Methodology:

Samples from erratic cobbles and boulders perched on presently exposed mountain peaks next to Larter and Lucchitta glaciers were collected by Dr Joanne S Johnson (British Antarctic Survey) during the 2019-20 austral summer. Latitude, altitude, and elevation measurements of the location of the samples were made using a Javad Triumph II GPS receiver. The sample heights above ellipsoid were corrected to orthometric heights (height above geoid EGM08 in metres above sea level) using Precise Point Positioning software. Laboratory chemistry was conducted in the CosmIC Laboratory at Imperial College London, UK. Quartz purification followed the methods of Kohl and Nishiizumi (1992). Isotope dilution chemistry and beryllium purification used the methods of Corbett et al. (2016). 10Be/9Be ratios were measured by accelerator mass spectrometry at ANSTO using the methods of Wilcken et al. (2017; 2022) and normalised to the standard values of Nishiizumi et al. (2007). Laboratory data were then converted to 10Be concentrations and, finally, exposure ages, using an online exposure age calculator (Balco et al., 2008). Exposure ages were calculated using the default global production rate calibration dataset and two different cosmogenic nuclide scaling methods known as St (Stone, 2000) and LSDn (Lifton et al., 2014).

Data quality:

Quartz purity was tested using inductively coupled plasma optical emission spectrometry (ICP-OES). For isotope dilution chemistry, one procedural blank (carrier-only) was processed for every 11 samples, and each batch of samples is assigned a batch number. The procedural blank used to correct the 10Be concentration of each sample is dependent on the time period during which it was processed in the CosmIC Laboratory (processed_blank_data.csv). Beryllium was extracted from samples in batch one at the same time as the samples presented in Balco et al. (2023), Nichols et al. (2023), and Marschalek et al. (2024). Therefore, the blank value used to correct the 10Be concentrations in batch one is the same as that used to correct 10Be concentrations in the aforementioned studies (the mean and standard deviation of 12 batches of samples, 10Be/9Be = 1.73 x 10-15 ± 4.82 x 10-16; 30287 ± 8456 10Be atoms; processed_blank_data.csv). Samples in batch two were processed in the year following those in batch one. Therefore, samples in batch two are corrected with a different blank value (the mean and standard deviation of two process blanks, BLK300822 and BLK050922, 10Be/9Be = 2.71 x10-15 ± 3.61 x 10-16, 47410 ± 6255 10Be atoms; processed_blank_data.csv). Samples in batches three, four, and five were processed later in the same year as those in batch two. A batch-specific blank value is therefore used to correct the 10Be concentrations of samples in batches three (BLK120922, 10Be/9Be = 2.13 x10-15 ± 3.41 x 10-16, 37088 ± 5939 10Be atoms), four (BLK260922, 10Be/9Be = 1.86 x10-15 ± 3.88 x 10-16, 32550 ± 6786 10Be atoms), and five (BLK041022, 10Be/9Be = 4.50 x 10-15 ± 5.31 x 10-16, 78589 ± 9262 10Be atoms; processed_blank_data.csv). In all cases, uncertainties in the sample and blank values are propagated in quadrature.