An analysis of all published boulder exposure ages from eroding moraines reveals considerable age variation; I calculated the normalized variation by dividing the difference between the youngest and oldest age in any given range by the oldest age within each temporally uniform unit (e.g. moraine). The mean variation of all series is 43%. This substantial variation conflicts with the conventional assumption that surface boulder ages simply equate to the age of the landform. I suggest that the observed large boulder age distribution is a natural phenomenon, caused by erosion and subsequent exhumation of fresh boulders to the moraine surface. To study how moraine surface lowering is reflected in cosmogenic boulder exposure ages I run a surface degradation model where the moraine flattens and widens with time as fine-grained sediment is transferred down slope from the moraine crest, exposing previously buried boulders there. The Figure shows a typical model result; the exposure age distribution for 500 boulders from a modeled 100 kyrs old moraine surface that has eroded from initial 60 to 29 meters height at the crest. If no erosion took place all 500 boulders would yield the same age (100 ka).

The published boulder age distributions are used to constrain the surface degradation model by finding the least difference between the modeled and observed age distributions. The amount of surface erosion is solved by making appropriate assumptions about the moraine initial form and age of the moraine. The fit between modeled and observed age distributions and the implied erosion rate will be discussed at the meeting.