Simulation of snow shielding corrections for cosmogenic nuclide surface exposure studies

Snow cover reduces cosmogenic nuclide production rates in bedrock. Corrections for snow cover can be more than 10% in mountainous, mid-latitude regions where many glacial chronologies have been constructed using cosmogenic nuclide surface dating of landforms. Most published snow corrections use historic climate data of limited duration that are not likely to reflect adequately the full range of snow conditions over the time of exposure. We present a model for describing the impact of snow burial on long-term exposure histories of landforms. The model applies an energy balance approach to snowpack evolution and incorporates both historic and long-term climate proxy data. Attenuation of cosmogenic fast neutrons is modeled alternatively as a simple exponential decrease with increased shielding or as a thin surface layer with constant production followed by an exponential decrease with increasing depth. The choice of attenuation model has little effect on the modeled results for the Cairngorms but will have a more significant effect in regions characterized by thinner, less dense snowpacks. Spatial variability in snow cover is modeled as a function of elevation only, ignoring local variability in snow accumulation as a result of slope aspect, wind redistribution and local topography. Thus, model results reveal general spatial and temporal trends in snow shielding effects, rather than site-specific corrections. Applications to data from the Cairngorm Mountains of Scotland show that the constant-plus-exponential (CPE) production rate-depth profile reduces but does not eliminate snow-shielding effects. Under present-day conditions, snow at 900 m in the Cairngorm Mountains reduces average production rates by 6% using the CPE profile and 9% with the exponential profile (EP). Long-term climate simulations from 15.5 ka through today produce larger snow shielding effects, mainly because they predict an increased proportion of precipitation as snowfall during the Younger Dryas. At 900 m, this long-term simulation reduces average cosmogenic isotope production rates by 12% (CPE) and 14% (EP). These results indicate that snow-shielding corrections based on historic climate records may be a potential source of systematic error in midlatitude mountainous regions.