The Xianshuihe (XSH) fault in eastern Tibet is one of the most active faults in China, with the next large earthquake most likely to occur along its SE part, where the fault splits into three parallel branches: Yalahe, Selaha and Zheduotang. Precisely quantifying their slip rates at various timescales is essential to evaluate regional earthquake hazard. Here, we expand our previous work on the Selaha fault, to the nearby Zheduotang and Moxi faults, and add observations on the Yalahe fault and on the newly discovered Mugecuo South fault zone. Using tectonic-geomorphology approaches with 10Be dating, we had determined average late Quaternary slip rates of 9.75±0.15 and 4.4±0.5 mm/yr along the NW and SE Selaha fault, respectively. Using the same methods here, we determine a slip rate of 3.7-5.4 mm/yr on the Zheduotang fault and of 10.4-14.8 mm/yr on the Moxi fault. This is consistent with the southeastward slip rate increase we had proposed along the XSH fault system from 6-8 mm/yr (Ganzi fault) to ~10 mm/yr (Selaha fault), and >10.4 mm/yr (Moxi fault). We eventually propose a new model for the SE Xianshuihe fault, where the large-scale Mugecuo pull-apart basin lies within an even larger scale compressive uplift zone in the XSH fault’s restraining bend, where the highest peak in eastern Tibet is located (Gongga Shan, 7556 m). Our slip rates determination allows to estimate a relatively high regional earthquake hazard of Mw~7 at present in the SE Xianshuihe fault.

The pattern and timing of deformation in the southeast Tibet resulting from the India-Asian collision remain poorly constrained. Detailed field mapping, structural analysis and geo-thermochronogic data within a newly-defined Ludian-Zhonghejiang fold-thrust belt stretching over 120 km between the Diancang Shan and Xuelong Shan metamorphic belt in western Yunnan, China document Early Cenozoic tectonic evolution of the conjunction area between the Lanping-Simao and South China blocks. The study area is cut by two major northwest-striking, southwest-dipping brittle faults, named Ludian-Zhonghejiang fault and Tongdian fault from east to west. Kinematic measurements and indicators of S-C fabrics and striations, as well as juxtaposition of Triassic meta-sedimentary rocks overlying on Paleocene strata indicate thrusting along the Ludian-Zhonghejiang fault. Similarly, structural analysis show the Tongdian fault is also a reverse fault. Between these structures, fault-bounded Permo-Triassic and Paleocene strata are strongly deformed by upright southwest-vergent folds with axes that trend nearly parallel to the traces of the principal faults, consistent with reserve faulting related to regional NE-oriented compression. Zircon and apatite (U-Th)/He and apatite fission track ages from a Triassic granitic pluton in the hanging wall of the Ludian-Zhonghejiang thrust assisted by inverse modeling reveal a period of accelerated cooling from 50 Ma to 37 Ma, which is interpreted to record the lifespan of the fold-thrust system collaborated by the intrusive relations between magmas of ~35 Ma dated by zircon U-Pb and the fold and thrust belt. Since 37 Ma, decreasing cooling rates implies cessation of the thrusting. Early Cenozoic activity of the deformation system likely controlled deposition of the Jianchuan-Liming basin evident by coeval sediments derived from the proximal hanging wall of the fold-thrust belt. These results, together with tectonic records of contraction in east Tibet, suggest crustal shortening related to the India-Asian collision and convergence prevailed the southern and eastern part of the Tibetan Plateau, which predated Oligo-Miocene onset of extrusion tectonics in southeast Tibet.

Understanding the development of key thrust faults in southeastern Tibet is significant to reconstructing the geodynamic and topographic processes. Detailed structure analysis along the ~400 km long Jinhe-Qinghe thrust belt (JQTB) indicate thrust motion with a minor left-lateral component. The exhumation history of the Baishagou granite, based on apatite (U-Th)/He and fission-track thermochronology and thermal modeling, suggest an accelerated exhumation rate (~0.42 km/Myr) between 20 and 15 Ma. We interpret that fast exhumation due to the activation of the Nibi thrust, a northern branch of the JQTB. The ~1.5-2.2 km of exhumation that occurred corresponds to the present topographic difference across the thrust belt. In the Early Miocene, significant relief along JQTB was generated by thrusting. When compared with previous studies it appears that Cenozoic exhumation and relief creation in southeastern Tibet cannot be explained by a single mechanism. Rather, at least three stages of relief creation should be invoked. The first phase is an Eocene NE-SW compression partly coeval with Eocene sedimentation. During the Late Oligocene to Early Miocene the second thrusting phase occurred along the Yulong and Longmenshan thrust belts, and then migrated to the JQTB further to the southeast during 20-15 Ma. A third phase involved the activation of the Xianshuihe fault and the re-activation of the Longmenshan thrust belts and the Muli thrust. The interaction between thrusting and fast river erosion triggered by climate change is not certain but thrusting along thrust belts appears to explain most of the present-day relief in the southeastern Tibetan Plateau.