GLACIATION HISTORY OF THE PAMIR AND ADJACENT MOUNTAINS RECONSTRUCTED BY BE-10 SURFACE EXPOSURE DATING OF ERRATIC BOULDERS

In an attempt to reconstruct the late Quaternary glaciation history of Central Asia, we have sampled erratic boulders from different moraine stages in the Southern Alichur Range and in the areas of Lake Yashilkul and Ailuitek Pass, Pamir, Tadshikistan, as well as in the the Alay Range, Kyrgystan, for surface exposure dating using in-situ cosmogenic Be-10. Our results show that remnants of an extensive plateau glaciation in the southeast of the Pamir date to before the MIS 4 and probably are of middle Pleistocene age. During the late Pleistocene, glaciers in the Pamir were restricted to the valleys and left distinct high lateral as well as arcuate terminal moraines, which in the Pamir frequently show a hummocky relief, locally called "tschukur". Exposure ages from the more extensive of these "tschukur" moraines show a wide scatter ranging from the end of MIS 4 to the end of MIS 2, showing that they have been actively degrading, especially by melting out and sublimation of buried ice and permafrost, for all this time. A MIS 3-4 moraine in the Koksu valley of the Alay Range, which is situated below the permafrost limit, shows no such scatter in boulder ages and no "tschukur" relief. At Lake Yashilkul, a single lateral moraine outside the "tschukur" moraine dates to the beginning of MIS 4, when glaciation in this area was at its late Pleistocene maximum. Behind the "tschukur" lobes in the Pamir, there frequently lies a succession of recessional moraines, which lead back to a more pronounced, but less extensive stage dating to the MIS 2. During the MIS 2, glaciation in all studied valleys was less extensive than during MIS 3-4, extending at most a few km from the cirque headwalls. Older moraines frequently show clear symptoms of degradation and relief activity. This is reflected in the distribution of the boulder surface ages. This distribution in most cases is not random, but ages younger than the deposition age tend to cluster around times of younger glacier advances in the same valley. This supports the idea that times of glacier advances are also times of relief instability in the surrounding periglacial area.