QUANTIFYING THE INFLUENCE OF BEDROCK LANDSLIDING ON DETRITAL THERMOCHRONOMETER DATA

Thermochronometers are a widely applied tool for quantifying exhumation rates over long time scales (>10⁵ y), but generally lack sensitivity to short-term events. Detrital thermochronometer samples from modern river sediment are a notable exception, as they depend on the timescale of both long-term exhumation processes and short-term surface processes (~1-100 y) that supply mineral grains to the river. In mountainous settings, detrital datasets frequently aim to characterize the exhumation history of the drainage area upstream of sample locations, assuming that the dated grains provide a reasonable representation of the age distribution within the full basin. However, in many orogens bedrock landsliding is a major source of river sediment and detrital samples may be biased towards grains derived from individual landslide events. To quantify the influence of landslides, we modified a 3-D thermokinematic and thermochronometer age prediction model to generate detrital age distributions in modern channels resulting from sampling landslide-generated sediment. Landslide distributions were predicted for 16 basins in the Marsyandi River valley in central Nepal that range in area from <1 to >200 km². For each basin, 10-500 grain-ages were randomly selected from the landslide sediment to create the landslide age distributions, which were compared to the age distribution generated by sampling the entire basin. We find that the landslide age distributions are strongly sensitive to the number of contributing grain-ages and the residence time of landslide sediment in the basin. For a given basin area, decreasing the number of grain-ages leads to more landslide age distributions that are statistically identical to the full-basin distribution. In addition, short (~0.25 y) and long (~100 y) residence times yield a greater number of landslide and full-basin age distributions that are equal. Lastly, landslide distributions from small basins tend not to differ from the full basin for short residence times, but can differ at residence times of ~100 y. Overall, this suggests that care should be taken in selecting basins for detrital sampling as landslides may provide a non-uniform sediment supply and lead to erroneous conclusions about basin exhumation histories.