Response of the Irish Ice Sheet to abrupt climate change during the last deglaciation

Abstract

We summarize 121 ¹⁴C and in-situ cosmogenic (¹⁰Be and ³⁶Cl) ages that constrain fluctuations of the Irish Ice Sheet (IIS) since the Last Glacial Maximum (LGM) that can be linked to abrupt climate changes in the North Atlantic region. These data provide a robust means to date ice-sheet fluctuations because similar-age events can be identified from widely spaced sites, they are constrained by stratigraphy, and they can be related to large changes in the configuration of a dynamic ice sheet. The following events are recognized. (1) AMS ¹⁴C ages and ¹⁰Be and ³⁶Cl ages as well as offshore data suggest that the last maximum advance of the IIS occurred between ∼27 and 23 cal ka. (2) Deglaciation began on the western continental shelf and in the Irish Sea Basin at ∼23 cal ka. Dated sites from around Ireland constrain subsequent widespread retreat of ice-sheet margins from the shelf onto Irish coastlines during the Cooley Point Interstadial ≥20 cal ka. AMS ¹⁴C ages suggest that the Cooley Point Interstadial continued until ≤18.2 cal ka B.P. During this interstadial, retreat of the ice-sheet margin into the northern Irish Sea Basin indicates that the IIS may have lost up to two-thirds of its mass. (3) Dated sites from northwestern and eastern Ireland indicate that the Cooley Point Interstadial was terminated by ice readvance during the Clogher Head Stadial at ∼18.2 cal ka. (4) The Linns Interstadial was a brief interval (≥17.0–≥16.5 cal ka) of ice recession following the Clogher Head Stadial that is identified from raised marine sediments in eastern and northern Ireland and cosmogenic ages elsewhere in Ireland. (5) The IIS subsequently readvanced during the Killard Point Stadial, reaching its maximum extent ∼16.5 cal ka, indicating that the readvance began sometime earlier. (6) Widespread retreat of the IIS began by ∼15.5 cal ka during the Rough Island Interstadial. (6) Readvance of cirque glaciers in western Ireland occurred during the Younger-Dryas equivalent Nahanagan Stadial.

Introduction

The last glaciation was characterized by abrupt millennial-scale climate changes that involved changes in the Atlantic Meridional Overturning Circulation (AMOC) and associated ocean heat transport (Broecker et al., 1990, Ganopolski and Rahmstorf, 2001, Clark et al., 2002) with a strong feedback from changes in sea ice (Mikolajewicz et al., 1997, Li et al., 2005, Liu et al., 2009). Model simulations with appropriate glacial boundary conditions demonstrate that these changes would cause substantial changes in surface air temperature and precipitation downwind of the North Atlantic (Hostetler et al., 1999, Liu et al., 2009), thus strongly affecting the mass balance of Eurasian ice sheets (Clark et al., 2007).

Because of its proximity to the North Atlantic as well as its small size (<0.5 m sea-level equivalent, or <7% of the modern Greenland Ice Sheet), the Irish Ice Sheet (IIS) should have been particularly sensitive to the abrupt climate changes induced by changes in the AMOC. To a first order, we thus expect the abrupt climate changes of the last glaciation to have caused corresponding fluctuations of the IIS margin. Early models of deglaciation of the IIS, however, inferred monotonic retreat following its last maximum extent, albeit interrupted by intervals when the margin stabilized (Charlesworth, 1928). This perspective changed, however, when Synge (1969) and McCabe (1969) first inferred large shifts in centers of ice dispersion from patterns of streamlined bedforms and cross-cutting moraines and striae. In particular, this reconstruction showed that, after a period of ice retreat, the remaining ice masses readvanced during what Synge (1969) referred to as the Drumlin Readvance, providing the first evidence of a dynamic ice sheet that experienced substantial changes in ice flow. Subsequent work documented further evidence for large-scale changes in ice flow during the last deglaciation (McCabe et al., 1998), and detailed stratigraphic and geochronologic investigations documented widespread fluctuations of the IIS margin that corresponded to North Atlantic abrupt climate changes (Bowen et al., 1996, Bowen et al., 2002, McCabe and Clark, 1998, McCabe et al., 2005, McCabe et al., 2007b, Clark et al., 2009a, Clark et al., 2009b).

Despite the overwhelming evidence for significant fluctuations of the IIS during the last deglaciation, recent work based on undated flowlines has reverted to a model of monotonic thinning and retreat of the IIS in which the evidence for ice-margin readvances was either disputed or relegated to events of only local significance (Greenwood and Clark, 2009b; Clark et al., in press). Moreover, Greenwood and Clark (2009a) suggested that “there has been relatively little attention paid to deciphering the flow geometry of the ice sheet” (p. 3086), thus overlooking over 100 years of published research on just this topic (e.g., Charlesworth, 1928, Synge, 1969; McCabe et al., 1998, and references therein).

A reconstruction of slow incremental retreat of the IIS margin during a time of known large and abrupt climate changes would require that ice sheets are relatively insensitive to climate change. Insofar as this has broad implications to our understanding of past and future ice sheet–climate interactions, a detailed assessment of this claim is required. To address this issue, we provide a comprehensive summary of the geochronology of the IIS by incorporating all of the published age constraints derived from a number of sites that can be directly associated with stratigraphic evidence of ice-sheet margin fluctuations. In particular, this synthesis provides a rigorous test of whether previously documented fluctuations were synchronous (within dating uncertainties) around the ice margin, or isolated events that do not uniquely constrain a response by the IIS to the abrupt climate changes of the last deglaciation. In developing this synthesis, we note that prior to 1998, there were only two radiocarbon dates from Ireland that, other than limiting ages on post-glacial lakes and peat bogs, provided any information on the last deglaciation (McCabe et al., 1986). Since McCabe and Clark (1998) and McCabe et al. (1998) reported the first AMS ¹⁴C ages on monospecific foraminifera, nearly 120 new ages have been published that now provide a well-constrained geochronology for the IIS during the last glaciation, particularly since the Last Glacial Maximum (LGM). This synthesis reinforces and elaborates on previous work in demonstrating a dynamic IIS that responded to the abrupt climate changes in the North Atlantic region.