¹⁰Be exposure dating of the timing of Neoglacial glacier advances in the Ecrins-Pelvoux massif, southern French Alps

Highlights

• A new TCN-based Neoglacial glacier record for the western European Alps.

• At least five LIA-like advances occurred during the last 4200 years.

• Direct dating of a prominent glacier advance coeval with the 4.2 ka event.

• Other LIA-like advances occurred 3.7 ka, 2.1 ka, 1.3 ka and 0.9 ka ago.

• We review NH evidence for glacier advance related to the 4.2 ka event.

Abstract

Alpine glacier variations are known to be reliable proxies of Holocene climate. Here, we present a terrestrial cosmogenic nuclide (TCN)-based glacier chronology relying on 24 new ¹⁰Be exposure ages, which constrain maximum Neoglacial positions of four small to mid-sized glaciers (Rateau, Lautaret, Bonnepierre and Etages) in the Ecrins-Pelvoux massif, southern French Alps. Glacier advances, marked by (mainly lateral) moraine ridges that are located slightly outboard of the Little Ice Age (LIA, c. 1250-1860 AD) maximum positions, were dated to 4.25 ± 0.44 ka, 3.66 ± 0.09 ka, 2.09 ± 0.10 ka, c. 1.31 ± 0.17 ka and to 0.92 ± 0.02 ka. The ‘4.2 ka advance’, albeit constrained by rather scattered dates, is to our knowledge exposure-dated here for the first time in the Alps. It is considered as one of the first major Neoglacial advance in the western Alps, in agreement with other regional paleoclimatological proxies. We further review Alpine and Northern Hemisphere mid-to-high latitude evidence for climate change and glacier activity concomitant with the ‘4.2 ka event’. The ‘2.1 ka advance’ was not extensively dated in the Alps and is thought to represent a prominent advance in early Roman times. Other Neoglacial advances dated here match the timing of previously described Alpine Neoglacial events. Our results also suggest that a Neoglacial maximum occurred at Etages Glacier 0.9 ka ago, i.e. during the Medieval Climate Anomaly (MCA, c. 850-1250 AD). At Rateau Glacier, discordant results are thought to reflect exhumation and snow cover of the shortest moraine boulders. Overall, this study highlights the need to combine several sites to develop robust Neoglacial glacier chronologies in order to take into account the variability in moraine deposition pattern and landform obliteration and conservation.

Introduction

The considerable quasi-global glacial wastage that has been taking place for more than a century is currently accelerating in the European Alps (Vincent et al., 2017) in response to a warming trend that is higher than the hemispheric average (Auer et al., 2007). Glacier mass changes measured over the last three decades cannot be explained without accounting for anthropogenic forcing (Marzeion et al., 2014). Conversely, glacier mass losses immediately following the end of the Little Ice Age appear to have been primarily driven by natural forcings (Vincent et al., 2005, Lüthi, 2014, Marzeion et al., 2014, Sigl et al., 2016). The knowledge of the timing and amplitude of Holocene LIA-type glacial events and the underlying natural climate forcings could improve our understanding of the glacier evolution and responsible climate drivers since the LIA. Though, the temporal and spatial patterns of such glacial events remain poorly constrained, making causes highly debated (Wanner et al., 2011, Solomina et al., 2015, Solomina et al., 2016). A better spatio-temporal knowledge of Holocene climate characteristics is necessary to decipher forcing factors and improve climate model simulations (Schmidt et al., 2014, McCarroll, 2015).

Small to medium-sized alpine glaciers react sensitively to short-term climate variations (Johannesson et al., 1989, Oerlemans, 2005, Six and Vincent, 2014, Roe et al., 2017). This has last been evidenced in the Alps by the decadal-scale period of minor climate deterioration in the second half of the 20th century that led to significant glacier advances and moraine deposition in the early 1980s (Patzelt, 1985). The glacial-geologic record (i.e. moraine ridge stratigraphy) is hence often considered as one of the most straightforward climate proxies beyond the instrumental period (e.g. Putnam et al., 2012, Young et al., 2015). However, full use of this archive as paleoclimate proxy is restricted by its discontinuous nature, by the selective preservation of glacial deposits and by possible non-climatic controls on moraine deposition, such as glacier stagnation or advance due to coverage of the glacier tongue by rockfall debris.

The glacial geologic record could differ between nearby sites due to different glacier hypsometry and ice-flow dynamics (Winkler et al., 2010, Barr and Lovell, 2014). For this reason, dating of moraine complexes provides not only insights into the timing of past glacier-friendly periods, but also allows interpreting the glacial record in terms of landform deposition, conservation and obliteration (Gibbons et al., 1984, Kirkbride and Winkler, 2012, Barr and Lovell, 2014). Well-distributed regionally-significant chronologies based on a homogeneous set of glaciers (size, hypsometry) are therefore requisite to obtain a more complete picture of glacier advances in response to local climate changes. This is particularly true with regard to the Neoglacial, a period during which the glacier advances were of nearly the same magnitude throughout the Northern Hemisphere (Solomina et al., 2015), favouring self-censoring of the moraine record. These chronologies will also permit to assess if climate was the main driver of the glacier variations, because non-climatic controls should not affect several glaciers simultaneously.

There is still a dramatic lack of direct constraints on Holocene glacier variability in the westernmost Alps. Investigations on the French side of the Alps have mainly focused on the Mont Blanc massif (see Le Roy et al., 2015 and references therein). The Ecrins-Pelvoux massif (hereafter EPM) is the south-westernmost currently glaciated area in the European Alps – with the exception of isolated and soon-disappeared glacierets located in the Belledonne massif to the west and in the Ubaye and Maritime Alps massifs to the south (Fig. 1a). To deal with the problem of representativeness and selective preservation of moraine deposits we sampled the outermost Holocene ridges at four different glacier forefields located throughout the EPM (Fig. 1b). The aim of this study was to obtain a local-scale overview of the timing of Holocene/Neoglacial maxima. We present ¹⁰Be exposure ages constraining glacier advances that were almost as extended as the late-LIA stages, marked by the so-called ‘AD 1850 moraines’. Given the location of the dated moraine segments, at most a few tens of meters outboard of the LIA extent, we consider in the following text that glacier size during these advances was virtually the same as during the ensuing LIA maxima.