Glacial erosion rates of primary bedrock from in situ 14C-10Be measurements are low

Glacial erosion shapes alpine landscapes, produces chemically reactive mineral surfaces integral to the carbon cycle, and informs glacier dynamics applied in ice sheet models. Quantifying primary bedrock erosion has remained elusive due to the inaccessibility of the ice-bed interface. Many erosion estimates thereby rely on basin-wide sediment accumulation rates that can include reworked sediment, potentially causing overestimates of glacial erosion. Here, we quantify glacial erosion of primary bedrock using 60 paired cosmogenic in situ ¹⁴C-¹⁰Be measurements from new and published bedrock samples spread across 10 glacier forefields from 60° N to 16° S. We apply a Monte Carlo forward model that tests millions of scenarios of glacier exposure, burial, and erosion to identify scenarios capable of replicating the measured nuclide concentrations. Our new data are from a glacier in southeast Alaska where samples were collected at two scales: landform-scale along a single roche moutonnée to investigate abrasion versus plucking and valley-scale from the modern glacier terminus to its pre-industrial moraine to constrain glacier length fluctuations. The other 9 sites are across-valley transects abutting the terminus of the modern glacier. We compare our results to modeled erosion rates from a power-based abrasion law and Elmer/Ice glacier model simulations. The cosmogenic nuclide-based erosion rates are consistent across scales and sites, overlapping with the modeled erosion rates that are concentrated below 0.3 mm yr^-1. These findings suggest glacial erosion rates of primary bedrock are much lower than predicted from modern sediment supply studies that reach up to 10 mm yr^-1. Our millennial-scale glacial erosion estimates of crystalline bedrock support a modern bias in erosion estimates (e.g. Ganti et al., 2016) with implications for landscape evolution and sediment delivery models.