Early Younger Dryas glacier culmination in southern Alaska: Implications for North Atlantic climate change during the last deglaciation

Abstract

The transition from the glacial period to the Holocene was characterized by a dramatic reorganization of Earth’s climate system linked to abrupt changes in atmospheric and oceanic circulation. In particular, considerable effort has been placed on constraining the magnitude, timing, and spatial variability of climatic changes during the Younger Dryas stadial (YD; 12.9–11.7 ka) within the North Atlantic region. Whereas the YD is clearly expressed in some climate archives, the record of mountain glacier change through the YD remains enigmatic and has elicited debate concerning the overall magnitude and seasonality associated with YD temperature change. Here, we report 19 new 10Be ages from a location in the Pacific sector—the Ahklun Mountains, southern Alaska—that constrain the age of a lateglacial terminal moraine to 12.52 ± 0.24 ka, in the middle of the YD stadial. Our new 10Be ages, combined with additional Northern Hemisphere records of glacier change, reveal that glacier culminations occurred in the early and/or middle YD, followed by glacier recession through the remainder of the YD. Widespread early-to-middle Younger Dryas glacier culminations imply modest summer cooling (i.e., seasonality) that briefly punctuated an overall warming trend through the YD stadial. This pattern of glacier culminations occurring in the early-to-middle YD followed by retreat through the remainder of the YD largely mimics the pattern of YD temperature change displayed in Greenland ice cores. INTRODUCTION Earth’s climate is abruptly changing, and glaciers worldwide are in decline (Roe et al., 2017). The relatively short duration of today’s ongoing climate change event, however, leaves uncertainty regarding its teleconnections around the globe and its regional manifestation in the coming decades and centuries. A longer view of Earth-system response to abrupt climate change is contained within the paleoclimate record from the last deglaciation. Often considered the quintessential example of abrupt climate change, the Younger Dryas stadial (YD; 12.9–11.7 ka) has received significant attention (Alley, 2000; Buizert al., 2014); yet the expression of the YD in records of glacier change, both throughout the North Atlantic Ocean region and beyond, remains ambiguous. The relatively brief duration of the YD, coupled with its abrupt transitions as expressed in ice-core and marine-sediment archives, requires high-precision absolute chronological tools for dating moraines and glacial-sediment records. Many glacial chronologies lack the reso lu tion required to attribute glacier advances to YD climate change (e.g., Gosse et al., 1995; Kelly et al., 2008), thereby inhibiting our ability to constrain the spatio-temporal climatic footprint of YD climate change. However, as newer and more precise glacier chronologies become available, there is increasing evidence that some glacier advances in the Northern Hemisphere, including Norway (Andersen et al., 1995), Scotland (Bromley et al., 2018), and Greenland (Jennings et al., 2014), culminated during the early or middle YD rather than at the abrupt termination of the cold event. Moreover, the apparently limited response of Greenland mountain glaciers, which were located directly adjacent to the canonical record of extreme YD temperature change (i.e., Greenland ice cores), highlights the likely role of seasonality in YD climate change (Denton et al., 2005). There also remains uncertainty regarding the global versus hemispheric climate forcing of glacier change during the last deglaciation (Clark et al., 2012; Shakun et al., 2012; Bereiter et al., 2018). Methodological advances in cosmogenicnuclide exposure dating over the past 15+ yr now provide an opportunity to determine the finer structure of Arctic climate change displayed within moraine records, and in particular glacier response to the YD stadial. Here, we use 19 new high-precision cosmogenic 10Be exposure ages to report an updated chronology of moraines deposited during the Mount Waskey (Alaska) advance, which broadly dates to around the time of the YD, as originally reported by Briner et al. (2002). SETTING In the Mount Waskey region (Fig. 1B), prominent, well-preserved moraines are located ~4 km down-valley of the local cirque glacier complex and mark what is likely a readvance of the valley glacier during overall retreat, as marked by the crosscutting relationship between the Mount Waskey moraines and ice-flow features in the main east-west–trending trunk valley (Briner et al., 2002). The Mount Waskey moraines are, notably, clast-supported features, and they impound a lake; a macrofossil-based radiocarbon age from basal lake sediments constrains the age of the Mount Waskey moraines to >11,010 ± 250 cal yr B.P. (Levy et al., 2004). Briner et al. (2002) originally presented seven 10Be ages from Mount Waskey moraine boulders and a neighboring equivalent moraine with a mean age of 11.31 ± 0.71 ka, after excluding two older outliers (n = 5; see the GSA Data Repository1). The clast-supported structure of the Mount Waskey moraines encourages exceptional moraine stability and largely circumvents issues 1GSA Data Repository item 2019204, materials, methods, and 10Be age calculations, is available online at http:// www .geosociety .org /datarepository /2019/, or on request from editing@ geosociety .org. CITATION: Young, N.E., et al., 2019, Early Younger Dryas glacier culmination in southern Alaska: Implications for North Atlantic climate change during the last deglaciation: Geology, v. 47, p. 1–5, https:// doi .org /10 .1130 /G46058.1 Manuscript received 30 January 2019 Revised manuscript received 26 March 2019 Manuscript accepted 27 March 2019 https://doi.org/10.1130/G46058.1 © 2019 Geological Society of America. For permission to copy, contact [email protected]. Published online XX Month 2019 Downloaded from https://pubs.geoscienceworld.org/gsa/geology/article-pdf/4678802/g46058.pdf by University at Buffalo Libraries user on 07 May 2019 2 www.gsapubs.org | Volume 47 | Number 6 | GEOLOGY | Geological Society of America that have plagued 10Be chronologies in geomorphically unstable environments, such as Alaska (Briner et al., 2005). Moreover, since the initial 10Be measurements of Briner et al. (2002), the cosmogenic isotope community has made remarkable improvements in: (1) isolating 10Be from quartz, (2) accelerator mass spectrometric measurement of the 10Be/9Be ratio (Rood et al., 2013), and (3) precisely constraining 10Be reference production rates (e.g., Balco et al., 2009; Young et al., 2013; Putnam et al., 2019). Combined, the Mount Waskey moraines provide an opportunity to develop a rare, and precise, centennialto millennial-scale record of late-glacial ice extent in Alaska using 10Be. MOUNT WASKEY 10Be MORAINE CHRONOLOGY We present eight 10Be ages from the outer moraine crest (M1), seven 10Be ages from an inner moraine (M3), three 10Be ages from erratic boulders located immediately inboard of the Mount Waskey moraines, and a single 10Be age from a boulder perched on bedrock upvalley of Waskey Lake (Fig. 1B; using Baffin Bay production rate and Lm scaling; Young et al., 2013; see the Data Repository for 10Be details). The 10Be ages from the outer moraine range from 12.60 ± 0.27 ka to 12.39 ± 0.36 ka and have a mean age of 12.52 ± 0.07 ka (n = 7) after excluding one younger outlier (11.34 ± 0.17 ka; >2σ younger than population mean; Table DR1 in the Data Repository). The 10Be ages from the inner moraine range from 12.76 ± 0.32 ka to 11.58 ± 0.24 ka and have a mean age of 12.09 ± 0.44 ka (n = 6) after excluding one younger outlier (9.42 ± 0.20 ka). Three 10Be ages from boulders inboard of the innermost recessional moraine are 11.75 ± 0.25 ka, 11.64 ± 0.22 ka, and 11.59 ± 0.30 ka (mean = 11.66 ± 0.08), and one erratic boulder up-valley of Waskey Lake has a 10Be age of 10.39 ± 0.25 ka (Fig. 1B). DISCUSSION Including the production-rate uncertainty (1.8%; Young et al., 2013), our 10Be ages reveal that the culmination of the Mount Waskey phase occurred at 12.52 ± 0.24 ka—early within the YD stadial (Fig. 2). Following this culmination, the Mount Waskey glacier retreated 1 km through the YD, yet it remained relatively near its YD maximum extent until 11.66 ± 0.23 ka, as constrained by 10Be ages from inside the Mount Waskey moraine complex before retreating up valley (Figs. 1B and 2). Indeed, the Mount Waskey moraine complex appears to represent a response of the Mount Waskey glacier to YD cooling. Records of climate variability in southern Alaska also point to pronounced climate changes through the YD interval. At Arolik Lake (Fig. 1A), a record of biogenic silica—considered a proxy for changes in summertime temperature—clearly captures YD cooling (Fig. 3; Hu et al., 2003). In a review of qualitative lacustrine summer-temperature–sensitive proxy data, Kaufman et al. (2010) provided evidence from several additional sites in southern Alaska showing early YD cooling. Offshore, Gulf of Alaska (GOA) planktonic δ18O values also capture a YD signal (Praetorius and Mix, 2014), and a complementary record of GOA alkenone-based ocean temperatures also reveals pronounced YD cooling, with lowest temperature occurring in the middle YD (Fig. 3; Praetorius et al., 2015). We cannot rule out that changes in winter snowfall contributed to the oscillation of the Mount Waskey glacier. However, pollen records indicate that southern Alaska, including the Ahklun Mountains, remained relatively dry during the YD (Peteet and Mann, 1994; Hu et al., 2002), suggesting that oscillations of the Mount Waskey glacier during late-glacial times were likely driven by summer temperature versus a marked increase in winter snowfall. Although several independent proxy records point to YD-related temperature variability in the southern Alaska–GOA region, the exact nature of this temperature variability is less clear. The mean-annual temperature depression during the YD displayed in Gree