Beryllium-10 exposure ages of erratic boulders in southern Norway and implications for the history of the Fennoscandian Ice Sheet

The margins of the Fennoscandian Ice Sheet (FIS) during the Last Glacial Maximum (LGM) are fairly well constrained, yet the thickness of the FIS and its temporal evolution are poorly known. Here, we use beryllium-10 (Be-10) exposure ages of glacially transported boulders along three vertical transects in southern Norway to determine past ice-sheet elevations. The ice surface elevation during the LGM at our westernmost site, Skala, was likely at approximately 1440 in, coinciding with a previously mapped trimline. At our central site, Blaho, the ice surface elevation at the LGM was at least at similar to 1620m, likely closer to 1800m. At Elgahogna, our easternmost site, LGM ice was above the summit elevation of 1460m. Rapid deglaciation began at Skala and Blaho at approximately 13.5-14.5 (10)Beka and at approximately 12 Be-10 ka at Elgdhogna. By approximately 9 Be-10 ka, the FIS had thinned to below our lowest sample at each location. Periods of rapid thinning correlate with an exposure dated ice-margin history, as well as the glaciation curve for western Scandinavia. Our data, when combined with previous bedrock and blockfield (highly weathered bedrock on summits) exposure ages, resolve a long-standing debate over the age and geomorphic interpretation of blockfields and weathering zones in southern Norway. Frozen-bed, low-erosive ice clearly covered the summit blockfields at Blaho, and especially at Elgdhogna, where we find a large and progressive divergence between boulder and bedrock exposure ages with elevation. We conclude that the lower limit of blockfields in these areas should be interpreted as an englacial thermal boundary, rather than the upper limit of the LGM FIS, and that blockfields have survived multiple glacial cycles. We also compare our deglaciation record with several models of the FIS. Models that predict how the FIS changed during deglaciation, either through inversion of crustal rebound or through climate-dependent forward modeling, generally match our exposure ages reasonably well during deglaciation, but predict larger LGM surface elevations than our data indicate. On the other hand, the CLIMAP maximum model, which provides LGM elevations only, agrees with our results fairly well. (C) 2007 Elsevier Ltd. All rights reserved.