Glaciers preserve climate variations in their geological and geomorphological records, which makes them prime candidates for climate reconstructions. Investigating the glacier–climate system over the past millennia is particularly relevant first because the amplitude and frequency of natural climate variability during the Holocene provides the climatic context against which modern, human-induced climate change must be assessed. Second, the transition from the last glacial to the current interglacial promises important insights into the climate system during warming, which is of particular interest with respect to ongoing climate change.
Evidence of stable ice margin positions that record cooling during the past
12 kyr are preserved in two glaciated valleys of the Silvretta Massif in the
eastern European Alps, the Jamtal (JAM) and the Laraintal (LAR). We mapped
and dated moraines in these catchments including historical ridges using
beryllium-10 surface exposure dating (
The Jamtal and Laraintal moraine chronologies provide evidence that millennial-scale EH warming was superimposed by centennial-scale cooling. The timing of EH moraine formation coincides with brief temperature drops identified in local and regional paleoproxy records, most prominently with the Preboreal Oscillation (PBO) and is consistent with moraine deposition in other catchments in the European Alps and in the Arctic region. This consistency points to cooling beyond the local scale and therefore a regional or even hemispheric climate driver. Freshwater input sourced from the Laurentide Ice Sheet (LIS), which changed circulation patterns in the North Atlantic, is a plausible explanation for EH cooling and moraine formation in the Nordic region and in Europe.
The transition from the Younger Dryas (YD; 12.9–11.7 ka; e.g., Alley, 2000) to the Holocene (ca. 11.7 ka to present, e.g., Walker et al., 2008) is an important period for studying the climate system, its forcings and its feedbacks. Climatic conditions shifted from glacial to full interglacial conditions within approximately 2 millennia, between 12 and 10 ka (e.g., Cheng et al., 2020; Marcott et al., 2013; Rasmussen et al., 2006). This general warming trend was interrupted by abrupt centennial-scale cooling that appears linked to freshwater input into the North Atlantic (e.g., Bjorck et al., 1997; Hald and Hagen, 1998; Nesje et al., 2004; Thornalley et al., 2010). The climatic shift from the YD to the early Holocene (EH) was accompanied by a multitude of major environmental changes that are interconnected, including the melting of ice caps and glaciers in both hemispheres, changes in the atmospheric composition and in circulation patterns, and the reorganization of ocean currents (e.g., Clark et al., 2012; Denton et al., 2021; Shakun et al., 2015). Human-induced warming since the beginning of the industrial era is on a trajectory to lead to changes of similar magnitude in our environmental system, yet at an even faster pace (Beniston et al., 2018; Gobiet et al., 2014). By investigating the timing of YD–EH warming and its perturbations, we can broaden our knowledge on natural drivers and physical mechanisms that modulated the climate system at that time. Information on climate oscillations obtained from this major natural transition – from glacial to interglacial conditions – provides a valuable foundation for disentangling natural and anthropogenic forcings and their respective relevance. New knowledge in this field is particularly useful in the light of the ongoing transition from an interglacial to an industrialized world.
Glaciers respond to climate fluctuations sensitively and are important
elements for understanding the climate of the past (Huston et al., 2021;
Roe et al., 2017). Reconstructing former ice margins allows for deciphering
glacier fluctuations across time and space and informs us of climate
variations that drove these changes. Mountain glaciers in alpine,
melt-dominated regimes are most sensitive to changes in summer temperature and
to a lesser extent to changes in precipitation (e.g., Oerlemans, 2005;
Rupper and Roe, 2008; Steiner et al., 2008). At the end of the Late Glacial
(LG), YD cooling resulted in glacier stabilization or readvance in the
European Alps and led to the deposition of moraine sets, whose estimated
Equilibrium Line Altitudes (ELAs) are approximately 250 to 350 m below ELAs
of glaciers during the Little Ice Age (LIA; e.g., Ivy-Ochs, 2015). These moraines are
termed “Egesen” moraines and have been subject of numerous cosmogenic
nuclide studies that have advanced our understanding of glacier responses
to cooling during the LG (e.g., Cossart et al., 2012; Federici et al.,
2008; Ivy-Ochs et al., 2009, 2006; Kelly et al., 2004;
Kerschner and Ivy-Ochs, 2008). In parallel, the first attempts had been made to
produce direct ages of younger moraines that were identified inboard the
presumable Egesen moraines but outboard historical LIA margins (Ivy-Ochs
et al., 2006; Kerschner et al., 2006). Based on their morphostratigraphy,
these moraine ridges were postulated as type localities for Preboreal
glacier advance, for instance the Kartell moraines in the Verwall area and
the Kromer moraines in the Silvretta Massif, both in the Eastern Alps
(e.g., Faedrich, 1979; Gross et al., 1978). Kartell moraines
are today placed into the latest YD (Egesen-III). Recalculated
To intensify our knowledge on EH glacier configurations in the Eastern Alps, where directly dated moraine ages remain sparser compared to the Western Alps and Central Alps, we conducted a geochronological study in two glaciated catchments in the Silvretta region in the Eastern Alps. We applied state-of-the-art cosmogenic nuclide techniques to date moraines inboard presumable LG ice margins to constrain the timing of Holocene cold phases recorded in the moraine record. We chose this region for two reasons: first, moraine sets that postdate the LG phase are well preserved in the Silvretta Massif and show multi-ridge structures. These geomorphological features promise insights into repeated Holocene cooling at times when glaciers were larger than during the LIA. Second, in addition to comparable cosmogenic nuclide records in the region (Braumann et al., 2020; Ivy-Ochs et al., 2006; Moran et al., 2016b), high-quality paleoenvironmental, archeological, and historical information on Holocene climate, which complements the moraine record, is available (Dietre et al., 2014; Kasper, 2015, 2013; Nicolussi, 2010; Patzelt, 2019).
The primary objective of this study is to generate more detailed and robust moraine chronologies in the eastern Alpine region, which contribute to our understanding of the spatial and temporal pattern of glacier advances throughout the Holocene with an emphasis on the EH. We correlate the new moraine chronologies with moraine records and climate proxy data from the Alpine region and from other glaciated regions in the Northern Hemisphere and identify climate signals that are coherent with Holocene glacier and climate evolution in the Silvretta Massif. Finally, we discuss possible links between climatic trigger events and EH cold snaps, which manifest in the moraine record of the Northern Hemisphere.
The study sites are located at the north-facing side of the Silvretta
Massif, a mountain range in the transition zone between the eastern European Alps and
western European Alps at the border of Austria and Switzerland (Fig. 1a).
Moraine sets from two adjacent valleys, the Jamtal (JAM) and the Laraintal
(LAR), were investigated and used for glacier reconstructions. Both valleys
are north–south oriented, drain northwards into the Danube catchment, and are
at present glaciated only in their highest sections (
Location and lithology of investigated glaciated catchments.
The closest meteorological station recorded a mean annual atmospheric
temperature of 3.1
The Silvretta Massif contains some of the oldest rocks of the Eastern Alps with a presumed depositional age in the Precambrian followed by several metamorphic events (Bertle, 1973; Maggetti and Flisch, 1993). Lithology in the region consists of crystalline rocks, which are part of the Upper Eastern Alpine tectonic unit, more precisely the Silvretta–Seckau nappe (Fuchs and Oberhauser, 1990; Schuster, 2015). Rocks at the Jamtal and Laraintal formed during the Permian and experienced repeated faulting prior to and throughout the Alpine orogeny and are thus metamorphic (Friebe, 2007, and references therein). Predominant rock types in the study area are different gneiss variations and amphibolites (Fig. 1c and d). Rock samples that were collected from moraines had quartz yields – the target mineral for the applied cosmogenic nuclide method – ranging from 0.3 % to 26.1 %, with a median of 3.9 % (Table S1 in the Supplement).
Glaciers erode into bedrock and transport rock material to their margins.
When glaciers are stationary for several years (or longer), linear landforms
– moraines – that consist of glacial debris accumulate at their ice
margins. Dating these moraines unravels the timing of glacier stabilization
or rather the beginning of glacier retreat and allows the reconstruction of
glacier configurations of the past. For
We build upon geological and geomorphological maps, which were produced in
previous studies and which were the basis for further detailed field
investigations in the years of 2018 to 2020 (Fischer et al., 2019, 2015; Fuchs and Oberhauser, 1990; Hertl, 2001). In the
course of a general survey of the Jamtal and the Laraintal area, we updated
preexisting maps according to our own mapping. We then focused on the
mapping of glacial features and placed particular emphasis on the fine
structure of moraines that were presumably deposited during the LIA, and
ridges that were identified outboard these moraines. The dating of these
structures promises to shed light onto the timing of climate perturbations,
which favored moraine formation when glaciers were still relatively large
compared to their present-day configurations. In order to ensure robust
landform age calculations, we took three or more rock samples from each
selected ridge, provided that they fulfilled our sample selection criteria
described in detail in Braumann et al. (2020; their Appendix S-Table 1). In total, 27 samples were extracted from boulders using hammer
and chisel and an electric saw. Geographic coordinates of sampled boulders
were measured using a hand-held GPS device. Strike and dip of sampled
surfaces were quantified with a geological compass. Sample elevations were
taken from the DEM of 2018 (
All samples were processed at the Cosmogenic Isotope Laboratory of the
Lamont-Doherty Earth Observatory (LDEO) following the geochemical standard
protocol for quartz preparation and the extraction of
Exposure ages were calculated using the online calculator formerly known as
the CRONUS-Earth online calculator (v3; Balco et al., 2008). We applied
the regional Swiss production rate (Claude et al., 2014), and “Lm”
scaling to account for site-specific nuclide production. All
Corrections of exposure ages for seasonal snow cover were not applied.
First, samples were primarily taken from boulder tops or their upper
sections, preferably located at windswept locations to minimize potential
snow cover (see Supplement, Sect. S5). Second, if exposure ages were
significantly influenced by snow effects, age dispersion would be expected
among boulders embedded in the same moraine but whose shapes and exposures
vary. Our data do not show a significant bias of this type; therefore, snow
cover effects appear to be insignificant at our study sites. However, if a
snow correction was applied to a boulder assuming a 1 m thick snowpack that
is preserved over 4 months and has a snow density of ca. 0.3 g cm
The preservation of striations on rock surfaces and the general condition of
boulder surfaces in the valleys suggest that erosion has not significantly
impacted their surfaces since deposition. Therefore, all ages that are
presented and discussed in the following represent values without any
erosion correction applied. However, in some studies addressing the Holocene
timescale, an erosion rate of 1 mm ka
In both valleys, distinct moraine sets, which mark Holocene paleo-ice margins, are preserved. We numbered the moraines from the youngest (J0) to the oldest (J5) at Jamtal and moraines at Laraintal from L1 to L5 in analogy.
At Jamtal, the innermost moraine we address in this article is
J0. The moraine was deposited at the left-lateral valley flank,
inboard the presumable LIA moraine (J1; Fig. 2). The age of
deposition of J0 falls into the period between the end of the LIA and the turn of
the 20th century according to Fischer et al. (2019) and a
historical map that was composed in the years between 1870–1877
(K. u. k. Militärgeographisches Institut,
1870–1887). J0 and J1 are punctuated by an approximately 100 m wide drainage
channel, which evolved along the flowline of a former tributary glacier
(Totenfeld, Figs. 3a and A1c). With numerous bedrock outcrops along the
valley flank and a slope of
Holocene moraine chronology of Jamtal. J0 (red) has been deposited after the LIA but prior to the 20th century (Fischer et al., 2019; K. u. k. Militärgeographisches Institut, 1870–1887). J1, J3R, and J4R were dated in this study. Ages along the J1 moraine (pink) indicate that Jamtalferner reached its historical maximum during the second half of the 18th century and during the Neoglacial (NG). Earlier phases of glacier stabilization that exceeded subsequent Holocene culminations are evidenced by moraines J3R and J4R, which both date to the EH. DEM provided by © Land Tirol (resolution 1 m).
Photographs of Jamtal.
J3 and J4, two parallel, curved moraines about 20 to 30 m further uphill relative to J2, are right-lateral moraines of Chalausferner and evidence the convergence of this tributary glacier and the Jamtalferner (Fig. 3c). Boulder surfaces embedded in these moraines are populated with black and green lichens and show signs of weathering, for instance cracks and exfoliation. On the valley floor, a moraine with a frontal position at an altitude of about 2120 m a.s.l. is preserved. The ridge consists of weakly weathered material and is in this respect and with respect to geometry the terminal equivalent of the lateral J1 moraine (Figs. 2 and A2c). This correlation is in accordance with glacier outlines of the Austrian Glacier Inventory (AGI; Fischer et al., 2015), with a geomorphological map compiled by Hertl (2001, p. 226), and with results from a recent study on vegetation dynamics at the Jamtal, which includes ice margin reconstructions since the end of the LIA (Fischer et al., 2019). North of the terminal section of J1 is an area covered with angular and subangular blocks, whose surfaces are significantly more weathered compared to J1 and which exhibit extensive lichen population (Fig. A2c–e). Many blocks have cracks and are fractured, which may indicate impacts associated with gravitational movement. The morphology outboard J1 is convex – unusual for in situ rockfall deposits, which typically form lobate structures with large boulders in frontal positions. We therefore hypothesize that these deposits stem from rock failures along one (or both) valley flank(s) farther uphill. The material collapsed onto the formerly larger glacier and was transported downstream through glacial flow. As the glacier retreated, these rockfall deposits melted out and accumulated on the valley floor. An additional argument supporting this scenario is the provenance area of a potential rockfall event, which could not clearly be identified along the surrounding walls and peaks. The blockfield was in part overprinted by one (or multiple) glacier advances, as evidenced by the position of moraine J1.
A set of ridges in the right latero-frontal section outboard the J1 moraine
appears to be somewhat displaced (Figs. 2 and A2f). A scarp above this
moraine set and a stabilized sliding mass below caused an offset of formerly
connected crests. Together with rockfall deposits from bedrock outcrops in
higher-up sections, these moraines are not considered as prime candidates
for
Evidence of older, LG terminal positions farther downstream is scarce. Hertl (2001, p. 77) describes a lineament that dips to the valley floor about 2.5 km downstream of J5, at an elevation of 1900–1920 m a.s.l. The author tentatively interprets the structure as a latero-frontal moraine deposited towards the YD termination. Around 8 km down valley from J5, a tripartite moraine set (Gaffelar settlement) is attributed to an earlier YD phase but not the YD maximum. Lateral LG moraines are absent in the main valley but are preserved in the Futschöl tributary valley, which joins the Jamtal from the east in the area of the Jamtal hut (Fig. 1b).
At Laraintal, we focused on valley sections outlined in Fig. 1d and
detailed in Fig. 4. Texture, relative positions, and structure of moraines
in this valley resemble moraine sets at the Jamtal. L1 – the presumable LIA
ridge – consists of fresh, sparsely vegetated debris and is traceable along
both valley flanks. On the eastern side, we identified a fine-structured set
of moraines that we refer to as L2, L3R, and L4R in Fig. 4. Similar to J2,
L2 with a width of approximately 8–10 m is less prominent compared to L1
(
Holocene moraine chronology of Laraintal with moraine ages
displayed. The
Photographs of Laraintal.
Analogous to the Jamtal, the terrain at the western valley side is generally
steeper compared to the eastern side, with slope angles of ca.
Approximately 200–250 m further downstream, at an elevation of 2130 m a.s.l., is another set of ridges on one of which a small hut (“Zollhütte”) was built (Fig. 4). This Zollhütte ridge (L5) is framed by a blockfield consisting of a blend of angular and rounded boulders. A massive debris cone west of Zollhütte, which has a layer of coarse and medium-sized fresh material on top, points to continuous sediment supply from the left-lateral wall. To the east, a scarp with a concave surface below indicates former (and possibly ongoing) sliding processes directed towards Zollhütte. Moreover, rockfall events with material disintegrating from the wall below the “Hoher Kogel” peak have been witnessed during field work in the year 2019, with boulder volumes of multiple cubic meters that have crashed on the valley floor (Figs. 1d and A4a–b). Such events have probably also occurred in the past as the wall exhibits multiple lighter sections, which are indicative of removed material and thus of previous rock failures. Due to the manifold processes that impacted the Zollhütte area, the J5 ridge is scientifically risky to tackle with SED of boulders, even though it probably delimits a terminal glacier position.
Kernel plot of LIA ages produced from boulders embedded in the
historical moraine at Jamtal. The gray-shaded bar illustrates the 1
Kernel plots of Holocene moraine ages.
Five samples were collected from boulders along moraine J1 (Fig. 2,
Table 1). Three of them (JAM-18-06, JAM-18-17, JAM-18-18) were deposited
during the second half of the 18th century and yield a rounded mean age
of 260
Sample LAR-19-23 (700
Analytical results from samples, which were spiked with Fe, show that ages
calculated from both samples are consistent with boulder ages obtained for
the same landforms but processed according to the standard protocol (L3R:
LAR-19-13; L4R: LAR-19-12 and LAR-19-15). Analytical uncertainties of
corresponding samples amount to 2.0 % (LAR-19-16) and 2.5 %
(LAR-19-14) and are within the expected range of
The classical LIA moraines of both valleys (J1 and L1) feature boulders
deposited within the expected time interval, i.e. between 1250 and 1850 CE
(Fig. 8f; e.g., Grove, 2004; PAGES 2k Consortium, 2013). An early LIA
advance of Larainferner to L1's position is suggested by LAR-19-23 and may
have occurred at the beginning of the 14th century. Three consistent
boulder ages from J1 are aggregated to a mean age of 260
The youngest part of the
Besides LIA-aged boulders along the LIA moraine, we sampled two blocks of J1
that were deposited during the first millennium of the Common Era. The
younger boulder, JAM-18-16, dates to the beginning of the Medieval Warm
Period (MWP). By that time, glaciers in the region were likely smaller
relative to their LIA maxima (e.g., Solomina et al., 2016). The boulder's
position and its bedding were re-evaluated in the field after age
calculation, and we cannot exclude that the boulder has tilted (Fig. S7).
Therefore, we interpret the exposure age as a minimum age. The older
neoglacial boulder, JAM-18-07, was exposed 1500
In summary, samples collected along the classical LIA moraine at the Jamtal (this study) and at the adjacent Ochsental (Braumann et al., 2020) yield ages that fall into the regional DACP and the LIA. These results are consistent with the timing of glacier advances across the Alps and in other places of the Northern Hemisphere. The advance of Silvretta glaciers coincides with cooling trends captured in local, regional, and hemispheric proxy data. Moraines J1 and L1 are probably composite moraines that have represented ice margins at least once prior to the LIA. J1 and L1, in the following sections termed “Holocene composite moraines”, mark the maximum of glacier advances and corresponding temperature minima since the end of the YD–EH transition.
Moraine records at Jamtal and at Laraintal are remarkably similar and point to synchronous glacier dynamics throughout the Holocene, particularly during its onset. In both valleys, we identified up to three lateral ridges just outboard the J1 and L1 composite moraines and their terminal equivalents, albeit in varying states of preservation. The outermost ridges in both valleys, J3R and L3, and J4R and L4, respectively, yield statistically identical landform ages (Fig. 7a–d). Interestingly, they are chronostratigraphically inverse, i.e., JR3 and L3 are systematically several centuries older than JR4 and L4. An explanation for this age pattern may be decadal- to centennial-scale pre-exposure of J3R and L3 boulders, which would lead to an overestimation of ages. However, if pre-exposure was a problem in the data set, we would expect greater scatter in boulder ages inferred for the same moraine. Another explanation for age inversion is post-depositional displacement such as sacking or tilting of J4R and L4 boulders, for instance through the thawing of permafrost, which would lead to underestimation of corresponding ages. Yet again boulder ages along JR4 and L4 are in good agreement, making this explanation unlikely. A process, which may have affected JR4 and L4 boulder surfaces and could cause a systematic shift towards younger ages, is katabatic winds, when the glacier abandoned moraines JR4 and L4 and halted for a few centuries at the positions of or close to JR3 and L3. Consolidation of the age difference between moraines JR3/L3 and J4R/L4 would necessitate the removal of approximately 2.5 mm of rock from boulder surfaces along the outermost moraines (J4R and L4) within 500 years.
The age pattern of EH moraines in the Jamtal and Laraintal is noteworthy,
but we emphasize that moraine ages of JR3, L3, JR4, and L4 overlap within
1
Proxy records capturing the YD-EH transition.
We correlate EH Jamtal and Laraintal moraine chronologies with moraine
chronologies and glacier proxy records at the local scale and propose a
concept of YD–EH deglaciation. MFI 3–4 falls well into the EH and is
different from advances during the LG. Presumable YD moraines are identified
at considerable distance downstream and outboard of landforms addressed in
this study (Hertl, 2001, pp. 220 and 225), which implies that
glaciers shrank from their LG ice margin to a position close to the LIA
maximum within a few centuries. Rapid deglaciation is a direct response of
glaciers to an increase of summer temperatures by several degrees during the
YD–EH transition in the eastern Alpine region and across the Alps (Fig. 9c–f; e.g., Affolter et al., 2019; Heiri et al., 2014; Ilyashuk et al.,
2009; Larocque-Tobler et al., 2010b; Samartin et al., 2012). This warming
trend was interrupted by brief cold spells, which manifest in moraine
records in the Silvretta Massif and in the adjacent Verwall mountains to the
northeast. Based on these moraine chronologies, the following local YD–EH glacier
history emerges (Fig. 10).
Moraines that formed within this period in the region have been dated at the
Kartell site, which is around 15–20 km northeast of Jamtal and Laraintal
(Ivy-Ochs et al., 2006). Because of the previously
higher The shift
from glacial to interglacial conditions is captured at the Jamtal and
Laraintal. Corresponding moraine chronologies indicate ice margins at
terminal positions some hundreds of meters outboard of the LIA maximum,
equivalent to an estimated ELA depression of approximately 70 m (AAR
method; Hertl, 2001, p. 80). These chronologies evidence abrupt cold snaps,
which punctuated the general warming trend during the EH and which caused
decadal- to centennial-scale glacier oscillations. The mean age of MFI 3–4
(11.0 Kromer moraines identified in valleys 10–15 km further towards the west
(Fig. 1b) were originally placed into the Preboreal (Gross et al.,
1978). Morphologically, these moraines resemble the blocky, multi-ridge
structures of J3R and J4R at the Jamtal and L3 and L4 at the Laraintal and yield similar
snowline depression estimates of 70–90 m. Updated Deglaciation patterns in the Ochsental to the west suggest
ice margin configurations similar to the LIA around 10 ka. The timing of
moraine formation adjacent to the lateral Holocene composite moraine
(equivalent to J1 and L1 in this study) was constrained to 9.9
The timing of moraine formation at the Kartell site overlaps with MFI 3–4,
just as MFI 3–4 overlaps with moraine formation at GrK in the Ochsental.
Instead of indicating individual phases of glacier advance or stabilization,
differences in the landforms' central ages could also be owed to
uncertainties of the dating method, which we cannot exclude, or to
catchment-specific effects such as shading or bedrock topography. We note
that catchments where the evaluated
Glacier retreat during the transition from glacial to interglacial conditions evidenced in moraine chronologies from the Silvretta and Verwall regions. KAR indicates Kartell moraines (Ivy-Ochs, 2015; Ivy-Ochs et al., 2006). The site is located approximately 15–20 km northeast of Jamtal and Laraintal. MFI 3–4 are as described in this study. GrK indicates the Grüne Kuppe moraine identified in the adjacent Ochsental, suggesting LIA-like glacier extents around 10 000 cal BP (Braumann et al., 2020).
Moraine formation during the transition from glacial to interglacial climatic conditions that is presented in Fig. 10 builds on glacier records in the Silvretta Massif and in the Verwall mountains in the Eastern Alps. This model is consistent with results of previous studies that have addressed glacier evolution during the LG and EH based on moraine chronologies (e.g., Baroni et al., 2017; Hofmann et al., 2019; Moran et al., 2016a, 2017; Protin et al., 2021, 2019; Schimmelpfennig et al., 2012, 2014; Schindelwig et al., 2012). Investigated moraine sets may differ with respect to their structure, state of preservation, or distance relative to the LIA maximum, but they share their position (outboard the LIA maximum but inboard the presumable LG ice margin) and their age of deposition between ca. 12 and 10 ka, which implies substantial large-scale cooling during this period. This pattern of EH moraine stabilization is not limited to the Alpine realm but has been observed in other places in the North Atlantic and Arctic region, for instance along the Fennoscandian ice sheet (e.g., Briner et al., 2014; Nesje, 2009), the Icelandic ice sheet (e.g., Sigfusdottir and Benediktsson, 2020), at Svalbard (e.g., Farnsworth et al., 2020), along the eastern part of the LIS (e.g., Corbett et al., 2016; Ullman et al., 2016; Young et al., 2020), and along the Greenland ice sheet (e.g., Biette et al., 2020; Levy et al., 2016; Young et al., 2020) (Fig. 9g). Atmospheric temperatures were certainly different in these regions during the EH (Fig. 9b–f), and glaciers in the European Alps retreated much earlier to positions inboard their subsequent historical margin compared to Arctic glaciers, which continued to deposit moraines outboard their LIA at least 2 millennia longer. However, concurrent moraine stabilization during the EH raises the question what caused this synchronicity in climatic cooling during the first millennia of the Holocene.
Freshwater influx into the North Atlantic and into the Arctic Ocean is known as a driver for climate of the Northern Hemisphere and acts as a plausible cause for abrupt centennial-scale cold snaps during the LG and the EH (e.g., Bjorck et al., 1997; Fisher et al., 2002; Hald and Hagen, 1998; Nesje et al., 2004). Prominent examples are repeated outbursts of the North American proglacial Agassiz lake, whose final drainage caused a sharp temperature drop in the Northern Hemisphere, the 8.2 ka event detected in Greenland ice cores (Alley and Agustsdottir, 2005; Clarke et al., 2009; Hillaire-Marcel et al., 2007; Teller et al., 2002; Thornalley et al., 2010). Besides abrupt high-volume releases of freshwater to the North Atlantic or Arctic Ocean via major lake drainages or iceberg armadas, there is evidence of more subdued glacial discharge during the EH that results in a deceleration of the thermohaline circulation (Bamberg et al., 2010; Renssen et al., 2010; Thornalley et al., 2009). Weakening of the AMOC leads to less heat transported to the North Atlantic region, which can prompt brief decadal-to-centennial-scale cold snaps in the north and also at lower latitudes. During the PBO and subsequent centennial-scale cold phases, harsher climate conditions are reported in the North Atlantic region (e.g., Bos et al., 2007; Knudsen et al., 2008; Paus et al., 2015; Timms et al., 2021), with cooling extending towards western and central Europe (Fig. 9c–f). Glaciers in North Atlantic regions and in the Alps responded to these climate perturbations with stabilization or advance and hence moraine deposition.
To test the linkage between EH moraine formation and freshwater discharge of
the LIS, we review corresponding markers in marine sediment cores,
temperature proxy records, and moraine records in these regions. We begin
with the YD termination, when icebergs and meltwater plumes were released
into the North Atlantic, evidenced by ice-rafted debris and layers of
“foreign” sediment enriched with detrital carbonates in marine sediments.
These layers, often referred to as Heinrich-0 (H0) and characterized by a
detrital carbonate peak (DCP) date to the earliest Holocene, were
identified in Baffin Bay (Simon et al., 2014), in the Labrador Sea
(Andrews et al., 1995; Rashid et al., 2011), and at the coast of Newfoundland
(Pearce et al., 2015) including the Flemish cap (Li and Piper,
2015). Jennings et al. (2015) found multiple subsequent detrital
carbonate peaks (DCP 1–7 in Fig. 9a) between 11 500 and 8000 cal BP in a
core from the Labrador shelf, which is attributed to freshwater sourced from
Hudson Strait. DCP1 detected around 11 500 cal BP coincides with
the end of H0 and with the onset of 11.4 ka event (ca.
11 450 to 11 350 cal BP; Rasmussen et al., 2007). During the subsequent PBO
captured in Nordic records, mean annual and summer temperatures in the
European Alps declined (Fig. 9e–f; e.g., Affolter et al., 2019; Heiri et
al., 2014; Lauterbach et al., 2011). The propagation of this cooling trend
towards western and central Europe is supported by cooler and more humid
climate conditions in these regions ca. 11 300–11 150 cal BP (Magny et
al., 2007). In parallel,
A similar chain of events may have occurred some centuries later. The deposition of the DCP2 ca. 10 600 cal BP was preceded by the so-called Gold Cove advance of the LIS's Labrador sector across the Hudson Strait and its subsequent retreat (Jennings et al., 2015; Kaufman et al., 1993; Rashid et al., 2014). The resulting freshwater input may have weakened the AMOC, which in turn led to a drop in mean annual temperatures in Greenland and moraine formation in the Arctic (e.g., Biette et al., 2020; Young et al., 2020). Temperatures in the Alps decreased or stagnated around that time (Fig. 9c–d, f). Moraine formation between 10 700 and 10 500 cal BP, concurrent with DCP2, is observed across the European Alps (e.g., Moran et al., 2017; Protin et al., 2019; Schimmelpfennig et al., 2012, 2014).
The linkage between DCPs, freshwater input into the Atlantic and Arctic Ocean, and subsequent EH glacier advances has been put forward before in the context of the North American and Arctic region (e.g., Andrews et al., 2014; Nesje, 2009; Young et al., 2020). The authors of a recent geochronological study carried out in the Western Alps go a step further and propose that freshwater forcing in the North Atlantic region acted as a driver for moraine formation in the Mont Blanc Massif (Protin et al., 2021). They suggest that a decrease in AMOC strength led to extended sea ice periods during winter in the North Atlantic, which in turn caused a southwards shift of the westerlies. Cold air was then transported to Europe and led to moraine deposition in the European Alps. With our new moraine chronologies from the Silvretta region, we complement the glacier record from the Western Alps with robust evidence for EH moraine formation in the Eastern Alps and corroborate the hypothesis that centennial-scale cold phases occurred at a regional (hemispheric) scale between 12 and 10 ka.
Glaciers at both study sites, the Jamtal and the Laraintal, stabilized or
advanced at least twice during the EH, which is evidenced by moraines
deposited outboard the historical moraine (LIA maximum) and inboard the
presumable LG ice margin. The timing of moraine formation is consistent
across both valleys and is constrained with Based on Holocene moraine chronologies of the Silvretta Massif and the
adjacent Verwall mountains, we propose the following local model that
describes alternating phases of glacier retreat and stabilization between 12
and 10 ka: the YD termination (Kartell; Ivy-Ochs, 2015; Ivy-Ochs et al.,
2006), the YD–EH transition (MFI 3–4, this study), and the Holocene mode
(GrK; Braumann et al., 2020). The proposed concept
confirms the hypothesis formulated by Patzelt and
Bortenschlager (1973) almost 50 years ago that glaciers in the Eastern Alps
deposited moraines during the Preboreal and had retreated to their
subsequent historical ice margins by ca. 10 to 9.5 ka. During the rest of the
Holocene, the magnitude of cooling was most likely too small in the Eastern
Alps to force advances which exceeded dimensions that glaciers had around 10 ka. Our data suggests that glaciers in the Silvretta region advanced to a
position close or equivalent to their LIA maximum around 500 CE. The timing
of this advance is concurrent with the migration period in Europe, which is
often associated with regional climate deterioration. In the Silvretta
Massif, this hypothesis rests on a few Silvretta glaciers may have advanced to their LIA maximum as early as ca.
1300 CE. A subsequent advance to the same position took place in the second
half of the 18th century, documented by three boulders along the
historic moraine yielding a mean age of ca. 260 The classical LIA moraine in the European Alps, traditionally referred to as the “1850 moraine”, marks the LIA maximum glacier extent that was reached
multiple times during the LIA but perhaps also earlier during the Holocene,
most likely around 500 CE. As the moraine comprises glacial sediments
deposited during several glacier advances during the past millennia, we
propose that these landforms should rather be viewed as “Holocene composite
moraines” instead of “1850 moraines” or “LIA moraines”. Our observations document sensitive glacier response to the natural warming
in the early Holocene. Accelerating anthropogenic warming has been driving
the rapid retreat of highly temperature-sensitive glaciers over the last
century (e.g., Oerlemans, 2005; Zemp et al., 2019) and will lead to the
disappearance of most glaciers in the European Alps unless greenhouse gas
forcing is substantially reduced in the very near future (Zekollari
et al., 2019).
Jamtal.
Jamtal.
Laraintal.
Laraintal.
Overview of Silvretta and Verwall regions addressed in local correlation of moraine records (Sect. 5.2.1).
All analytical information associated with cosmogenic nuclide measurements is listed in the tables within the paper and the Supplement.
The supplement related to this article is available online at:
SMB and JMS designed the study. SMB, MF, and JMS carried out field work. SMB was responsible for cosmogenic nuclide sample preparation. AJH performed sample measurements and developed the strategy for low-level
The contact author has declared that neither they nor their co-authors have any competing interests.
Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation. Publisher’s note: Copernicus Publications remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Sandra M. Braumann is grateful to the LDEO cosmogenic nuclide laboratory staff Roseanne Schwartz and Jean Hanley for their help and advice during sample processing. Furthermore, Sandra M. Braumann thanks Günther Gross for helpful discussions on the Holocene moraine record in the Silvretta Massif, Dominik Blasenbauer for assistance during fieldwork, and Herbert Formayer for providing meteorological data. The authors express their gratitude to their hosts at the Jamtal hut, at Larainalpe, and at Pension Türtscher, who have facilitated the field campaigns. Finally, the authors thank two anonymous reviewers, whose comments greatly improved the manuscript, and Alberto Reyes, who was the editor of the manuscript. This work was performed in part under the auspices of the U.S. Department of Energy by Lawrence Livermore National Laboratory under Contract DE-AC52-07NA27344; this is LLNL-JRNL-821951.
This research has been supported by inatura Museum GmbH Dornbirn (funding period 2017–2019) and is based upon work supported by the National Science Foundation under grant no. NSF-1853881. Sandra M. Braumann is a recipient of a DOC Fellowship of the Austrian Academy of Sciences (OeAW) at the Institute of Applied Geology, University of Natural Resources and Life Sciences (BOKU) Vienna (grant no. 24709). For research visits at the LDEO, Sandra M. Braumann received a Marietta Blau scholarship sponsored by OeAD GmbH (grant no. ICM-2018-11638), a Marshall Plan scholarship provided by the Austrian Marshall Plan Foundation (grant no. 967116921 222019), and financial support from BOKU's “Transitions to Sustainability” (T2S) doctoral school.
This paper was edited by Alberto Reyes and reviewed by two anonymous referees.