MOVING BEYOND THE AVERAGE AND BEYOND STEADY STATE IN COSMOGENIC NUCLIDE STUDIES OF LANDSCAPE EROSION AND SOIL PRODUCTION (Invited Presentation)

Cosmogenic nuclides have transformed understanding of surface processes in what has become a decades-long, global effort to quantify landscape erosion and soil production rates in mountain landscapes around the world. However, landscape erosion rates are commonly based on stream sediment containing a fraction of the sizes on the bed and thus assume that the narrow, sampled range of sizes is representative of the catchment average. Meanwhile, soil production rates are commonly based on samples of saprolite under the assumption of steady soil thickness and density over the timescale of nuclide buildup. Here we introduce methods for testing these assumptions and apply them at sites that are representative of the global database of landscape erosion and soil production rates. In our test of the assumption about stream sediment, we found that cosmogenic nuclides in sediment draining a steep catchment are inversely correlated with sediment size, reflecting climate-related spatial variability in the production of the different sizes, which we inferred from detrital thermochronological age distributions. This confirms that a single cosmogenic nuclide measurement from a narrow size range can be insufficient to quantify the spatially averaged erosion rate in steep catchments. Likewise, in our test of the steady state assumption, we found a mix of steady- and non-steady-state soils, confirming that non-steady conditions may be common in mountain landscapes. Together, these results challenge key assumptions of two of the foundational approaches to quantifying surface processes from cosmogenic nuclides. Yet these results also demonstrate how (1) a much richer picture of spatially variable sediment production and erosion can be obtained by measuring both cosmogenic nuclides and detrital thermochronological ages in each of the sizes present on the stream bed, and (2) cosmogenic nuclides from multiple depths can overcome the limitations of the steady-state assumption in soil production rate studies.