the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 28 Sep 2023
Abrupt warming and alpine glacial retreat through the last deglaciation in Alaska interrupted by modest Northern Hemisphere cooling
Jason P. Briner
Nicolas E. Young
Joerg M. Schaefer
Abstract. Alpine glacier-based temperature reconstructions spanning the last deglaciation provide critical constraints on local-to-regional climate change and have been reported from several formerly glaciated regions around the world yet remain sparse from high northern latitude regions. Using newly and previously 10Be-dated moraines, we report paleo-glacier equilibrium line altitudes (ELA) for 15 time slices spanning the Last Glacial Maximum (LGM) to the Little Ice Age (LIA) for a valley in the western Alaska Range. We translate our ELA reconstructions into a proxy for summer temperature by applying a dry adiabatic lapse rate at each reconstructed ELA relative to the outermost LIA moraine. We observe ~4 °C warming through the last deglaciation at our site that took place in two steps following initial gradual warming: ~1.5 °C abrupt warming at 16 ka, ~2 kyr after global CO2 rise, and ~2 °C warming at ~15 ka, near the start of the Bølling. Moraine deposition and modest summer cooling during Heinrich Stadial 1 and the early Younger Dryas (YD) suggest that despite these events expressing more strongly in wintertime, the classic blueprint of North Atlantic climate variability extends to the western Arctic region.
- Preprint
(2318 KB) - Metadata XML
- BibTeX
- EndNote
Joseph P. Tulenko et al.
Status: final response (author comments only)
-
RC1: 'Comment on cp-2023-75', Anonymous Referee #1, 13 Oct 2023
Dear editor,
Thank you for the opportunity to review the manuscript by Tulenko et al. The authors provide a thorough chronology for the late Holocene moraines in the Revelation Mountains and pair their cosmogenic exposure ages with ELA modeling to parse deglacial to Holocene summer temperature change. This work is an important contribution to the field in that it combines moraine exposure dating with glacier and ELA modeling to derive temperature change, which is not commonly undertaken. The glacier modeling accounts for some of the issues that arise from inferring climate from dated moraines alone, such as how to account for glacier hypsometry; their methods also provide a temperature record outright which is an intriguing result from the LGM to the Little Ice Age.
I recommend that the manuscript is accepted with minor revisions. The cosmogenic exposure dating follows well-established methods, and the study area has already been published on twice and referenced accordingly. The ELA modeling, though less of my expertise, also follows published methods used by the alpine glacier community.
There a few small areas where the manuscript could in my opinion be improved. First, I would like there to be a table that includes the data to calculate nuclide concentrations. Secondly, I would like to see a visualization to help assess moraine age robustness and outliers. I suggest adding a PDF plot that shows all the samples from the late Holocene moraines at once. Third, I also make some notes about the role of precipitation in the Holocene versus the deglacial period in Alaska that comes up in the discussion. Overall, I think the authors do a good job in the discussion of accounting for the most obvious line of contention: the adiabatic lapse rate and how it may have changed through time.
In sum, the paper is clear and well-written; it is a contribution to the field that should be published. Below, I provide my specific comments as line-by-line notes. I thank the authors for their time in preparing the manuscript.
Line-By-Line Comments
Line 58. ‘South’ should be lowercase.
Line 64 and 66. I’m not totally clear on the use of the word ‘surveyed’ in this paragraph, really between Line 64 and Line 66. “After a thorough moraine dating campaign of deglacial moraines in the valley, we further surveyed…” It seems like the first usage is about selecting a moraine series to date, whereas the second usage on Line 66 is about a more quantitative surveying, namely mapping and dating. Just a note.
Line 70. Blending Methods and Results is fine by me in Sections 3 and 4. That said, could you mention in the subheadings for Sections 3 and 4 something about the Methods? E.g. ‘Cosmogenic exposure dating and late Holocene glacier fluctuations in Alaska’ or ‘ELA Modeling and climate change in the Revelation Mountains through deglaciation’ (and variations more creative than that). Would help for organizational clarity.
Line 82 (Figure 2). As someone who works in the Holocene, I find the convention of using calendar dates in Figure 2 confusing. I would prefer to see the ages in annums or ka. The ages that go negative are quite difficult to understand, and commonly the convention is to present as exposure duration. I don’t mind using calendar dates in the text, however, especially in the Discussion when other climate records are brought in (in fact it’s quite helpful there). Consistency between the ‘Younger Dryas’ moraine (F2-IV E, reported in exposure duration) and the Holocene moraines (reported in calendar dates) is another reason to stick to exposure duration. It is also how results are presented in Table 2. As a suggestion, I’ve seen other deglacial to Holocene papers where ‘ka’ is used for ages > 1,000 years and ‘a’ for ages < 1,000 years and liked that convention. Up to the authors and editor, however.
Line 93 (General comment on data replicability). I appreciate that all the samples are already on ICE-D and age recalculation is easy. However, there does not seem to be any of the raw data necessary if I want to recalculate the concentration of 10Be atoms in the sample, which is necessary for verification. In particular, the reported uncertainties seem low to me; it may be completely fine, but as is I cannot recalculate nuclide concentrations and independently verify their work, which is one of my objectives when I review a paper. Could the authors add a table (to a supplement is fine) with the requisite information? Data such as 10/9Be ratio from the AMS, grams of quartz used (7 g versus 40 g will impact measurement uncertainty, which I currently cannot evaluate), carrier concentration, blank 10/9Be ratios, etc. It ensures the longevity of the data and is necessary for accordance with standards for geochemical labs, data replicability, etc.
Line 94. Would it be possible to put Table 1 here? Near Line 108 would also work for me. It feels out of order as is. It’s useful for interpreting the paragraph from Lines 108–116, so at least right after that paragraph I’d find helpful.
Line 102. This was said already, or at least nearly so, on Lines 78–81. Consider combining these and streamlining.
Line 106. Circling back to exposure ages vs. calendar date, I’ve thought about this a good bit and think it is important the data are reported as exposure ages, simply because a year in time is not an exposure age. Here is my logic: The convention is to report ages in years of exposure, or exposure age. That differs from specifying a point in time in which samples were exposed, or a calendar year. Since the measurements are referred to as ‘exposure ages,’ they should be reported as exposure duration, in years or annums. At the very least, this is what ‘exposure age’ implies in the literature. The authors are absolutely free to provide calendar dates in the parenthetical (like on Line 108) and Table 1, or even completely switch to that format once they provide the exposure ages. Again, I think it is even helpful to do so for comparison with the dendrochronology data, and a good idea by the authors. As written, I am not sure it is correct, but either way I think it would improve the manuscript to start with exposure duration before switching to calendar year of exposure.
One other note on this subject is that the date of sampling is not presented in the paper. I see from ICE-D that it was in 2019, but with the calendar year, is that value being subtracted from 2019? (I think so, but then there’s radiocarbon’s conventions it being reported relative to 1950, etc etc. The clarity will help, that is all.) Would be worth saying ‘samples were collected in 2019’ around Line 71.Lines 110–116. I find it difficult to evaluate the robustness of the moraine ages as presented. I request that the authors add a PDF plot that shows all the data at once. In Figure 5, the older outliers are not on the plot. I think this is an important plot to add, since n = 5 on each moraine and on one moraine (LH-2), 3 of 5 ages are discarded as outliers. A PDF plot could go here or after Lines 150–155.
Line 119 (Figure 3). Same thing, I recommend switching to years of exposure.
Lines 149–150. Citation needed.
Line 189 (Figure 4). Might be nice to add the moraine ages in the numbered side-panels. Helps for interpretation. Just some small text with age and uncertainty in the lower-right-hand corner of each panel, for example. This is subjective, though; feel free to ignore.
Lines 199–202. This is a really interesting finding and a contribution to the field, assuredly.
Lines 221–226. I’m not sure about this. This gets into my general comment about the role of precipitation. While I do think the authors do a good job addressing the elephant in the room (adiabatic lapse rate changing between Deglacial and Holocene), I had to look into the references to fully understand the authors’ point here. Could you add a sentence or two clarifying what specifically you draw from these references? I discuss this further in the next line comment for Line 246…
Lines 246–247. This a surprising result, really. Commonly the YD is considered much colder than the LIA; this is clear from the Greenland ice core record the authors plot in Figure 7. To me, it is completely possible that a wetter Holocene may explain why the YD moraine and the late Holocene moraine are only 0.1° C apart in temperature space (assuming no precip change). A colder, dryer deglacial leading to a warmer but wetter Holocene could potentially cause advances that approach the YD moraines. I think it matters which way the authors are interpreting the precipitation change during the YD and into the Holocene, and therefore my only request is a bit more clarity to this end. I acknowledge that the authors make clear their preference for interpreting ELA change as overwhelmingly summer temperature, but this is not agreed upon in the literature. Sea level (source moves closer), ocean circulation, and the jet stream changed dramatically from the deglacial to the Holocene; it is not inconceivable to me that precipitation could be significantly different. The authors support their claims well and I have no problem with their interpretation; it is what the Discussion section is for.
The authors prefer ~4° C LGM to LIA change because it matches other local records, but as they correctly point out, using a more standard lapse rate lowers this value. To me, a significantly different precipitation regime in the Holocene could explain this mismatch.Line 250. Same thing, can you explain this? 1 more sentence maybe. If the YD is much colder than the LIA in the North Atlantic (Figure 7), why does your record—which suggests the LIA is nearly as cold as the YD—suggests North Atlantic climate forcings extend to Alaska? Later (Line 261) you say the response is dampened. What’s written is not necessarily in conflict, just seeking some clarity.
Line 269 (Figure 7). I’m not sure about the dashed lines temperature reconstructions in Figure 7. Or said another way, the Revelations reconstruction has the actual moraine ages plotted and the dashed lines superimposed on it, which is reasonable to me. The Alps data too seems to have minimal inference between points and clearly lays out the data (moraines, which ELAs are derived from) as a solid line vs interpretation as a dashed line. The Ohau Glacier and Patagonian glaciers are only the dashed lines. Can you make this consistent between the plots? Something to make the Ohau and Patagonia more like the Revelations data or the Alps data.
I also think the paper would be strengthened by including some sort of precipitation data in Figure 7, perhaps the data referenced in Viau et al., 2008 or Kaufman et al., 2010. This is not a requirement, just a suggestion.Again, the paper is clear and well-written; I recommend minor revisions. Thank you to the authors for their work.
Citation: https://doi.org/10.5194/cp-2023-75-RC1 -
RC2: 'Comment on cp-2023-75', Anonymous Referee #2, 26 Oct 2023
Dear editor,
I am pleased to provide a review for this manuscript.
This study uses glacier reconstructions to investigate the climate during deglaciation in Alaska. The manuscript makes an interesting contribution and is likely suitable for publication following a few minor modifications/clarifications.
Here, I focus my comments most on the glacier/ELA modelling component of the study.
The areas where I recommend some modification:
- In the glacier/ELA modelling several methodological decisions are made without being clearly explained and without the impacts of these decisions (on study results) being tested. For example, ELA is calculated using the AAR method. However, several studies highlight the merits of using the AABR method instead (i.e. it better approximates measured ELAs for modern glaciers). I suggest calculating ELA using the AABR (alongside AAR if required) at least to demonstrate the impact this has on results.
- On page 10, you mention setting the basal shear stress to 1 hPa. I presume this value is 100 kPa? In addition, there is no mention of the sensitivity of results to different shear stress values.
- The paper’s focus is temperature reconstruction. Past precipitation is mentioned but only really to justify the choice of lapse rate. In practice, if past precipitation differed from present and/or changed throughout deglaciation then the temperature reconstructions presented here would be quite different (since precipitation impacts glacier dimensions). I feel there is scope for further consideration of this impact.
- Linked to the above, presumed former aridity is used to justify a large (ish) lapse rate, but there is little mention of the impact this continentality might have on glacier dimensions. Where the amplitude of annual temperatures varies through time (i.e. where continentality changes), glacier dimensions can fluctuate without any corresponding changes in summer temperatures (see Golledge et al., 2010). I think this is something to discuss/consider.
Golledge, N., Hubbard, A. and Bradwell, T., 2010. Influence of seasonality on glacier mass balance, and implications for palaeoclimate reconstructions. Climate Dynamics, 35, pp.757-770.
Citation: https://doi.org/10.5194/cp-2023-75-RC2
Joseph P. Tulenko et al.
Joseph P. Tulenko et al.
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 273 | 79 | 8 | 360 | 3 | 7 |
- HTML: 273
- PDF: 79
- XML: 8
- Total: 360
- BibTeX: 3
- EndNote: 7
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1