Cosmogenic 36Cl surface exposure ages obtained for multiple boulders from single landforms are usually characterized by a variance larger than that of the analytical methods employed. This excessive boulder-to-boulder variability, progressively more profound with increasing age of landforms, is due to removal of soil and gradual exposure of boulders at the surface. In our gradual exposure model, boulders are initially buried in moraine matrix. With time, erosion lowers the moraine surface and the boulders are gradually exposed to cosmic rays. Because the cosmic ray intensity changes with depth, the boulders are subjected to variable production rates of the cosmogenic 36Cl. Initial depth of boulders and their chemical composition are variable, which results in different amounts of the accumulated cosmogenic 36Cl and thus different apparent ages of boulders. The shape of the resulting distribution of the apparent ages and the coefficient of variation depend on the erosion depth, while the first moment is a function of the true surface age and the erosion depth. These properties of the apparent age distributions permit calculation of the surface age, the erosion depth, and also the average erosion rate. We tested the model calculations using 26 boulders from a late Pleistocene moraine at Bishop Creek, Sierra Nevada, California. The set exhibited a bimodal distribution of the 36Cl surface exposure ages.

We interpreted the older mode as the result of gradual exposure and the younger one as the result of surficial processes other than soil removal. The 10 samples that constitute the older mode produced a distribution which closely matches the modeled distribution calculated using an age of 85 kyr and erosion depth of 570 g cm-2. This age is the same as an independent estimate obtained from cation ratio studies, and the calculated erosion depth is very close to the erosion depth of 600 g cm-2 based on a simple analytical model of soil erosion. These results indicate that our statistical model adequately describes effects of soil erosion on accumulation of cosmogenic 36Cl. The approach can be used to simultaneously obtained the true landform age and the erosion rate from apparent 36Cl ages and therefore may help in evaluation of surface exposure ages of eroding landforms. ¿ American Geophysical Union 1994