Cosmogenic 10Be and 26Al exposure ages of tors and erratics, Cairngorm Mountains, Scotland: Timescales for the development of a classic landscape of selective linear glacial erosion

The occurrence of tors within glaciated regions has been widely cited as evidence for the preservation of relic pre-Quaternary landscapes beneath protective covers of non-erosive dry-based ice. Here, we test for the preservation of pre-Quaternary landscapes with cosmogenic surface exposure dating of tors. Numerous granite tors are present on summit plateaus in the Cairngorm Mountains of Scotland where they were covered by local ice caps many times during the Pleistocene. Cosmogenic ¹⁰Be and ²⁶Al data together with geomorphic relationships reveal that these landforms are more dynamic and younger than previously suspected. Many Cairngorm tors have been bulldozed and toppled along horizontal joints by ice motion, leaving event surfaces on tor remnants and erratics that can be dated with cosmogenic nuclides. As the surfaces have been subject to episodic burial by ice, an exposure model based upon ice and marine sediment core proxies for local glacial cover is necessary to interpret the cosmogenic nuclide data. Exposure ages and weathering characteristics of tors are closely correlated. Glacially modified tors and boulder erratics with slightly weathered surfaces have ¹⁰Be exposure ages of about 15 to 43 ka. Nuclide inheritance is present in many of these surfaces. Correction for inheritance indicates that the eastern Cairngorms were deglaciated at 15.6 ± 0.9 ka. Glacially modified tors with moderate to advanced weathering features have ¹⁰Be exposure ages of 19 to 92 ka. These surfaces were only slightly modified during the last glacial cycle and gained much of their exposure during the interstadial of marine Oxygen Isotope Stage 5 or earlier. Tors lacking evidence of glacial modification and exhibiting advanced weathering have ¹⁰Be exposure ages between 52 and 297 ka. Nuclide concentrations in these surfaces are probably controlled by bedrock erosion rates instead of discrete glacial events. Maximum erosion rates estimated from ¹⁰Be range from 2.8 to 12.0 mm/ka, with an error weighted mean of 4.1 ± 0.2 mm/ka. Three of these surfaces yield model exposure-plus-burial ages of 295 _{- 71} ^{+ 84} , 520 _{- 141} ^{+ 178} , and 626 _{- 85} ^{+ 102} ka. A vertical cosmogenic nuclide profile across the oldest sampled tor indicates a long-term emergence rate of 31 ± 2 mm/ka. These findings show that dry-based ice caps are capable of substantially eroding tors by entraining blocks previously detached by weathering processes. Bedrock surfaces and erratic boulders in such settings are likely to have nuclide inheritance and may yield erroneous (too old) exposure ages. While many Cairngorm tors have survived multiple glacial cycles, rates of regolith stripping and bedrock erosion are too high to permit the widespread preservation of pre-Quaternary rock surfaces.