Nature And Timing Of Large Landslides Within An Active Orogeny, NE Pamir, China

Large-scale landslides/rock avalanches exceeding>10⁶m³in volume are significant events thatresult in erosional degradation of mountains, change orogenic mass balance, and imperil humanpopulations and their infrastructure. Large landslides may be very important in landscapedevelopment in high mountains, yet there are still few studies that define their magnitude, frequency,timing and causal factors.Determining distribution of landslides through time is the key to assessing hazards and forunderstanding trigger mechanisms, such as climatic change, phases of enhanced earthquake activity,and/or post-glacial stress relaxation. Although there are many traditional dating methods forlandslide and these methods play an important role in dating landslides, for exampleDendrochronology, Radiocarbon dating, Lichenometry and OSL etc, they maybe have some defectswhen they are used in special geologic settings. Terrestrial cosmogenic nuclides （TCN） dating hasdeveloped well in last few decades. At present, a few foreign geologists used this TCN dating forlandslides. However, few geologists in China studied landslides using this method. Moreover, fewstudies have been undertaken on large landslides in the high mountains of Central Asia because ofthe inaccessibility of the region and associated problems of mapping and sampling for numericaldating. To help assess their importance in high mountains, six large landslides are identified inGoogle Earth imageries and in the field, and four of them （Bulunkou, Muztagh, Taheman andYimake） were mapped and dated using¹⁰Be terrestrial cosmogenic nuclides （TCNs） in the NEChinese Pamir at the westernmost end or the Himalayan-Tibetan orogen which is influenced by theIndian summer monsoon and mid-latitude westerlies.The work contents completed in this thesis are as follows:1) The six large landslides were mapped in the field using high resolution Google Earth imagery.I estimated their volume using3arc-second （⁹0m） Shuttle Radar Topography Mission（SRTM） digital elevation models （DEMs）.2) Large quartz-rich boulders on surfaces of Bulunkou, Muztagh and Taheman landslide deposits,and disintegrated quartz boulders and quartz clasts on the Yimake landslide deposit weresampled for¹⁰Be surface exposure dating. All samples were processed in TCN laboratory atUniversity of Cincinnati. The resultant BeO was mixed with Nb powder and loaded in steel targets for the measurement of the¹⁰Be/9Be ratios by the accelerator mass spectrometry atPRIME Laboratory in Purdue University. Caculating TCN ages were calculated using dataprovided by PRIME Laboratory and explaining their results.3) Twenty-five published¹⁰Be ages of landslides in the Himalayan-Tibetan orogen wererecalculated based on the model of Lal（1991）/Stone（2000）.4) Using the data for the large landslides, I estimated the amount of debris involved in largelandsliding since¹4.3±0.8ka and erosion rate due to large landslides in the study area.The conclusions of this work are described below:1) Six large lanslides （volume>10⁶m³） were identified in Pamir active orogeny. This thesisdescribed their characteristics and estimated their volumes.2) The ages for Bulunkou, Muztagh, Taheman and Yimake landslides are2.0±0.1ka,14.3±0.8ka,6.8±0.2ka and7.1±0.6ka, respectively. All ages for each landslide are consistent, andsuggest that TCN dating is an effective method for landslide dating and the effect of variant ofnuclides production rate with time is less important. Moreover, erosion rate is small and haslittle effect to the four landslides.3) Comparison of the mean slip surface dips with typical angles of internal friction of theBulunkou, Yimake, Taheman, Aerpa Aigezi and Bile Jiyi landslides show that they could havebeen triggered by increased pore water pressure, seismic shaking, or some combination of thetwo.4) The timing of many other large landslides in the Himalayan-Tibetan orogen at²-3.8ka,⁴.9-16ka and³2-39ka also suggests that during the periods of increased precipitation largelandsliding was prevalent. No landslides dated to²0ka when the intense monsoon decreased.This evidence suggests that landsliding intensity is directly proportional to the intensity ofmonsoon activity.5) Geologic factors may contribute to triggering of landslides in this region. Bedrock in thelandslide source area is fractured metamorphic rock with joints and shists. Footwallexhumation has led to steeply dipping foliation and the formation of triangular facets along therange. In addition, glaciation and periglacial weathering steepened and fractured slopes.Landslides could easily occur in this kind of setting.6) This thesis argues that there are strong climatic controls on the development of large landslides in the Himalayan-Tibetan orogen, but also recognizes the importance of earthquakes as amechanism to help initiate landslides in these tectonically active regions because Bulunkou,Muztagh and Taheman landslides were offset by Holocene faults. In addition, gully scales,rock varnish thickness, offset scarp and dating results suggest that the Taheman landslide wasnot trigged by the1895Tashkorgan earthquake.7) The total volume of debris moved by large landslides since the Late Glacial is²168millionm³. This equals to⁷8mm （2168million m³/27,700km2） of surface lowering of thelandscape by large landslides since the Late Glacial, or equivalent to a lowering of thelandscape of0.005mm/a over that time.