the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
Equilibrium line altitudes of alpine glaciers suggest summers in Alaska were not more than −2 – −5 °C colder than the pre-Industrial during the Last Glacial Maximum
Jason P. Briner
Joseph P. Tulenko
Stuart M. Evans
Abstract. The lack of continental ice sheets in Alaska during the Last Glacial Maximum (LGM; 26–19 ka) has long been attributed to arid and relatively warm summer conditions. Records of this aridity across Alaska are relatively abundant, yet quantitative temperature reconstructions have been comparatively lacking until recently. Climate model outputs, a few isolated paleoclimate studies, and global paleoclimate synthesis products show mild summer temperature depressions in Alaska compared to much of the high northern latitudes. This suggests the importance of summer temperature in controlling the relatively limited glacier growth during the LGM. We present a new statewide map of LGM alpine glacier equilibrium line altitudes (ELAs), LGM ∆ELAs (LGM ELA anomalies relative to the Little Ice Age [LIA]), and ∆ELA-based estimates of temperature depressions across Alaska to assess paleo-precipitation and -temperature conditions. We mapped glacier extents and reconstructed paleoglacier surfaces in ArcGIS to calculate ELAs using an accumulation area ratio (AAR) of 0.58 and an area-altitude balance ratio (AABR) of 1.56. We calculated LGM ELAs (n = 480) across every glaciated massif in the state, excluding areas in southern Alaska that were covered by the Cordilleran Ice Sheet. We see a similar trend of increasing ELAs from the southwest to the northeast during both the LGM and the LIA indicating a consistent southern Bering Sea and northernmost Pacific Ocean precipitation source. Our ∆ELAs from the Alaska and Brooks ranges, and the Kigluaik Mountains, average to −355 ± 176 m, well above the global LGM average of ca. −1000 m. Using atmospheric lapse rates, we calculate minimum summer cooling of −3.5 ± 1.7 ºC and maximum summer temperature depressions of −1.9 ± 0.9 ºC. Our results are consistent with a growing number of local proxy reconstructions and global data assimilation syntheses that indicate mild summer temperature across Beringia. Limited summer temperature depressions could be explained by increased incoming solar radiation across Alaska during the LGM. Limited summer temperature depressions – and general aridity – in Alaska during the LGM have been previously hypothesized as resulting from the complex influence of North American ice sheets on atmospheric circulation.
- Preprint
(5303 KB) -
Supplement
(12940 KB) - BibTeX
- EndNote
Caleb K. Walcott et al.
Status: open (until 07 Jun 2023)
-
RC1: 'Comment on cp-2023-20', Anonymous Referee #1, 13 May 2023
reply
Title Equilibrium line altitudes of alpine glaciers suggest summers in Alaska were not more than −2 – −5 °C colder than the pre-Industrial during the Last Glacial Maximum
Summary
This study presents a state-wide compilation of equilibrium-line altitudes (ELAs) for Last Glacial Maximum (LGM) glacier extents in Alaska and ELAs for select Little Ice Age (LIA) glacier extents. The authors use these data to discuss paleoclimate implications for the spatial trends in ELAs during the LGM and LIA, and quantify the difference in ELAs between the LGM and ELA to estimate the LGM temperature depression from pre-Industrial. Using past glacier extents or glacial geologic features to reconstruct spatial patterns in glaciers during the LGM across Alaska has been done before, however, previous studies are based on outdated and inconsistent methods, and/or have been applied to smaller regions. This works provides a more comprehensive and consistent approach, and generates what appears to be a comprehensive state-wide LGM dataset.
Another novel aspect of this study, is the determination of LGM ΔELA values by comparing LGM glacier extents to LIA glacier extents. This approach differs from that of previous studies that have different baselines, but often use modern glaciers, which are not in equilibrium with climate. While I was intrigued to see this approach, the mapping of LIA glacier extents is not as comprehensive, and concentrated almost exclusively in the Alaska Range, making statements about state-wide LGM-LIA temperature depressions (including in the manuscript’s title) seem overstated. The poor coverage of LIA glacier extents is a major weakness of the study, and most conclusions about the LGM-LIA temperature depression would not be as believable/justified without references to previous work from these regions (based on different methods).
Overall, this study will be of interest to the CP readership and presents high quality data. Although the ultimate findings and paleoclimate implications are not tremendously different from previous works, this study does provide an updated assessment based on LGM data compiled data from across the state using more recent methods for past ELA reconstructions, and has results that are compared to the most recent proxy and paleoclimate model data. I do not think the limits of the LIA mapping should prevent this from being published, but I do recommend additional mapping if possible or at least that more information be provided about the mapping of both LGM and LIA glacier extents as described below in my two major comments below.
ALso, I commend the authors for an extremely well-written and structured manuscript. It was very clear, for the most part, and I have only a few minor comments. I think the updated assessment of LGM glacier extents and the characteristics of LGM temperature and precipitation changes in Alaska are well documented here and will be well received by the community.
Major Comments
1. More information/justification is needed on the LGM and LIA datasets used for this study.
-LGM glaciers: In Section 3.1 you state that “…we used the Alaska PaleoGlacier Atlas v2 to guide our mapping of LGM ice extents.” Were LGM glaciers mapped that are not in the Atlas? If so, information should be presented to document how that was done. If not, be clear in the text that ‘all’ individual LGM glaciers mapped are from the Alaska PaleoGlacier Atlas v2.
-LIA glaciers: (i) It is not clear at all how LIA glacier systems were chosen for mapping. Are any of these glaciers systems/moraines chosen based on previous mapping by the authors, or other published information that should be cited (e.g. Sikorski et al., 2009)? (ii) Overall, the LIA data are quite insufficient outside of the Alaska Range. It seems a bit of an overstatement to say that you have calculated ‘state-wide’ LGM-LIA ΔELAs when 90% of the data are from the Alaska Range. Only two of the LIA glacier systems are located outside of the Alaska Range….is this due to a lack of identifiable LIA moraines at other locations in the Brooks Range or the other massifs (which seems hard to believe). Is it possible to include more LIA glacier extents? At least, more information should be provided on the LIA data used, and possibly to temper statements about state-wide trends (or provide more reference to previous work before making interpretations about state-wide trends).
2. Background.
The authors do a nice job of summarizing previous work on ELA reconstruction methods and how they have been applied in the past to LGM glaciers in Alaska. However there is no information on the LIA in Alaska. Since the LIA is used as a benchmark for calculating DELAs, there should be some introduction to the LIA in Alaska, including previous work mapping LIA glacier extents, how they compare to the extent of modern glaciers, the timing of the LIA, etc.
Specific Comments:
Page 6, line 1: Not clear how your data will be able to say anything about aridity.
Page 7, line 10: extra “.” before citations
Page 8, line 8: Sentence beginning with “These glaciers deposited….” You should provide references to work that has documented/characterized LIA advances in at least some of these regions in Alaska
Page 14, line 16: The parenthetical statements in the middle of the sentence starting with “In the Alaska Range..” make it difficult to understand the information presented. Consider moving to the end of the sentence or presenting in a different way. Same for how the first sentence of the following paragraph is structured.
Page 14, line 22: “We see no statistical relationship between longitude and ∆ELA.” I don’t think you have enough data from outside of the Alaska Range to suggest that this is significant. If you had better spatial coverage of LIA glacier extents, this would be more of an interesting result.
Page 17, line 14: Sentence starting with “Coversely, in Alaska…” needs a reference. What is the justification that Alaska was drier during the LGM, provide reference.
Page 20, line 19: Use “low” or “small” not both, for describing summer temperature depression
Page 21, line 18: “Our range DELA-derived…” insert ‘of’
Figure 3: A title is not needed since this information is in the caption. Also, the text is too small in the inset with labels for “LGM” and “LIA” points. The x-axis units are also not intuitive and do not make for easy comparison to locations of sites in the other figures… would be easier to understand if they were in degrees longitude.
Citation: https://doi.org/10.5194/cp-2023-20-RC1 -
RC2: 'Comment on cp-2023-20', Anonymous Referee #2, 17 May 2023
reply
This study presents an updated assessment of equilibrium line altitudes (ELAs) for Last Glacial Maximum (LGM) and Little Ice Age (LIA) glaciers in Alaska. ELAs are calculated for the LGM and LIA glaciers and are then used to regional ELA trends across Alaska. Additionally, the ELA data are used to quantify differences in ELAs (ΔELAs) between the LGM and LIA, as well as to estimate the temperature differences between the LGM and LIA. Although there have been several previous studies focused on ELAs in Alaska, this study provides a comprehensive, regional assessment using up-to-date methods. An important aspect is the use of the LIA (rather than modern glaciers) as a baseline for LGM ΔELA calculations. Overall, this study will be of interest to the readership of Climate of the Past.
Although this paper has merits and would be of interest to the CP readership, there are some issues that need to be addressed. I have four main comments on the manuscript.
1. Approach used to delineate the former glacier limits: based on the manuscript, it is unclear how the authors mapped and established the LGM and LIA glacier limits. For the LGM glaciers, the authors mention that the ‘Alaska PaleoGlacier Atlas v2’ was used to guide the mapping of the glacier limits (p. 10), but it is unclear if the authors undertook any original geomorphological mapping. It is also unclear how the limits of the “well-defined LIA glaciers” (p. 11) were identified. There need to be more detailed and transparent explanations of the mapping methods, with reference to established mapping practices (e.g. Chandler et al., 2018).
2. Chronological constraints on the glacier limits: linked to the mapping of the glacier limits, it is unclear how the age of the glacial landforms/limits were determined, as there is no reference to any chronological evidence/constraints. The authors need to provide details on how the ages of the limits were established. Are the ages based on ‘absolute’ chronological data, or are they based on relative methods and/or changes in the glacial landform/landsystem signature? Or is it simply assumed that the limits are either ‘LGM’ or ‘LIA’ glacier limits based on their geographical position?
3. Spatial distribution of the LIA glaciers: the authors indicate that glacier reconstructions were produced for “selected valleys with well-defined LIA glaciers outlines” (p. 11), but it is unclear how these were chosen. Moreover, the limited spatial distribution of these LIA glaciers presents an issue in terms of the regional ELA assessments, as the LIA glaciers are limited to the Alaska Range, Kigluaik Mountains, and northeastern Brooks Range. Given the limited spatial distribution of the selected LIA glaciers, the ΔELA calculations should be restricted to the aforementioned localities. There is insufficient data to justify the state-wide ΔELA calculations, given that 90% (22 out of 24) of the LIA glaciers are located in the Alaskan Range.
4. Uncertainties associated with asynchronous glacier maxima: a key uncertainty in regional ELA assessments is the timing of the ice masses reaching their maxima, and it is generally assumed that all the glacier maxima were reached synchronously. However, this is unrealistic, and the ice masses would almost certainly have reached their maxima diachronously. While it is not possible to quantify and account for this uncertainty, there needs to be some acknowledgement and discussion of this issue in the manuscript.
Specific comments
p. 7, l. 13: I disagree with the statement that ELA reconstructions are a “labor efficient” approach to assessing regional palaeoclimate and that they are “more easily applied to a large region”. While the ELA calculations themselves may not be time consuming, the process of glacier reconstruction is time consuming, especially when dealing with multiple ice masses across a large study area.
p. 8, l. 11: “if” should be of.
p. 10, l. 23: did you apply a uniform basal shear stress value along the whole glacier flowline? Were variations in basal topography accounted for in the basal shear stress values?
p. 11, l. 8: the AABR method is not one of the simplest methods of calculating ELAs, it is one of the more robust and sophisticated methods.
p. 14, l. 5: why are there data for the LIA glacier systems (total numbers, distribution), but not for the LGM glaciers?
p. 14, l. 8: why have you not plotted the LIA ELA data on a map, as for the LGM glaciers?
p. 14, l. 15: a general point regarding the use of the term “temperature depression(s)”. I think it would be better to state “temperature differences” or similar, as the use of temperature depressions does not make any logical sense in this context. The LGM temperature cannot be depressed from the LIA because it occurred before the LIA (assuming that time is linear!).
p. 14, l. 16 ff.: the text in the parentheses make this sentence difficult to read. Explanations of the error values reported would be better placed in the Methods section.
p. 14, l. 21: the ΔELA calculations should be restricted to the localities where there are LIA glaciers, as the limited spatial distribution of the LIA glaciers could skew the state-wide assessment.
p. 15, l. 2: as above, explain the errors in the Methods section.
P, 15, l. 10 ff.: Section 5.1 is excessively long and could be streamlined substantially. Could you present some of the data in a table format and then synthesise the previous work? There is no need to describe each previous study individually.
p. 15, l. 14–15: this compares ELA values at different scales. Why not directly compare the ELA values for the Brooks Range calculated in your study with those from Balascio et al. (2005)?
p. 16, l. 2: it would be better to indicate that the THAR method has been superseded by more robust methods of ELA calculation.
p. 17, l. 19–21: the independent clause after the colon is incomplete.
p. 19, l. 1 ff.: Section 5.3 is excessively detailed and could be streamlined.
p. 22, l. 2: it is an overstatement to suggest that the ELA gradients are similar, given the very limited data for LIA glaciers outside of the Alaskan Range.
Citation: https://doi.org/10.5194/cp-2023-20-RC2 -
CC1: 'Comment on cp-2023-20', Darrell Kaufman, 31 May 2023
reply
Walcott et al. do an excellent job summarizing past work on Last Glacial Maximum ELA reconstructions in Alaska, and they present a new dataset of uniformly calculated ELAs from multiple mountain ranges across the state. Their resulting minor ELA lowering during the LGM and its implication for correspondingly relatively minor summer temperature reductions in this high-latitude region will be of interest to Climate of the Past readers.
I encourage the authors to further consider three aspects of their study:
(1) The authors place their findings in context of Arctic amplification, an important emergent feature of the climate system. However, Arctic/polar amplification is typically defined in terms of annual temperature, which can be related to the energy balance of the earth, whereas the glacier-based reconstructions are likely more strongly correlated with summer temperature. In order to make the link between the temperature reductions in this paper and those globally (i.e., Arctic amplification), it’s important to compare annual to annual temperature. This is especially true because polar amplification is weak in the summer compared with other seasons, so comparing Arctic summer temperature with global annual can be misleading.
Rather than focusing on Arctic amplification, the authors might be better off comparing their Alaska summer temperature lowerings to those from the Atlantic sector of the Arctic and refocus the discussion on the longitudinal asymmetry and its possible causes. Alternatively, they might consider how their ELA-derived summer temperatures relate to annual temperatures so that they can more confidently discuss implications related to Arctic amplification. For example, assuming that seasonality is driven by orbitally controlled insolation, it’s possible that temperature lowering during summer is a maximum estimate for annual changes.
Furthermore, previous studies of polar amplification during the LGM need to be considered. Polar amplification during the LGM is assessed in IPCC-AR6-WGI section 7.4.4.1.2. Based on the proxy data compilation in the report, LGM terrestrial air temperatures (Fig. 7.13f) and sea surface temperatures (Fig. 7.13i) from north of 60° latitude are not noticeably different than the global mean annual temperature (although they are variable). Some site-level LGM proxy records are even warmer than present. The glacier-based inferences in this study, albeit more strongly tied to summer temperature, agree with the IPCC’s compilation, a point worth explaining.
Finally, the Arctic-wide average of -18 ± 7°C (Miller et al., 2010), which is cited on page 21, assumes that summer temperatures in Beringia were very low. Miller et al. featured the estimate from Elias et al. (1996) who concluded that Beringia summer temperature was 20°C lower during the LGM. While I agree that Alaska cooled less than over the ice-covered regions of the Arctic, it’s rather misleading to use an Arctic-wide temperature estimate that assumes much lower temperatures for Beringia than are currently accepted, and that are the point of this paper. Instead, an important implication of this study is that the very low estimate of Arctic-wide LGM cooling needs to be updated.
(2) I agree with the two anonymous reviewers that the study is missing ELA lowering estimates for multiple massifs where Little Ice Age glacier extent has been mapped and need to be added to this analysis for a more complete state-wide analysis. Easy pickings include the Ahklun Mountains, Kenai Mountains, and Brooks Range.
(3) The paper would benefit from some sort of take-away figure (or maybe a table) that summarizes how the ELA-based temperature estimates compare with other paleoclimate information discussed in the text. This could include key boundary conditions during the LGM, especially high-latitude summer insolation, which should be described someplace in the paper.
Minor comments:
Page 4 Line 22: Change “global data syntheses” to “data assimilation”, or more specifically, “blending of proxy-based sea-surface temperatures with a global climate model”. The warm Alaska temperatures in these products strongly reflect the choice of the model prior. This point is explored in detail in Annan et al.’s (https://doi.org/10.5194/cp-18-1883-2022) more recent data assimilation product for the LGM.
Page 5 line 8: Annan et al.’s LGM data assimilation includes terrestrial data. Also, it’s the underlying model that enables the spatial coverage of the DA reconstruction.
Page 5 line 13: “lack of terrestrial records” is an overstatement. Try, “limited availability” or similar.
Page 6 Line 16: Polar amplification is better defined as surface temperature at high latitude compared with the global average or sometimes hemispheric average, not the “mid-latitudes”.
Page 9 line 13-14: What are the specific “features of [these two] previously published studies” that this paper is building on?
Page 21 line 13: Where did this estimate (-5 ± 2°C) of summer temperature change come from?
Page 21 line 15: While I agree that Alaska cooled less than over the ice-covered regions of the Arctic, it’s rather misleading to use an Arctic-wide temperature estimate that assumes much lower temperatures for Beringia than are currently accepted, and that are the point of this paper.
Citation: https://doi.org/10.5194/cp-2023-20-CC1
Caleb K. Walcott et al.
Caleb K. Walcott et al.
Viewed
HTML | XML | Total | Supplement | BibTeX | EndNote | |
---|---|---|---|---|---|---|
329 | 90 | 11 | 430 | 16 | 3 | 3 |
- HTML: 329
- PDF: 90
- XML: 11
- Total: 430
- Supplement: 16
- BibTeX: 3
- EndNote: 3
Viewed (geographical distribution)
Country | # | Views | % |
---|
Total: | 0 |
HTML: | 0 |
PDF: | 0 |
XML: | 0 |
- 1