Pull-apart basin and push-up structure are the two most common and important structures formed within a strike-slip fault system. The term of pull-apart basin was firstly introduced when researchers discussing the formation of Central Death Valley, California. A pull-apart basin typically forms and develops between en echelon surficial strands or along a releasing bend of the transform fault. The diagonal cross-basin fault, formed diagonally within a pull-apart basin, connects the two en echelon strands bounding the basin. This type of fault not only plays an important role in the extinction of pull-part basin, but also controls the occurrence of large earthquakes. Therefore, it is of great significance to study the formation and evolution of cross-basin faults. However, compared with extensively studied pull-apart basins, fewer studies have been conducted on cross-basin fault, which greatly hampers our understanding of the structural evolution process of pull-apart basin and strike-slip fault. In this study, we take the Ganyanchi(Salt Lake)basin, the largest pull-apart basin located in the central part of the Haiyuan Fault, northeastern corner of the Tibetan plateau, as an example to investigate the character and general mechanism of the cross-basin fault. Geomorphological investigation, shallow artificial seismic exploration, and composite drilling geological survey are carried out along the cross-basin fault in Ganyanchi Basin. The main conclusions are listed as below: 1)The cross-basin fault in Ganyanchi Basin is a reverse strike-slip fault dipping to SW, rather than the previously claimed a strike-slip fault with significant normal component; 2)Although with a classic rhombic shape, the Ganyanchi Basin is actually an asymmetric pull-apart basin, which is mainly controlled by the northern boundary fault, i.e., the Nan-Xi Hua Shan Fault. Under the control of the Nan-Xi Hua Shan Fault, more than 680m thick growth strata were accumulated in the basin, and a rollover anticline was formed by the growth strata; 3)The example of the Ganyanchi asymmetric pull-apart basin suggests that a possible mechanism for the reverse cross-basin fault is probably the “straightening of strike-slip fault”, that is, the earlier formed antithetical normal fault adopts reverse component of straightened basin boundary fault, and then undergoes rotation and finally becomes a synthetical reverse fault with great strike-slip component. The rollover anticline, mainly controlled by the master boundary fault of an asymmetric pull-apart basin, may also partly contribute to the rotation of the cross-basin fault. As more natural pull-part basins needed to be investigated, caution should be taken when this suggested model is applied elsewhere.

The theories, techniques and methods for the exploration of active faults in the terrestrial domain are relatively mature, while such efforts in the water domain remain very few. In this study, the AAE shallow profiler was used to detect the underwater three-dimensional topography and active faults in the Qionghai area, Xichang for the first time. Based on the SKUA-GOCAD software platform and its DSI interpolation method, three-dimensional modeling of the exploration data was carried out. The survey profiles clearly reveal three different reflection interfaces, including the underwater interface, the interface between the silt layer and shallow sedimentary layer, and the bottom of the shallow sedimentary layer. The three-dimensional topography of the Qionghai area was mapped initially. Moreover, evidence of active faults was first found in several survey profiles from the reflection interface cutoff. This study also analyzed and discussed the working principle and characteristics of the AAE shallow profiler, including their parameters and various factors of exploration. The mapped three-dimensional topography and active faults in the Qionghai area of Xichang can provide a reference for research on the active tectonics underwater in the future.

Based on the differences of seismic activity,focal mechanism,geological structure,and hydrologic condition between reservoir induced seismicity(RIS)to natural earthquake,the basic features of reservoir induced earthquake from natural earthquake are studied statistically by the RIS database of 150 cases of reservoir induced earthquakes world-wide,and the results are shown as follows: ① Using statistical principle,we find that RIS occur in the areas near the reservoir banks in the impounding period,most of reservoir induced earthquake are distributed within a distance of 10km to the bank,with depth less than 10km,and a few events extend outwards as far as over 10km along fault zones,karst caves,and(thermal)springs; ②The frequency and magnitude of reservoir induced seismicity change with the reservoir water level. Most maximum magnitude earthquakes take place in the period before the first highest water level is reached,or in the following 2-3 impounding stages after the first highest water level,after then,a few would take place; ③The sequence of RIS is complete,usually as "foreshock-main shock-aftershock" or "foreshock-swarm-aftershock",most of them are microseisms of M_L1-3, so the b value is larger than natural earthquakes; ④ RIS differ from natural shocks in higher epicenter intensity,higher frequency of ground motion and higher peak ground acceleration,but faster attenuation with distance; ⑤The stress drop of RIS is smaller than that of natural earthquake of the same size,and the source size is larger; ⑥ The RIS occur in areas under lithological and geologic conditions of good permeability. Tensional fracture,tensional fault zone,unconsolidated fault plane,karst,and(thermal)springs are all good water penetration channels.

Using the observations of the Zipingpu reservoir seismic network(including seven short-period seismic stations,with an average station distance of 10km)of the period from July 2004 to 2009,and the data from January-December 2009 recorded by the intensive seismic observation network(six short-period seismic stations with the station spacing of 3km),930 foreshocks near Zipingpu reservoir,the main shock,and 5789aftershocks,as well as the initial rupture of the M_S 8.0 Wenchuan earthquake were relocated with double difference location algorithm.Combined with the geological field investigations,the analysis of seismic activity in Zipingpu reservoir area,the distribution of coseismic surface rupture,the depth of initial rupture location of Wenchuan M_S 8.0 earthquake,and the aftershock distribution,we find that the occurrence of M_S 8.0 Wenchuan earthquake has close relation with Zipingpu reservoir: 1) The seismic swarms at the Shuimo,and Dujiangyan and Shenxigou area are 10 kilometers away from the reservoir,the seismic strain release caused by Zipingpu reservoir increased by 200%,and the release was closely related to water level changes and accelerated before the Wenchuan earthquake; 2) There are two northeast directed coseismic surface rupture zones with 1m displacement along Zhongtanpu Fault; 3) the aftershocks of Wenchuan earthquakes are mainly distributed on northwestern wall of the Zhongtanpu Fault; 4) The relocation revealed that the Wenchuan earthquake occurred at 27minutes,59.5 seconds,the depth is between 6～9km; and 5) by comparison,the focal depths of the fore-shocks occurring on 2008-04-05 are about 7.8km,and the initial time of waveforms of this swarm is same as that of the M_S 8.0 Wenchuan earthquake.

815 aftershocks of Wenchuan earthquake recorded by 12 seismic stations in 2009 were relocated using the double difference algorithm to analyze the seismic activity of the Zipingpu reservoir.Relocation results show that the epicenters of aftershocks concentrate relatively in three regions.Focal depth distribution characters of different concentration regions take on obvious difference.This means earthquakes in different concentration regions may have different causes.Compared with relocations of earthquakes taking place before the Wenchuan earthquake done by other researchers,locations of the seismic concentration regions in the reservoir area changed noticeably after the Wenchuan earthquake.These variations are related with the stress adjustment of local part in the reservoir area and may also be related with the diffusion depth and range of increased pore pressure caused by rock failure in the course of Wenchuan earthquake.

136 earthquakes,taking place in the west of Xietan area,recorded by portable stations deployed in the Three Georges reservoir area were relocated using the double difference algorithm.The relocations show that the root-mean-square shifts of the relocations in the directions of E-W,N-S and U-D are 0.38km,0.33km and 0.98km,respectively.The earthquakes in clasolite area with the focal depths about 4～5km take on linear distributions from shallow to deep part.These earthquakes were deduced as reservoir induced earthquake of fault fracture type.In the contrast,the earthquakes in limestone pavement with the focal depths about 2～3km take on slightly divergent distribution and have the characters of reservoir induced earthquake of the karts collapse type.

In accordance with the requirements of "Eleventh Five-Year" national scientific and technological support projects,an intensive seismic observation network consisting of 26 seismic stations was established at the Three Gorges Reservoir area in the section of Hubei Province in March 2009(21 short-period seismographs,5 broadband digital seismographs).From March to December,2009,2995 M_L-0.8～2.9 earthquakes were observed during the trial impounding of the Three Gorges Reservoir(water level rose from 145m to 172.8m).Using the double difference earthquake location algorithm,2837 earthquakes were relocated precisely.The results show: The pattern of small local earthquake swarms in Three Gorges Reservoir area took on a linear distribution or mass-like cluster distribution,the mass-like clusters of events were generally within the distance of 5km from waterfront,and the linear distribution of the earthquakes could be extended to a distance of 16 kilometers away from waterfront.In Hubei section of Three Gorges Reservoir,earthquakes were mainly concentrated in the northern end of Xiannüfeng and Jiuwanxi Fault in the vicinity of the Xiangxihe river,and along the banks of the Yangtze river at the east of Xietan village and at Shenlongxi area on the northern bank in Badong region;the focal depths were less than 10km,and 4km in average.Earthquake frequency of the reservoir region had a positive correlation with reservoir water level fluctuation,indicating that the seismicity belonged to the reservoir induced earthquakes.Along the Shenlong river in the reservoir area,earthquakes showed three linear distributions in the northern Badong County,which was related to karst distribution.There are underground rivers along carbonate strata.When reservoir was impounded,water permeated into the underground rivers,thus inducing earthquakes.Earthquakes in the areas on the cross-river segment of Xiannüfeng Fault,the Jiuwanxi Fault and at the areas west of Xietan,Shazhen and Xizhen,may be related to the softening of discontinuities,such as the Nukou Fault,the Xiannüfeng Fault or the bedding joints,etc.,which would lead to failure of rock masses,thus,inducing earthquakes.However,convincing conclusions about the triggering mechanism still need further study.In addition,in the areas south of Wenhua and Yanglin of Zigui County and at Rangkou town east of Badong county,there are mining-induced earthquakes at the mines nearby.And on the shores of the reservoir,there are some collapse earthquakes.

Teleseismic receiver functions(RFs)migration method has been used to image the spatial variation of the converters in the crust and upper mantle along the north Hi-Climb broadband seismic array in the western Qinghai-Tibet Plateau.Results of RFs migration image show the crustal structure under Bangong-Nujiang Suture(BNS)takes on opposite underthrust form,i.e.,the Lhasa block exhibits northward overthrust and underthrust of upper crust and lower crust,respectively,the Qiangtang block southward overthrust and underthrust of upper crust and lower crust,respectively.The above underthrust form means Lasha and Qiangtang blocks may have complex assembly process.Combined with petrology achievements from previous researchers,terrain assembly model has been constructed based on the opposite underthrust form under BNS.Terrain assembly model illustrates that oceanized crust of the New-Tethys north ocean basin has been subducted southward beneath the Lasha block since the closure of the ocean basin and the continental collision of Lhasa and Qiangtang blocks.Since the continental collision of the Indian and Eurasian Plates at 60～50Ma ago,the lower crust of Lasha block however has been subducted northward and may have made the southward subduction plate delaminated,and then Indian Plate subducted beneath the Qiangtang block.

In the paper,according to the related factors and characteristics of identified reservoir induced earthquakes,the parameters of reservoir induced earthquake database and library structure have been determined.Based on the database functions of ARCGIG geographic information system software,the reservoir induced earthquake database contains 131 reservoir induced earthquake cases in the would, 110 Class-Ⅰlarge reservoirs(larger than one thousand million cubic meters),200 ClassⅡlarge reservoirs (larger than one hundred million cubic meters and less than one thousand million cubic meters) and 70 reservoirs with dams higher than 100 meters in China.It has the functions of quick query,statistics and icons.After the Wenchuan earthquake,aftershocks endangered many reservoir's dams and people who were taking part in earthquake rescue and relief operations.Two days after the M_s8.0 Wenchuan earthquake,the distribution and basic information of large reservoirs in Sichuan Province and nearby regions was submitted.The government made timely earthquake emergency response decisions to ensure a safe relief.The reservoir induced earthquake database is not only a method provided to the government for quick decision-making,but also can be used nation-wide.

In this paper,the theoretical foundation of classification of focal mechanism solutions is drawn by the theory of shear dislocation,where the mechanisms of earthquake can be divided into the strike-slip fault,reverse fault and normal fault.The 124 focal mechanisms from Harvard University have been classified by the triangle diagram method in the area of Sichuan and Yunnan Provinces in China.The focal mechanisms of strike-slip,reverse and normal faulting have been discussed.These are the three deformation types in this area.The results show a SE-directed block movement of the Sichuan-Yunnan region.Along the Xianshuihe Fault,the Anninhe Fault,the Zemuhe Fault and the Xiaojiang Fault,large strike-slip movement and deformation have occurred.On the Sagaing Fault and the Longmenshan Fault,the reverse faulting mechanism dominates.The quick movement of the Red River Fault towards southeast driven by the Xiaojiang Fault forms a large tensional stress area between the Jinshajiang and the Xiaojinhe Fault.This constitutes the latest pattern of recent crustal deformation of Sichuan-Yunnan region and its adjacency.

In the paper,the seism ic damage to reservoirs in Sichuan Province has been analyzed by field in-vestigation and data collection of the damages of the M_S8.0 W enchuan earthquake.The results show that the damage of the M_S8.0 W enchuan earthquake is 1 or 3 degrees higher than the seism ic intensi-ties provided by the reservoir seism ic safety assessments in Sichuan Province.This reveals certain questions behind the reservoir seism ic safety assessment in this region.Because of the short earth-quake record history,there has been no pred icting of earthquake larger than the largestmagnitude of historical earthquake,and thus,the risk of the Lomenshan active faultwas greatly underestimated,re-sulting that fissures,seepage and partial collapse occurred widely on the dams ofmany reservoirs.The paleoearthquake method is an importantmeans of d istinguishing active fault in reservoir seism ic risk assessment and improving the accuracy and reliability of the assessment results.