Constraints on the late Quaternary glaciations in Tibet from cosmogenic exposure ages of moraine surfaces

This contribution provides new constraints on the timing of Tibetan glacial recessions recorded by the abandonment of moraines. We present cosmogenic radionuclide Be-10 inventories at 17 sites in southern and western Tibet (32 crests, 249 samples) and infer the range of permissible emplacement ages based on these analyses. Individual large embedded rock and boulder samples were collected from the crests of moraine surfaces and analyzed for Be-10 abundance. We consider two scenarios to interpret the age of glacial recession leading to the moraine surface formation from these sample exposure ages: 1) Erosion of the moraine surface is insignificant and so the emplacement age of the moraines is reflected by the mean sample age; and 2) Erosion progressively exposes large boulders with little prior exposure, and so the oldest sample age records the minimum moraine emplacement age. We found that depending on the scenario chosen, the moraine emplacement age can vary by > 50% for similar to 100 ka-old samples. We consider two scaling models for estimating the production rates of 1 Be in Tibet, which has an important, although lesser, effect on inferred moraine ages. While the data presented herein effectively increase the database of sample exposure ages from Tibet by similar to 20%, we find that uncertainties related to the interpretation of the Be-10 abundance within individual samples in terms of moraine emplacement ages are sufficient to accommodate either a view in which glacial advances are associated with temperature minima or precipitation maxima that are recorded by independent paleoclimate proxies. A reanalysis of published data from moraines throughout Tibet shows that the variation we observe is not unique to our dataset but rather is a robust feature of the Tibetan moraine age database. Thus, when viewed in a similar way with other samples collected from this area, uncertainties within moraine exposure ages obscure attribution of Tibetan glacial advances to temperature minima or precipitation maxima. Our work suggests that more reliable chronologies of Tibetan glaciations will come from improvements in production rate models for this portion of the world, as well as a better understanding of the processes that form and modify these geomorphic surfaces. (C) 2010 Elsevier Ltd. All rights reserved.