Constraining decay of the British-Irish Ice Sheet on Lundy, Celtic Sea, during the Last Glacial Cycle

Much has improved about the glacial history of the British-Irish Ice Sheet (BIIS) during the Last Glacial Cycle (MIS 5d) thanks to rich data sets generated by the BRITICE Project.  However, the south-west sector, between the Celtic Sea and Bristol Channel, is challenging because it is largely secured by marine evidence. However, a glacial landscape, preserved on the granitic island of Lundy, exhibits clear evidence of glacial over-printing with remnant transported glacial boulders, glacially-eroded bedrock surfaces, ice scoured tor stumps and erratic gravels. Lundy marks the intersection of ice flow across the eastern boundary of the Celtic Sea and southern extent of the Welsh Ice Cap at the Bristol Channel.  The Celtic Sea transported one of the largest BIIS ice streams and thus has significant implications for understanding rapid deglaciation of large  ice-sheets. Together with the Isles of Scilly, further south by ~125 km, both locations are strategically placed to pin down the southern extent of the history of the BIIS  and answer the question – when during the Last Glacial Cycle did the BIIS overtop Lundy and if so,  did it extend up through the Bristol Channel during the global LGM 27-23 ka yrs ago.  Previous exposure ages ranging from 30-50 kyrs, from Lundy (Rolfe, 2012), notably on bedrock, point to a pre-MIS 3 glaciation (most likely MIS 4) ruling out the conclusion that LGM ice reached Lundy. Given that there is strong evidence (OSL and 10Be dating) for Scilly to have been glaciated during the LGM (Smedley et al 2017), which is also supported by BIIS modelling, debate surrounds assigning Lundy 10Be data to true exposure ages (Carr, 2017). The relatively large age spread possibly resulting from cosmogenic inheritance/erosional irregularities and pegmatite/beryl presence in Lundy granite (Mclintock, 1912) may complicate the interpretation of the exposure ages. 

We have re-visited Lundy and collected 11 new  samples comprising erratic boulders perched on polished bedrock, tor stumps and tops. The tor stumps (tops) would have been the least (most) persistent to preservation of inherited nuclides resetting . The new10Be exposure ages when compared to bedrock ages from Rolfe (2012) should confirm whether the timing of BIIS retreat at Lundy was before or during the LGM. Two OSL samples from cover sands over gravel will provide independent age control.

An interesting aspect of Lundy granite is the presence of beryl, topaz and other insoluble minerals (ie tourmaline). This required considerable care to quantify the intrinsic 9Be concentration. For example, in 3 quartz samples there was sufficient native 9Be to warrant zero addition of 9Be carrier.  Not including a native 9Be contribution would underestimate exposure ages. However, the leaching of meteoric 10Be from these insoluble minerals during quartz dissolution is more of a concern and would result in over estimating exposure ages (Corbett, 2023). The new ages and impact of Lundy granite chemistry for cosmogenic dating will be represented.

Rolfe, QSR, v43, 2012
Carr,  Proc. Geol. Assoc., v128, 2017
Mclintock, Mineral. Magazine, v16, 1912
Smedley, JQS, v32, 2017
Corbett, QG, v73, 2023