GRAIN-SIZE BIAS IN DETRITAL THERMOCHROMETRY: IMPLICATIONS FOR INTERPRETING SEDIMENT PROVENANCE AND LANDSCAPE EVOLUTION

Detrital thermochrometry is commonly used to study sediment provenance at the orogen scale and to quantify spatial variations in erosion rate at the catchment scale. Most applications assume that hillslopes shed sediment in proportion to their erosion rates, and that sampling stream sand yields a representative distribution of bedrock ages from the catchment. However, if sediment size varies across hillslopes, any one sediment size (e.g., sand) may be unrepresentative of the catchment as a whole. In previous work, we showed that this grain-size bias can lead to substantial errors in cosmogenic nuclide studies of catchment erosion (JGR Earth Surface, 121, 978-999). Here we explore its effects on detrital thermochrometry.

To evaluate the influence of grain-size bias on provenance studies, we used a forward model of catchment erosion to determine the probability of detecting an age population that occurs in a small fraction of the catchment. This simulates the challenge of fully characterizing age distributions of tectonically complex terrain. We found that stream sand underrepresents higher elevations when sediment size increases steeply with elevation. As a result, the chance of detecting an age population originating from the highest 20% of the catchment can be as low as 50%, depending on the number of grains analyzed.

To evaluate the influence of grain-size bias on quantifying spatial patterns in erosion rates, we considered catchments in which bedrock age increases with elevation. Our results show that higher-elevation sources are underrepresented in smaller grain sizes, such that the age distribution in sand is skewed toward younger ages from lower elevations. We explored this effect over a range in catchment relief and area, and found that age distributions from low-relief (<1.5 km) catchments are not sensitive to changes in sediment size. However, in large, high-relief catchments, age distributions in sand do not reflect spatial variations in erosion rates.

Both scenarios considered here show that changes in sediment size across landscapes can influence interpretation of detrital thermochronology in large, high-relief catchments. We show that limitations imposed by grain-size bias can be mitigated via analysis of a sufficient number of grains across a range of sediment sizes.