kennedy et al., 2020), the MOG samples are relatively fresh and unweathered. Second, while we concur with the reviewer that all rocks analysed at Otway Massif (not just the sandstones) might have experienced complex exposure, we note that sandstone boulders are extremely rare at Otway, which implies that they derive from only one source. In contrast, the much more common dolerite boulders are likelier to derive from a diversity of sources and therefore exhibit a diversity of inheritance and non-inheritance. Recognising that we cannot currently test for complex exposure in dolerites (owing to the single-nuclide approach in this lithology), we also want to avoid ‘oversteering’ by assuming all samples have been impacted in the same manner; that would be speculative and would ignore the various geomorphic processes that might cause the apparent age spreads (ice-core ablation, deflation, exhumation). To address the reviewer’s concerns on this point, we will clarify the text to support the notion that, whatever impacted the sandstone boulders is also likely to have influenced at least some of the neighbouring dolerite boulders, but as this cannot be quantified for the latter, we can only use the existing approach (age ranges) to assess the likely age for moraine deposition. Specifically, we will modify the text to remove any suggestion that erosion and/or complex exposure is inherent to sandstones but not dolerites.

Lines 375-376.  Newer erosion/exhumation data from this region (He et al., 2021) shows the maximum incision of Beardmore substantially predates these deposits, peaking in the late Eocene and mostly complete by the end of the Oligocene. Re-evaluate landscape evolution with these data in mind.

• We are grateful to the reviewer for highlighting more recent data on glacial-uplift histories in the central TAM, adjacent our study area. To clarify, we do not suggest that selective linear erosion (invoked as a plausible mechanism to explain the apparent long-term ice-surface lowering at Otway Massif) is occurring on the same scale as during the Eocene-Oligocene, which He et al. (2021) reported as the period of maximum glacial incision of Beardmore Glacier. Indeed, the relative ice surface change potentially attributable to isostacy at Otway Massif (and similar sites) is minor relative to the overall topographic relief of these sites, being on the order often of a few tens of metres. Rather, we propose that glacial incision over the course of our record has been relatively minor and gradual, reflecting in part the onset of cold-based glaciation and, if true, helping account for the relatively-minor-yet-persistent pattern we observe in the central and southern TAM. We will adjust the text in this section to clarify our meaning and to stress that the mechanism we invoke is highly localised.

 Figure 2.  Acknowledge the source of imagery.

• This was an inadvertent omission that will be corrected.

Figure 4.  See note about missing data for MOG.  For MON – using the mean doesn’t follow your criteria #1 as there is no ‘main’ population.  Revise this figure and related text to clarify that the ‘mean’ landform cannot be determined.

• We agree with the reviewer and will revise Figure 4 accordingly. We will also remove the mean value for MON given in Figure 2B. Given that both reviewers questioned the applicability of the statistical mean for Antarctic moraine datasets, we propose to refine our interpretation of results using the age ranges, which the text already defends as the most appropriate constraint.

Figure 6.  The gray shading in the background doesn’t make sense.  There are places where error bars overlap, yet there is no gray shading (mostly older than 9 Ma). In the caption specify ‘error’ bars (so it’s not confused with histogram bars).

• The reviewer highlights an error on the figure, which we will rectify for the revised manuscript. The caption will be adjusted accordingly.

Bader et al., 2017: https://doi.org/10.1016/j.quascirev.2016.12.005

Bergelin et al., 2022: https://doi.org/10.5194/tc-16-2793-2022

Bibby et al., 2016: https://doi.org/10.1002/2016GL069889

Dragone et al., 2021: https://doi.org/10.1029/2020JG006052

Dragone et al., 2022: https://doi.org/10.1128/msystems.01330-21

Graly et al., 2018.  https://doi.org/10.1017/jog.2018.4

He et al., 2021: https://doi.org/10.1016/j.epsl.2021.117009

Kaplan et al., 2017.: https://doi.org/10.1130/G39189.1

Siegert et al., 2007:  https://doi.org/10.1002/9781444304435.ch1

Wright et al., 2017: https://doi.org/10.3189/2014JoG13J014

Additional references cited in this response

Bromley, G.R., Hall, B.L., Stone, J.O., Conway, H. and Todd, C.E., 2010. Late Cenozoic deposits at Reedy Glacier, Transantarctic mountains: implications for former thickness of the West Antarctic Ice Sheet. Quaternary Science Review s 29,384-398: https://doi.org/10.1016/j.quascirev.2009.07.001

Hagen, E.H., 1995. A geochemical and petrological investigation of meteorite ablation products in till and ice of Antarctica. Unpublished, PhD thesis, The Ohio State University: https://www.proquest.com/openview/f671b2810060353a923c584227667c34/1?pq-origsite=gscholar&cbl=18750&diss=y

Author Responses to RC2

 RC2: 'Comment on cp-2024-21', Robert Ackert, 05 May 2024  reply 

This article, “East Antarctic Ice Sheet Variability In The Central Transantarctic Mountains Since The Mid Miocene” by Gordon R. M. Bromley et al. reports the distribution of glacial drift at Otway Massif and 52 exposure ages from boulders on lateral moraines of Mill Stream Glacier an outlet glacier of the East Antarctic Ice sheet (EAIS). The exposure ages are primarily 3He in clinopyroxene from Dolerite boulders, but three sandstone samples have paired 10 Be and 21 Ne.  Previously reported cosmogenic data from Roberts Massif and the Dominion Range are compiled to create a data base for the central TAM.  This paper highlights the difficulty in interpreting cosmogenic data from these old (< 1 million years) deposits which typically have large ranges and overlap those on adjacent deposits.  None the less, the data support the authors main conclusion that there is a several million-year depositional hiatus in the late Miocene to early Pliocene in which EAIS outlet glaciers in the central TAM were lower than present. This provides direct constraints of EAIS extent in this region which can be used in conjunction with marine records to access EAIS response to warmer climate regimes.

While certainly suitable for publication, the manuscript could be improved in several respects.  In particular, a more detailed discussion of the factors contributing to the large range of exposure ages from each moraine ridge is called for.

• We agree with the reviewer (and note this point was also made by Reviewer 1) and will revise the Discussion section to explore the potential causes of our age distributions more fully.

The authors do not adequately discuss the effects of erosion on cosmogenic nuclide concentrations, other than briefly mentioning variable erosion rates as contributing to the dispersion of the apparent exposure ages.  It should be stated that all the exposure ages presented are in this sense minimum ages and discuss how that affects interpretations, in particular the constraints on the timing of the non-depositional hiatus. 

• Together, erosion and inheritance can complicate the apparent surface-exposure age of any boulder, particularly in Antarctica where surfaces can be preserved for millions of years. The reviewer requests greater discussion about the influence of erosion on our sampled boulders and we agree that this will strengthen the paper; details of our proposed addition are given below in response to the reviewer’s comment on steady state erosion. With regard to our exposure dates being minimum ages, we argue that this cannot be the case for samples with affected by nuclide inheritance, which would serve to overstate the apparent exposure age. As discussed previously – and recognised by both reviewers – nuclide inheritance is a likely contributor to at least some of our moraine age distributions, and thus the reported moraine ages cannot be reported as minima for deposition.

It is notable that the dispersion of the exposure ages on boulders from the oldest (highest) drift is minimal.  This may partly reflect that the glacier only reached this elevation once, thus resulting in a simple exposure history. 

• It is indeed possible that the high internal consistency of this older deposit reflects a single period of deposition, with no subsequent burial by ice. We are open to including this in the Discussion as a plausible scenario and highlighting the possibility that the broader age ranges of the lower Otway deposits reflects (at least partially) multiple episodes of EAIS advance and deposition. At the same time, we would need to discuss how at neighbouring Roberts Massif there are younger and intermediate moraines with relatively low scatter (Balter-Kennedy et al., 2020), implying that the broader age ranges at Otway Massif do indeed reflect a greater degree of inheritance and/or exhumation.

A reasonable assumption is that these samples are at steady state erosion.  Thus, a long-term average erosion rate can be determined. The implied similar erosion rate of all six samples is likely a result of the common lithology and that boulders with higher erosion rates may have completely weathered away. Quantifying the erosion rate would be of interest for landscape evolution and could be compared to other estimates of Dolerite and other lithologies from the Trans Antarctic Mountains. This erosion rate could then be applied to the younger samples providing a better estimate of the integrated exposure and duration of the hiatus. Assuming the erosion rate falls within previous measurements (10-30 cm/Myr) it is worth keeping in mind that 1 m diameter boulders will disappear within 10 Ma.  Given that > 1 m boulders are relatively rare on the younger drifts, Surface deflation and exhumation of boulders probably contributes to the dispersion of the exposure ages. Finally, if the highest samples are a steady state they could be considerably older than 9 Ma.

• The reviewer makes an interesting suggestion regarding erosion rates, which even in the central TAM are almost certainly greater than zero. However, we note that with samples of this age, and assuming a similar erosion rate to that at neighbouring Roberts Massif (≤ 5 cm/Myr; Balter-Kennedy et al., 2020), the sampled boulders are unable to reach steady state erosion in the time available. The moraines would need to be ~20 million years old for our samples to have lost two attenuation lengths of material and achieved steady state, which we consider highly unlikely in the context of published multiple-nuclide data from the central TAM (Spector and Balco, 2020). Accepting that erosion rates at Otway Massif are greater than zero, we note that the apparent exposure ages for the Otway moraines increase by between 3–10% (younger moraines) and 20–25% (oldest deposits) when an erosion of 2.5 cm/Myr is applied; this is the maximum surface erosion rate for dolerites at neighbouring Roberts Massif (Balter-Kennedy et al., 2020) and thus broadly applicable at Otway Massif. While notable, this shift would not impact our interpretations significantly and, critically, does not change the overall nature of the apparent depositional hiatus, which can still be classified as coinciding with the Late Miocene–Early Pliocene. To address this comment, we propose to (1) add a discussion of the potential impact of erosion rates on the Otway record, drawing from recent, regionally pertinent published estimates, and (2) justify how incorporation of a plausible erosion rate does not change our overall interpretations.

I suggest not using averages for interpreting the exposure ages. Given that the relative contribution of prior exposure vs. erosion exhumation and cover to the scatter in apparent exposure ages on each moraine is unknown, there is really no reason to expect that the true age of the moraine is approximated by the average.  This is especially clear in the case for samples from the Montana Moraine.  Given the exposure ages of boulders range over 2 million years on all but the OLD moraine, it seems unlikely (impossible?) the dispersion is due primarily to erosion and fracturing of boulders. Rather, this is more likely reflecting prior exposure and exhumation. 

• We agree with the reviewer that the mean is not a strong approach when age ranges are broad, as is commonly the case in Antarctica. Moreover, we do not propose that age ranges reflect erosion and/or fracturing of boulders; we agree that much of this can be attributed to inheritance. Following on from our response to Reviewer 1, we propose to remove the mean values given in Figure 2 (map) and to revise our text to focus on age ranges as the most reasonable constraint. With the exception of the MON dataset, we will retain the mean age values reported in Table 1 and depicted in Figure 4 (box plots) as this information is still important to report to readers. However, we stress that the revised manuscript will reiterate our use of the range in estimating true deposit age.

Indeed, although the time scale is much longer, Fig 4 shows the same pattern observed for age-elevation transects of LGM and younger drift. As such it is worth considering a similar interpretation; that the youngest samples may indicate the last time the ice margin stood at that elevation. 

• While we accept the reviewer’s argument here, we avoid using this approach because it implies that geomorphic processes resulting in boulder exhumation (e.g., ice core ablation, deflation), and which therefor give exposure ages younger than the true moraine age, to be negligible. Yet, in these environments, long-term settling/modification of moraines is almost certainly responsible for at least part of the age spreads characteristic of central TAM surface-exposure datasets. To avoid this limitation, we propose to keep using the range as the most reasonable age estimate, but also improve the discussion of our results to include this matter in a fully transparent manner.

In that case, the older samples on each moraine may indicate earlier advances to the same elevation or higher.  Although the surface weathering supports a trend toward lower ice elevations over time, there is no reason to assume that this process was monotonic; all of these moraines may have been overridden more than once, and not all advances necessarily left significant moraine ridges. 

• We agree with the reviewer that the apparent surface lowering is unlikely to have been monotonic and accept that some moraines at Otway may well include ‘composite’ ages from multiple advances of the EAIS. Indeed, this conceptual model potentially makes sense for the MON moraine, which includes ages that align with the older part of the Otway record and ages that align with the younger part. This point will be made in the revised Discussion.

This possibility is supported by the three sandstone samples with paired 10Be-21Ne measurements.  These samples all fall below the simple exposure field (Fig 5) requiring at least one burial event.  Based on this result, it is not unreasonable to assume that most if not all Dolerite samples also have complex exposure histories. Given these observations, greater discussion of the likely complex glacial history and the limitations of single nuclide cosmogenic dating on these old Antarctic surfaces is warranted.  

• We acknowledge the reviewer’s concern here (which aligns with the similar comment by Reviewer 1) and will clarify in the revised manuscript how complex exposure may also have impacted the dolerite boulders at Otway Massif. Nonetheless, we argue that it would be inaccurate to assume all boulders (of both lithologies) share similar exposure histories and reiterate that the scarcity of sandstone boulders at Otway implies they are from a single source, while the dominant dolerite boulders probably originate from numerous sources (and thus exhibit a diversity of inheritance/non-inheritance).

#177: youngest to oldest should be changed to lowest to highest.  Chronology is inferred. In this descriptive section, inference should be avoided.

• We will make these changes as suggested.

#279: Ackert 2002 (here and elsewhere) should be Ackert 2000.

• This error will be rectified as suggested.

#361: Uplands of Antarctic interior.  These are all high elevation sites where significant warming could occur before a change in glacier regime. 

• We will make these changes as suggested.

#381:  I suggest that some landforms change to all landforms, see discussion above.

• We agree and will make this change as suggested.

#393: change to accuracy of single nuclide age estimates in cold-based glacial regimes remains poor and that in most (all?) instances ranges are the only reliable means for presenting ages.

• We will ament this text in the revised manuscript.

#397:  Fig 6.  Given the argument for not using average ages, I would remove Fig 6c.  In any case, it does not seem to provide useful information.  You might consider including some long-term ice modelling results.  See for example Fig 6 of Mukhopadhyay et al 2012.  The timing of modelled ice volume changes is broadly consistent with data presented here indicating resumption of glacial deposition 3-4 Ma. Note also. That modelled ice elevations gradually increase.  This makes sense in that climate gradually cooled.  Why would the earliest post hiatus glacial advances be the largest?

• We agree that the existing panel 6C (mean ages) is not the optimal illustration of these data and propose to replace the mean values with the age ranges, thereby aligning the figure with the text/graphic revisions described above. We do not plan to add model output to Figure 6 as this isn’t a modelling paper: our objective is to present the available geological data that need to be accounted for in climate/ice sheet reconstruction and/or modelling experiments. Moreover, moraine elevation at Otway Massif is almost certainly a composite of EAIS surface elevation and isostatic adjustment, so to pick just one of those (e.g., by incorporating the suggested model output) would misrepresent the full picture.

#452:  change to possible episodes of smaller-than-present ice extent cannot be ruled out from the available subaerial geologic record.

• We will adjust the text as suggested.

Additional references cited in this response

Ackert Jr, R.P. and Kurz, M.D., 2004. Age and uplift rates of Sirius Group sediments in the Dominion Range, Antarctica, from surface exposure dating and geomorphology. Global and Planetary Change 42, 207-225: https://doi.org/10.1016/j.gloplacha.2004.02.001

Spector, P. and Balco, G., 2021. Exposure-age data from across Antarctica reveal mid-Miocene establishment of polar desert climate. Geology 49, 91-95: https://doi.org/10.1130/G47783.1