LANDSLIDE DAMMED FLUVIAL SEDIMENTS AT THE ANNING-DADU DIVIDE, EAST MARGIN OF THE TIBET PLATEAU

The eastern Tibetan Plateau margin are etched by giant parallel drainages following north-south courses. Among those the Dadu River, a major tributary to the Yangtze River, makes an abrupt (90°) change flows eastward through a narrow deep gorge into the Sichuan basin. In contrast, the Anning River, also a tributary to the Yangtze River, flows in a broad, alluviated valley with gentle river gradients close to its headwaters, which is 30 km south to the abrupt bend of the Dadu River. This scenario leads to hypothesis that the originally south flowing Dadu/Anning River has been captured by an east flowing river that became the lower Dadu River. The fluvial sediments preserved at the pass that separates the Dadu River from the headwaters of the north-south oriented Anning River was suggested to represent the wind gap.

If the fluvial sediments at the pass were transported by the Dadu River, we should expect Miocene zircons from the Gongga Shan granite, which is 50 km to the north, be found within the sediment. However, U-Pb detrital zircon analysis on the sand lens samples contains no Cenozoic component. In addition, along the present Nanya River that cut in the fluvial sequence, the sediment is fining very fast down-stream. It changes from gravel and cross-bedding sand dominated sequence to horizontal bedding greyish silt or peat with abundant plant debris in a distance of 2500 m. This fact indicates prominent decrease of fluvial dynamics northward. We suggest that the decrease in fluvial energy is likely due to damming in the river valley.

From satellite image and field investigation, a remanent landslide dam has been identified downstream. The pulverized landslide body may be derived from the steep slope on the right bank of the Nanya River. The landslide may be resulted from an earthquake at the Xianshuihe-Anninghe Fault, which has been highly active.

The fluvial sediments preserved were described as Pliocene in geological map. However, the age is constrained by neither paleontological nor magnetostratigraphic methods and therefore can only be tentatively assigned. We applied AMS C-14 dating on two wood samples collected from the sediment, both of them are beyond the upper limit of this method (>45000 a). The very well preservation of wood may imply the sequence younger than Pliocene. Cosmogenic burial dating and OSL dating are in progress.