Rates of late Quaternary normal faulting in central Tibet from U-series dating of pedogenic carbonate in displaced fluvial gravel deposits

摘要

Along the main boundary fault of the central Tibetan Shuang Hu graben, two well-preserved fluvial terrace surfaces are vertically offset by ～1.3 m (terrace I) and ～14.8 m (terrace II). Using thermal ionization mass spectrometry (TIMS), we have determined 48 ~(230)Th-~(234)U-~(238)U ages of small (10-40 mg) samples from pedogenic carbonate rinds on clasts in the terrace deposits. Based on textural, microstratigraphic and geochemical criteria, we conclude that the U-series ages of innermost-rind samples provide reliable ages of 16.4 ± 1.9 ka for terrace I, and 233.1 ± 9.3 ka for terrace II. This constrains the average rate of vertical displacement along the normal fault to be 0.079 ± 0.011 mm/yr during the past ～16 kyr, and 0.064 ± 0.007 mm/yr during the past ～233 kyr. Combining these results with slip-rate estimates for other normal faults along the graben margins indicates that the cumulative vertical displacement on all normal faults did not exceed ～0.3 mm/yr during the late Quaternary. This new rate from central Tibet is distinctly lower than the rate of 1.9 ± 0.6 mm/yr during the late Quaternary. This new rate from central Tibet is distinctly lower than the rate of 1.9 ± 0.6 mm/yr inferred for normal faults bounding the Yadong-Gulu graben system in southern Tibet, consistent with more pronounced and common graben development in southern Tibet. We therefore suggest that normal faulting in southern Tibet is largely controlled by local processes, and that the age of its initiation is not a valid proxy for uplift of the Tibetan plateau, as has been widely assumed. Our data also place chronologic constraints on the termination of periglacial conditions in central Tibet during the last three interglaciations. Rind formation appears to have been relatively continuous on clasts in terrace I from ～16 ka. Clast rinds from terrace II, in contrast, are characterized by three main episodes of rind growth that are widely separated in time. The earliest and least well-preserved episode started at 233.1 ± 9.3 ka, a second episode occurred at 131.8 ± 7.8, and a third episode was approximately coeval with rind formation in terrace I after 16.4 ± 1.9 ka. In each case, the age of onset of carbonate rind accumulation coincides with independently estimated ages of global deglaciation. Accordingly, we conclude that periglacial conditions in the central Tibetan plateau terminated synchronously with early stages of global warming following the last three glacial maxima, and that fluvial aggradation, terrace formation, and pedogenic carbonate accumulation in terrace soils ensued rapidly.