The neotectonic activity is intense in the Taiwan Straits and the coastal area of South China. This region is one of the earthquake-prone areas of the world. In history, earthquakes of magnitude 6-7 occurred repeatedly in this region with a high recurrence rate. Therefore, this area has always been the focus of seismicity research and coastal earthquake prevention and disaster reduction. The exploration of active faults is the basis for seismic zoning, but the detection and identification of active faults in sea area are more difficult because of the coverage of sea water, which leads to a large number of “blind areas” in marine fault exploration for a long time. Seismic exploration methods are economical, suitable and efficient in detecting active faults in the sea area. This study compares the detection effect of different seismic sources.
In this study, geophysical exploration of active faults was carried out in the southeast Fujian uplift zone in the Taiwan Straits. A mini-multichannel seismic profile of GI gun source and sparker source at the same location was selected for comparative analysis and illustration. Five reflection interfaces(T₁—T₄, Tg)were interpreted on the GI gun profile, and five sets of seismic sequences(A—E)were classified. Six reflection interfaces(T'₁, T₁—T₄, Tg)were interpreted on the sparker source profile, and six sets of seismic sequences(A—D and E₁—E₂)were classified. Three basement faults and two shallow faults with small vertical extension were found, which are active since the late Pleistocene. Among them, the scale of fault F₁ is large, the displacement of the basement fault F₁ is 51ms, and the overall displacement of (T₁—T₄) in the sediments is 35ms. Faults F₂—F₅ are located on the continental side of fault F₁ and can be combined into grabens and horsts in forms, which are inferred to be the associated faults of Fault F₁. It’s found that basement faults can be identified by both GI gun profile and sparker source profile, while the small faults can only be identified by the sparker profile. At the same time, the depth of upper breakpoint on the sparker profile is shallower, and the latest fault activity can be traced back to the Holocene. The locations and geometrical shapes of the three basement faults are similar on the two profiles, but there are imaging differences in the formation shapes around the faults and the distribution patterns of the secondary faults due to the influence of resolution. The similarity of fault detection results shows the effectiveness of the two methods, while the difference of profile imaging shows the necessity of combined detection in practical work.
According to the comparison of the two kinds of data, the sparker profile reveals a finer shallow structure than the GI gun profile does, and the GI gun profile can obtain a clearer basement structure. Based on the fusion results of the two kinds of data, the structural attributes of fault F₁ are further analyzed and explained in detail in this paper, and the Fault F₁ is the result of the reactivation of a basement pre-existing fault in the late Pleistocene and is a depression-boundary fault with an activity pattern of extensional normal faulting, and it is considered in this paper to be part of the South China Binhai fault zone. Therefore, it is necessary to attach importance to the combination of multiple detection methods in marine seismic zoning and marine seismic hazard assessment in order to obtain more detailed fault information.

The Weixi-Qiaohou Fault is located in the west boundary of Sichuan-Yunnan rhombic block, and also the north extension segment of active Red River fault zone. Strengthening the research on the late Quaternary activity of Weixi-Qiaohou Fault is of great theoretical and practical significance for further understanding the seismogeological background in northwest Yunnan and the structural deformation mechanism of the boundary of Sichuan-Yunnan block. Based on the 1︰50 000 active fault mapping and the research results of the National Natural Science Fund project, this paper mainly elaborates the latest active times of the fault and paleoseismic events along it revealed by exploration trenches at Matoushui, Shiyan, and Yushichang. Matoushui trench revealed three faults developed in late Pleistocene and Holocene pluvial fan accumulation, and the latest ages of faulted strata are(638±40)a BP and(1 335±23)a BP, respectively. The Shiyan trench revealed six faults, three in the western section and three in the eastern section. The three faults in the western section dislocated the late Pleistocene and Holocene accumulation, and the ¹⁴C ages of the latest faulted strata are(4 383±60)a BP, (4 337±52)a BP and(4 274±70)a BP, respectively; the other three faults revealed in the eastern part of the trench offset the Holocene fluvial facies accumulation, the ¹⁴C age of the latest faulted strata in the footwall of the main fault is(9 049±30)a BP, and the ¹⁴C ages of two sets of faulted sag pond deposits in the hanging wall are(1 473±41)a BP and(133±79)a BP, separately. Five active faults are revealed in Yushichang trench. Among them, the F₁ and F₂ dislocated the gray-white gravelly clay layer and the black peat soil layer. The ¹⁴C age of the gray-white gravelly clay layer is(1 490±30)a BP, and ¹⁴C ages of the upper and lower part of the black peat soil layer are(1 390±30)a BP and(1 190±30)a BP, respectively. The F₃ and F₄ faults offset the gray-white gravelly clay layer, the black peat soil layer and the brown yellow sand bearing clay, and the OSL age of brown yellow sand bearing clay is(0.6±0.2)ka. The F₅ fault dislocated the gray-white gravelly clay layer, its ¹⁴C age is(1 490±30)a BP. According to the relationship between strata and the analysis of dating data, the Yushichang trench revealed two seismic events, the first one occurred at(1 490±30)~(1 390±30)a BP, as typified by the faulting of F₅, the second paleoseismic event is represented by the faulting of F₁, F₂, F₃ and F₄.The F₁ and F₂ faulted the gray-white gravelly clay layer and the black peat soil. Fault F₃ and F₄ dislocated the gravelly clay, the peat soil and the sandy clay, and a seismic wedge is developed between fault F₃ and F₄, which is filled with the brownish yellow sandy clay. The OSL dating result of the brownish yellow sandy clay layer is(0.6±0.2)ka. Judging from the contact relationship between strata and faults, F₃ and F₄may also faulted the upper brownish yellow sandy clay layer, but the layer was eroded due to later denudation. Therefore, fault F₁, F₂, F₃ and F₄ represent the second event. Combined with the analysis of fault scarps with a height of 2~2.5m and clear valley landform in the slope near the fault, it is estimated that the time of the second paleoearthquake event is about 600 years ago, and the magnitude could reach 7. The trench at Gaichang reveals that the seismic wedge, soft sedimentary structure deformation and the medium fine sand uplift(sand vein)and other ancient seismic phenomena are well developed near the fault scarp. All these phenomena are just developed below the fault scarp. The vertical dislocation of the strata on both sides of the seismic wedge is 35cm, and ¹⁴C ages of the misinterpreted peat clay are(36 900±350)a BP and(28 330±160)a BP, respectively, so, the occurrence time of this earthquake event is estimated to be about 28 000a BP. If the fault scarp with a height of 2m was formed during this ancient earthquake, and considering the 0.35m vertical offset revealed by the trench, the magnitude of this ancient earthquake could reach 7.The Matoushui trench revealed three faults, which not only indicated the obvious activity of the faults in late Pleistocene to Holocene, but also revealed two paleoseismic events. Among them, the OSL age of the faulted sand layer by fault F₁ is(21.54±1.33)ka, which represents a paleoearthquake event of 20 000 years ago. The faulted strata by fault F₂ and F₃ are similar, which represent another earthquake event. The ¹⁴C dating results show that the age of the latest faulted strata is(638±40)Cal a BP, accordingly, it is estimated that the second earthquake time is about 600 years ago. A clear and straight fault trough with a width of several ten meters and a length of 4km is developed from Meiciping to Matoushui. Within the fault trough, there are fault scarps with different heights and good continuity, the height of which is generally 3~5m, the lowest is 2~3m, and the highest is 8~10m. Tracing south along this line, the eastern margin of Yueliangping Basin shows a fault scarp about 5m high. After that, it extends to Luoguoqing, and again appears as a straight and clear fault scarp several meters high. In addition, in the 2km long foothills between Hongxing and Luoguoping, there are huge rolling stones with diameters of 2~5m scattered everywhere, the maximum diameter of which is about 10m, implying a huge earthquake collapse occurred here. According to the length, height, width and dislocation of the rupture zone, and combined with the experience of Yiliang M≥7 earthquake and Myanmar Dongxu M7.3 earthquake, this earthquake magnitude is considered to be ≥7.

In order to realize the rapid and efficient identification of earthquakes, blasting and collapse events, this paper applies the Convolutional Neural Network(CNN)in deep learning technology to design a deep learning training module based on single station waveform recording of single event and a real-time test module based on multiple stations waveform recording of single event.
On the basis of ensuring that the data is comprehensive, objective and original, the three-component waveforms of the first five stations that recorded the P-wave arrival time of each event are input, and the current mainstream convolutional neural network structures are used for learning test. The four main convolutional neural network structures of AlexNet, VGG16, VGG19 and GoogLeNet are used for learning training, and the learning effects of different network structures are compared and analyzed. The results show that in the training process of various convolutional neural network structures, the accuracy rate and the cost function curve of the training set and the test set of each network are basically the same. The accuracy rate increases gradually with the increase of the training times and exceeds 90%, and finally stabilizes around a certain value. The cost function curve decreases rapidly with the increase of the training times, and eventually the stability does not change near a relatively small value. At the same time, over-fitting occurred in all convolutional neural network structures during training, except for AlexNet. In the end, the cost function of each type of structural training set and test set is finally lower than 0.194, and the recognition accuracy of each type of structure for training sets and test sets is over 93%. Among them, the recognition accuracy of AlexNet network structure is the highest, the accuracy of the training set of AlexNet network structure is as high as 100%, the test set is 98.51%, and no overfitting occurred; the accuracy of VGG16 and VGG19 network structure comes second, and the recognition accuracy of GoogLeNet network structure is relatively low, and the trend curves of the accuracy and cost function in training and test set of each network in the training process are basically the same. Subsequently, in order to test the event discrimination efficiency of the CNN in deep learning in the real-time operation of the digital seismic network, we select the trained AlexNet convolutional neural network to perform event type determination test based on the waveform recording of multiple stations of a single event. The final result shows that the types of a total of 89 events are accurately identified in the 110 events with M ≥0.7 recorded by Shandong seismic network, and the accuracy rate is about 80.9%. Among them, the accuracy rate of natural earthquake is about 74.6%, that of explosion is about 90.9%, and that of collapse is 100%. The recognition accuracy of collapse and explosion events is relatively high, and it basically reaches or exceeds the recognition accuracy of manual determination in the daily work of the seismic network. The accuracy of natural earthquake identification is relatively low. Among the 18 misidentified natural earthquakes, up to 13 events were judged as blasting or difficult to identify due to distortion of waveforms recorded by some stations(They are determined to be explosion and earthquake each by the records of two of the five stations). If sloughing off the recognition type error events caused by waveform distortion due to the background noise interference that overwhelms the real event waveform or waveform drift, the recognition accuracy of earthquake will become 91.4%, and the recognition accuracy of all events will increase from 80.9%to 91.7%, which is basically equivalent to the recognition accuracy of manual judgment in the daily work of the seismic network. This indicates that deep learning can quickly and efficiently realize the type identification of earthquake, blasting and collapse events.

Geomagnetic depth sounding is an effective method for exploring deep structure of the earth. There are dense geomagnetic observatories in China, which lays a foundation to obtain the electrical structure of the transition zone and the upper part of the lower mantle beneath China. However, the corresponding C-responses estimation methods which are applied now cannot get the stable C-responses for many observatories. Thus, a large amount of geomagnetic data is wasted. Therefore, in order to make full use of the geomagnetic data, the estimation of C-responses needs to be systematically studied. Because of the heterogeneous characteristics of the data quality of China's geomagnetic observation data, such as the quality of the data, the length of the record, the types of data(absolute and relative observation)and data discontinuity condition, many geomagnetic data are abandoned, this limits the resolution of mantle electrical structure studies. In this paper, the following techniques are used to improve the stability of the data and increase the number of the available geomagnetic observatories, in the meantime, the stability of the C-responses curves can be effectively improved:1)obtaining the stable spectrums of the different components for each frequency by the BIRRP(Bounded Influence, Remote Reference Processing)software, and using the global smoothing technique to suppress data noise on geomagnetic data; 2)As for the geomagnetic data which only records the relative variation of the D, H and Z components and doesn't have the baseline value, the horizontal component is decomposed by the approximate estimation method to obtain the C-responses of the relative variation data, and then the relative variation data is used directly for the C-responses estimation; 3)the effects of discontinuous data and short-record data on C-responses estimation are discussed. Under normal conditions, the discontinuity of the data has little influence on C-responses, and when the data length is shorter than 5 years, we can hardly get the available C-responses whose periods are longer than 40 days. All these experiments can provide a basis for the data processing of these kinds of observation data; 4)for coastal observatories, the ratio method is used to eliminate the influence of ocean effect on the C-responses functions. After carefully processing the data of more than 100 geomagnetic observatories in China by the above techniques, the stable C-responses function of 42 observatories is finally obtained, among them, the number of the observatories with C-responses ranging from 1.3 to 113.7 days is 24, and the observatories with C-responses ranging from 1.3 to 42.6 days are 18. The techniques of this paper can process heterogeneous data well and obtain more stable C-responses, which provides more basic data for high-resolution geomagnetic depth sounding inversion researches in China.

The 2014 Jinggu M6.6 earthquake attacked the Jinggu area where few historical earthquakes had occurred and little study has been conducted on active tectonics. The lack of detailed field investigation on active faults and seismicity restricts the assessment of seismic risk of this area and leads to divergent view points with respect to the seismotectonics of this earthquake, so relevant research needs to be strengthened urgently. In particular, some studies suggest that this earthquake triggered the activity of the NE-trending faults which have not yet been studied. By the approaches of remote sensing image interpretation, structural geomorphology investigation and trench excavation, we studied the late Quaternary activity of the faults in the epicenter area, which are the eastern margin fault of Yongping Basin and the Yixiang-Zhaojiacun Fault, and drew the conclusions as follows:
(1)The eastern margin fault of Yongping Basin originates around the Naguai village in the southeastern margin of Yongping Basin,extending northward across the Qiandong, Tianfang, and ending in the north of Tiantou. The fault is about 43km long, striking near SN. The linear characteristic of the fault is obvious in remote sensing images. Structural geomorphological phenomena, such as fault troughs, linear ridges and gully dislocations, have developed along the faults. There are several dextral-dislocated gullies near Naguai village, with displacements of 300m, 220m, 146m, 120m and 73m, respectively, indicating that the fault is a dextral strike-slip fault with long-term activity. In order to further study the activity of the fault, a trench was excavated in the fault trough, the Naguai trench. The trench reveals many faults, and the youngest strata offseted by the faults are Holocene, with ¹⁴C ages of(1 197±51)a and(1 900±35)a, respectively. All those suggest that it is a Holocene active fault.
(2)The Yixiang-Zhaojiacun Fault starts at the southeast of the Jinggu Basin, passes through Xiangyan, Yixiang, Chahe, and terminates at the Zhaojiacun. The total length of the fault is about 60km, and is a large-scale NE-trending fault in the Wuliangshan fault zone. Four gullies are synchronously sinistrally dislocated at Yixiang village, with the displacements of 340m, 260m, 240m and 240m, indicating that the fault is a long-term active sinistral strike-slip fault. A trench was excavated in a fault trough in Yixiang village. The trench reveals a small sag pond and a fault. The fault offsets several strata with clear dislocation and linear characteristic. The thickness of strata between the two walls of fault does not match, and the gravels are oriented along fault plane. The offset strata have the ¹⁴C age of(2 296±56)a, (3 009±51)a, and(4 924±45)a, respectively, and another two strata have the OSL age of(1.8±0.1)ka, (8.6±0.5)ka respectively, by which we constrained the latest paleoearthquake between(1.8±0.1)ka(OSL-Y01)and(378±48)a BP(CY-07). This again provides further evidence that the fault is a Holocene fault with long-term activity.
(3)Based on the distribution of aftershocks and the predecessor research results, the 2014 Jinggu M6.6 earthquake and the M5.8, M5.9 strong aftershocks are regarded as being caused by the eastern margin fault of Yongping Basin, which is part of the Wuliangshan fault zone. The seismogenic mechanism is that the stress has been locked, concentrated and accumulated to give rise to the quakes in the wedge-shaped area near the intersection of the SN and NE striking faults, which is similar to the seismogenic mechanism in the southwest of Yunnan Province.

In order to understand the mechanism of the 1668 M_S8.5 earthquake occurred in Tancheng, it is important to probe the fine deep geological structure beneath the epicenter. A MT profile 20km south of the epicenter has been deployed. There are 17 sites along the profile, with a 3km average separation. Signals in Ex, Ey, Hx and Hy were measured in a cross manner, with x-axis orientated to the north. Record length for each site was at least 20h. The impedance and phase at sites in high cultural noisy environment were estimated by remote reference technique. As the Tanlu Fault Zone(TLFZ)is in NNE, nearly northerly, thus YX mode was considered as TM mode. Gauss-Newton inversion was done in 2-D mode with only the TM impedance and phase as input data. The electrical sections of 10km and 40km depth were respectively obtained after 8 iterations. The both initial models were created by Bostic approximation. The sections reveal the following features.
The TLFZ consists of five faults, from east to west numbered as F₀ to F₄. F₁ is the primary fault, steeply dipping west down to mantle, which has turned into a buried one overthrust by the east dipping Fault F₀. F₂ and F₃ dip east at 45 degrees, parallel to F₄, truncated by F₁ at depth. F₄ dips east in the shallow subsurface and gradually dips to west toward depth through the entire crust merging with F₁ to form a bigger one. These four faults constitute a flower-shaped structure, showing the nature of strike-slip of the TLFZ, associated with normal faulting in the late Yanshanian to early Himalayan. F₁ dips west, overthrust by east-dipping F₀, implying the compression from the westward subduction of the Pacific plate, thus present-day compression is superposed on the early tensile and strike-slip feature.
Based on MT data, it is inferred that the 1668 Tancheng M8.5 earthquake occurred at the junction of F₁ and F₃ about 15km deep. Thus it was likely resulted from westward compression of the Pacific plate, leading to thrust of the Sulu uplift along F₀, inducing activity of F₁ at depth, reactivated F₃, and adjusting the stress distribution in the region.

This paper tries to formulate the C-response of geomagnetic depth sounding(GDS)on an Earth model with finite electrical conductivity. The computation is performed in a spherical coordinate system. The Earth is divided into a series of thin spherical shells. The source is approximated by a single spherical harmonic P₁⁰ due to the spatial structure of electrical currents in the magnetosphere. The whole solution space is separated into inner and external parts by the Earth surface. Omitting displacement current, the magnetic field in the external space obeys Laplacian equation, while in the inner part, due to the finite conductivity, the electromagnetic fields obey Helmholtz equation. To connect the magnetic fields in the inner and external space, the continuity condition of magnetic fields is used on the Earth surface. The external magnetic fields are expressed by the inner and external source coefficients, from which a new parameter called C-response is computed from the inner coefficient divided by the external coefficient, thus normalizing the actual source strength. The inner magnetic fields in each layer can be recursively derived by the continuity boundary condition of both normal and tangential components of the magnetic field from the initial boundary condition at core-mantle-boundary. The consistency of our C-responses with that from a typical 1-D global model validates the accuracy of the proposed algorithm. Numerical results also show that the C-response estimated from the geomagnetic transfer function method will deviate exceeding 5%from the actual response at longer periods than about 10⁶s, which means that ignoring the curvature of the Earth at extreme long periods will make inversion result unreliable. Therefore, an accurate C-response should be computed in order to lay a solid foundation for reliable inversion.

The 3 August 2014 Ludian, Yunnan M_S6.5 earthquake has spawned more than 1, 000 landslides which are from several tens to several millions and over ten millions of cubic meters in volumes. Among them, the Hongshiya and Ganjiazai landslides are the biggest two with volumes over 1 000×10⁴m³. The Hongshiya and Ganjiazai landslides are two typical landslides, the former belongs to tremendous rock avalanche, and the latter belongs to unconsolidated werthering deposit landslide developed in concave mountain slope. Based on field investigations, causes and formation mechanism of the two landslides are discussed in this study. The neotectonic movement in the area maintains sustainable uplifting violently all the time since Cenozoic. The landform process accompanied with the regional tectonic uplifting is the violent downward erosion along the Jinshajiang River and its tributary, forming landforms of high mountains and canyons, deeply cut valleys, with great height difference. The regional seismo-tectonics situation suggests that:Ludian earthquake region is situated on the southern frontier boundary of Daliangshan secondary active block, and is seismically the strongest active area with one earthquake of magnitude greater than M5.0 occurring every 6 years. Frequent and strong seismicity produces accumulated effects on the ground rock to gradually lower the mechanical strength of slopes and their stability, which is the basis condition to generate large-scale collapse and landslide at Hongshiyan and Ganjiazhai. The occurring of Hongshiyan special large rock avalanche is associated with the large terrain height difference, steep slope, soft interlayer structure and unloading fissures and high-angle joints. The formation mechanism of Hongshiyan rock avalanche may have three stages as follows:Stage 1, when P wave arriving, under the situation of free surface, rocks shake violently, the pre-existent joints(in red)parallel to and normal to the river and unloading cracks are opened and connected. Stage 2, on the basis of the first stage, when S wave arriving, the ground movement aggravates. Joints(in green)along beds develop further, resulting in rock masses intersecting each other. Stage 3, rock masses lose stability, sliding downward, collapsing, and moving over a short distance along the sliding surface to the inside of the valley, blocking the river to form the dammed lake. The special large landslide at Ganjiazhai is a weathering layer landslide occurring in the middle-lower of a large concave slope. Its formation process may have two stages as follows:Firstly, under strong ground shaking and gravity, the ground rock-soil body around moves and assembles to the lower of the central axis of the large concave slope, which suffers the largest earthquake inertia force and firstly yields plastic damage to generate compression-expansion deformation, because of the largest water content and volume-weight within the loose soil of it. Secondly, in view of the steep slope, along with the compression, the plastic deformation area enlarges further in the lower of slope, giving rise to a tensional stress area along the middle of the slope. As soon as the tensional stress exceeds the tensile strength of the weathering layer, a tensional fracture will occur and the landslide rolls away immediately making use of momentum. This two large landslides are the basic typical ones triggered by the M_S6.5 Ludian earthquake, and their causes and mechanism have a certain popular implication for the landslides occurring in this earthquake region.

This study provides new seismo-electromagnetic data processing methods to extract the anomalous signals by combining the principal component analysis(PCA)and local correlation tracking(LCT)methods. The PCA method can separate signals of different frequencies by projecting them to different axes according to their energy. So it can solve the problems of identifying the relatively weak signals in strong interference background to a certain extent. The LCT method is more suitable for non-stationary signal processing compared with classical cross-correlation method. This method is based on the good spatial correlation between the magnetic field components of different ELF stations to pick up the correlation coefficient, so as to achieve the purpose of weak anomalies signals identification. As a case study of the M4.6 Jinggu earthquake in Yunnan, China, we investigated the electromagnetic data observed by ELF stations near the epicenter. First, we applied the PCA method to the magnetic-filed data and got the temporal variation of percentage of each principal component. The results indicate that the contribution of the second principal components, which may relate with the earthquake, increased significantly about a week before the earthquake. Then, we applied the LCT method to the magnetic-filed data processing as well, and the results of both north-south and east-west magnetic field components showed that local correlation coefficient saw anomalies about a week before the earthquake, which had a good corresponding relationship with the former results by PCA method. Both means for the same earthquake case got a consistent conclusion, which not only enhanced the reliability of the results, but also confirmed the effectiveness of two methods applied in the earthquake-related electromagnetic anomalies extraction. We also discussed the possible relationship between these anomalies and the earthquake. Although there are no direct evidence and supporting theories in terms of the relationship between abnormal electromagnetic signals and earthquakes at present, the studies in this paper may strengthen the understanding of seismoelectromagnetic phenomena and promote further research.

Since stratigraphic formation is influenced by tectonic activities and climate since late Pleistocene,it is important to build the stratigraphic sequence to improve the research of active tectonics,climatic change and landform factors.Zoige Basin is located in the eastern edge of Tibet Plateau where the tectonic is active and the Chinese monsoon is strong.The research of stratigraphic sequence is closely related to the tectonic activities and climate changes.Based on 26 typical stratum profiles revealed by lacustrine boreholes,terraces,peat deposits and trenching,203 isotope dating data were obtained by AMS and OSL methods.We conduct a stratigraphic correlation and classification in Zoige Basin since the Late Pleistocene.Sedimentary cycles are divided into six sedimentary rhythms (75~42ka,42~37ka,37~20ka,16~11ka,11~4ka and 4~0ka) and six marker beds (fine sand of 75~55ka and 22~20ka,gray silt deposit or gravel deposit of 13~9ka,black sandy clay containing carbonaceous deposits of 4ka,2ka and 0.3ka).There is a close relation between strata and tectonic-climate.On the one hand,sedimentary cycles coincide with climate change and have a good correspondence with ocean oxygen isotope.On the other hand,sedimentation characteristics is influenced by the persistent activities in neotectonic period of the east Kunlun fault zone on the north side and the Longrize fault zone on the west side.Marker beds and sedimentary cycles are compared with the strata in adjacent areas.It shows that climate change is the main factor affecting sedimentary cycle.The difference of stratum thickness and its spatial distribution is also affected by tectonic activity.

The Anqiu-Juxian Fault is a major branch fault and an active prominent fault of the Yishu Fault belt. The spatial distribution, geometric features and the latest activities of the Anqiu-Juxian Fault are studied by field survey and mapping in this study. The northern segment of the Anqiu-Juxian Fault between Juxian and Changyi can be divided into four segments, namely from north to south, the Changyi-Nanliu segment, the Anqiu-Mengtuan segment, the Qingfengling segment and the Mengyan segment. These segments are left-step en echelon arranged, and each of the fault segments consists of right-step en echelon arranged sub-segments. The Changyi-Nanliu segment is about 31km long and composed of 4 sub-segments in right-step en-echelon arrangement, namely, Wenshan sub-segment, Zhuli sub-segment, Shuangguan-Meicun sub-segment and Nanliu sub-segment, from north to south. The length of these sub-segments is 5km, 7km, 10km and 9km, respectively. The width of the stepover between them is about 2～3km. The Changyi-Nanliu segment generally strikes～15°, and the fault plane dips both west and east with dip angle 70°～80°. This segment offsets the widely distributed eolian yellow or orange fine sand and silt that were formed in the latest late Pleistocene, and it also offsets the mid-Holocene grey-yellow clay. The latest active age of the Changyi-Nanliu segment is the middle and late Holocene. This segment is characterized by right-lateral strike-slip motion with thrust and normal fault component, and the normal faulting activity is usually younger than the reverse faulting activity. The Anqiu-Mengtuan segment is about 50km long and exposes～21km. It strikes 15°～20°with the major fault plane dipping NWW with dip angle 70°～80°. This fault segment is characterized by right-lateral strike-slip motion with west-to-east thrust component. The segment can also be divided into two sub-segments, namely, the 13km long Anqiu-Guangong sub-segment and the 8km long Anshang-Mengtuan sub-segment, as in right-step en echelon arrangement, with a stepover of about 3km in width. The youngest offset stratum along the Anqiu-Mengtuan segment is the late Pleistocene, so, its latest active age is the late Paleocene and early Holocene. The Qingfengling segment is about 32km long, striking 15°～20°, dipping mainly southeast and partly west with dip angles more than 60° generally. This segment is characterized by right-lateral strike-slip motion with minor thrust component. It is composed of 4 sub-segments, which are the Xiaodianzi-Henhushan sub-segment, Kushan-Chezhuang, Maobu and Wangtaizi sub-segment, respectively from north to south. The length of these sub-segments is 6km, 8km, 14km and 4km, respectively. The former three sub-segments are aligned right-laterally. The Qingfengling segment offsets the upper late Pleistocene and the early Holocene strata; its latest active age is the early Holocene. The Mengyan segment exposes about 20km, striking 20° and dipping northwest with dip angle ～70°. It is also characterized by right-lateral strike-slip motion with thrust component, and its latest active age is the early Holocene.
The only historical earthquake that occurred on the north segment of the Anqiu-Juxian Fault between Juxian and Changyi is the 70BC Anqiu M7 earthquake. However, paleo-earthquake researches show that several strong earthquakes occurred along the Qingfengling segment and the Mengyan segment between the latest late Pleistocene to early Holocene. The time of the latest strong earthquake is ～3 500a BP, 2 084a BP (-70BC), ～10 000a BP, ～10 000a BP on the Changyi-Nanliu segment, Anqiu-Mengtan segment, Qingfengling segment, and Mengyan segment, respectively. Since the strong earthquake recurrence interval is still not known for each segment, the exact time for the next strong earthquake can't be predicted. However, according to the geometric features, latest active age, latest activity features, historic earthquake data and paleoearthquake documents of this active fault, the 4 segments do have seismotectonic conditions for generating M≥7 earthquake, and the potential earthquake risk does exist and may be rather high and imperative. Thus, the fault activities and the potential earthquake hazard should be considered during future earthquake hazard prevention and prediction.

The Zhaotong-Ludian Fault zone, composed mainly of three right-step en echelon faults, namely, the Zhaotong-Ludian Fault, the Sayuhe Fault and the Longshu Fault, strikes 40°～60° on the whole, with the Sayuhe Fault and the Longshu Fault dipping SE and the Zhaotong-Ludian Fault dipping NW, and they all together constitute a complicated thrust fault system. Based on years of field investigation results of geology and geomorphography, we elaborate the late Quaternary active features, the geological and geomorphic evidences of the latest activity of the Zhaotong-Ludian Faults. Our observation shows that: the late Cenozoic basins along the Zhaotong-Luian Fault zone are obviously dominated by the fault; there are many neo-active fault landforms, such as, flat and straight fault troughs, directional aligned fault facets and fault scarps, and the upper Pleistocene to Holocene strata are offset by the fault. The fault zone has been active since the late Quaternary. For example, the fault at Daqiaobian dislocated a set of strata of the Pliocene, and middle to upper Pleistocene, with an apparently reverse character. The fault trending NE is developed in the Holocene diluvium with oblique striation on the fault plane at Guangming Village. Deposits with an OSL age of(23.4±1.8)ka BP on T2 terrace of a small river near Beizha town was offset by the fault. There is a fault scarp trending NE 40°, 0.5～2.0m in height, on the first terrace of the Longshu River near the Longshu Village. Several Quaternary faults are revealed by the trench which offset the late Pleistocene to Holocene strata and there are three poleo-earthquake events discovered in the trench. At Yanjiao Village the gravel layer has risen steeply and is aligned in a line because of squeezing effect of the fault; the rivers and ridges nearby are synchronously offset dextrally up to 30～40m. The fault zone is dominated by reverse faulting with a small amount of right-lateral motion. Besides, there are some NW-trending faults interweaving with the NE-trending fault zone, some of which are active since late Quaternary as well, and they are the conjugate structures with the NE-trending faults. Surface deformation, such as NE- and NW-trending ground fissures and reverse scarp landforms, has been generated in the epicenter area of the 2014 Ludian M6.5 earthquake, the distribution of which is in consistence with the NE- and NW-trending faults. Because of far-field deformation response and energy exchange and transfer between blocks, the Liangshan active sub-block formed on the east of the Sichuan-Yunnan block, and the Zhaotong-Ludian Fault zone lies in the forefront of the SE movement of this sub-block. On account of its distinct location and its complicated geometric structure, the Zhaotong-Ludian Fault zone is one of main carriers of the tectonic deformation of the Liangshan active sub-block to absorb and accommodate the strains produced by the block's SE movement, and is the southern boundary of the Liangshan sub-block. From the point of view of the regional tectonic positions and the kinematic characteristics, the relation of Zhaotong-Ludian Fault zone to the Liangshan active sub-block is exactly as the relation of the Longmanshan Faults to Bayan Har block. Consequently, the Zhaotong-Ludian Fault zone has an important significance in the division of active block boundaries and the regional tectonic framework, and meanwhile, it is also an important seismogenic structure in the northeastern Yunnan.

The terraces along the upper reaches of the Minjiang River record rapid uplift around the Tibetan plateau since the Quaternary. However, the common method to obtain the terraces elevation data always relies on single point or line to represent the whole landform. The available results and further analyses are usually not continuous and systematic. All of these, therefore, restrict the development of further study. A large number of qualitative information and the digital terrace models are relied on the improvement of DEM processing technology. Our present paper applies fuzzy C-means algorithm to a bunch of the Minjiang River terrace cross sections in different dimensions to get the integrated distribution characteristics of the terraces in the Zhangla Basin. SPOT5 image is adopted when generating the DEM. Then we consider both the spatial correlation and distribution characteristics of terrace cross sections into analysis and we consequently find a successful way to extract different levels of terraces based on both whole indicator and internal correlation. The precision evaluation suggests that result of terrace extraction is highly consistent with the field survey data. Based on these, we discuss relationship between characteristics of terraces, the incision rates and the regional uplift patterns. We suggest that a clustering approach for incorporating spatial dependence into the automatic fluvial terrace extraction can be used to study and understand the regional tectonic and geomorphic features in the layered landscapes.

Since the Wenchuan earthquake,the seismic hazard of the northeastern segment of the Longmenshan Fault zone as well as the Hanzhong Basin has drawn more and more concerns. However,the essential data needed for further analysis on the seismic hazard in this region is scarce at present time,hence there is an urgent need for an in-depth study on the activities of faults around the basin. The faults around the Hanzhong Basin include five main faults,namely,the northern margin fault of Hanzhong Basin,the southern margin fault of Hanzhong Basin,the Qingchuan Fault,the Chaba-Lin'ansi Fault and the southern margin fault of Liangshan. Based on several detailed field investigations on the geometric distribution,movement nature and active ages of the five faults,and with consideration of previous work,our study shows that the late-Quaternary tectonic activity in the basin is relatively intense in west and weak in east. The west section of the north-margin fault of the Hanzhong Basin(east of Baohe)was active in early late Pleistocene,while its eastern section(west of Baohe)was active in middle Pleistocene. The south-margin fault of the basin was also active in middle Pleistocene. And the three faults in the southwest of the basin were all active in late Pleistocene. This activity pattern of high in the west and low in the east is also demonstrated by the difference in thickness of Quaternary system and the distribution of small earthquakes.

Based on the interpretation of satellite image,field investigation and geomorphic survey and sample dating of surface,the strath terrace and fill terrace at the outlet of Moleqiehe River on the western segment of Altyn Tagh Fault zone(ATF)are used to study the tectonic uplift rate,uplift model and aggradation rate,and cooperated with data of the regional climate,the response to climate evolution of development of terrace are discussed.The previous studies of the terraces related with the Altyn Tagh Fault zone are mainly focused on the horizontal offset in order to obtain sinistral-slip rate,but few studies involve the uplift using terraces.As a structural zone with strike-thrust characteristic,the ATF is a boundary structure of the northern fringe of Qinghai-Tibet Plateau,and its thrusting and uplifting movement is of significance for controlling the uplift of the northern fringe of the Plateau.Therefore, the study of uplift of the ATF will be helpful for understanding the uplift model and mechanism and promoting the kinematic study of the northern fringe of the Plateau.The formation of strath terrace is closely related with tectogenetic movement,the landform age of the terrace represents the starting time of uplift.Based on the height and strath and landform age,the uplift rate can be calculated.The fill terrace is formed by climate forcing,the surface age represents the end time of one aggradation event.If the starting time of aggradation is obtained,the aggration rate can be calculated.There are four stream terraces at the outlet of Moleqiehe River(T₄,T₃,T₂,and T₁). T₄ and T₃ are strath terraces,T₂ is fill terrace,and T₁ are fill-cut terraces.The landform ages of T₃,T₂,and T₁ are 18.98±1.42ka BP,13.08±1.01ka BP,and 5.72±0.43ka BP,respectively.The existence of T₃ reveals the uplift rate of 6.66±0.50mm/a since 18.98±1.42ka BP.The existence of T₃ and T₂ reveals the time of fast uplift movement and aggradation events between 18.98±1.42ka BP to 13.08±1.01ka BP,the uplift rate is bigger than 20mm/a and the aggradation rate bigger than 10mm/a.The model of tectonic uplift shows tilted uplift from south toward north across the ATF,and this model is one of the types of the Qinghai-Tibet Plateau extending toward north.The aggradations,that constructed the T₂,are the result of the coactions of fast uplift and deglaciation climate between 15ka BP to 12ka BP.

Magnetic Induced Polarization(MIP) is one kind of electrical conductivity methods,which measures magnetic field point by point rather than measuring the electric field difference between two electrodes in Electric Induced Polarization(EIP).Compared to EIP,MIP surveys have certain merits,such as:bigger detection depth,higher resolution,capable to provide useful information through highly conductive overburden and high-resistance or low-resistance cover,no need for grounding,and etc.It will be a deep exploration technology and applied to areas such as low-resistance overburden,Gobi desert,bedrocks which have been a major impediment to EIP surveys.In order to test the ability of obtaining IP anomaly in MIP surveys,we carried out a trial comparative study between MIP and EIP(Time-domain IP method and multi-frequency phase IP method included) in a mining area in China.Test results show that there is a good consistency in the IP anomaly about the two detection methods measured in the same survey lines and it will provide a basis for the further studies of MIP.In the paper the basic principles,techniques,and data processing of this method(MIP) are discussed,and the results between Magnetic Induced Polarization and Electric Induced Polarization are analyzed.

Cosmogenic dating(in situ cosmogenic nuclides)has been widely used in geologic and geographic domain along with the application of accelerator mass spectrometry(AMS)and highly sensitive conventional noble-gas mass spectrometry in the early 1980s,since Raymond Davis and Oliver Schaffer proposed that cosmogenic nuclides produced within minerals at the surface of the earth could be applied to geological problems.At present,this dating method is used to study glacier,desert and relief evolution,volcano development and active structure movement etc.There are a lot of materials which can be used in cosmogenic dating method.Because of its tight crystal structure,quartz can minimize possible contamination by meteoric ¹⁰Be,and also for its low ²⁷Al content,it becomes an ideal material for cosmogenic dating method.During the process of the isolation of ¹⁰Be and ²⁶Al from quartz sample,quartz isolation and purification are one of the key steps.HCl/H₂O₂ and HF/HNO₃ leaching is a safe method used to isolate and purify the quartz samples widely.However,there is defect about the method.It takes more time for sample preparation and curtails the life of instrument.Based on the above method,three experiments were designed to compare with the primary one.The results of experiment reveal that the purification efficiency can be enhanced by increasing the leaching solution concentration and the amount of sample,and ultrasonic cleaner can be replaced by hotplate /magnetic stirrer.

Calcite is a common matter in the fault zone and it is often related with fault movement,so its dating is of vital significance for studying the time of fault movement.At present,ESR method is one of the ways for measuring the age of calcite,but there are no final conclusions regarding the ESR signals and measurement conditions of calcite.The samples used in this article were picked from the east of Erhai,Yunnan.According to the preliminary study of the samples,we found that calcite was liable to generate unstable short-lived signal when it was artificially exposed.So before measuring,the samples were kept under the room temperature condition for at least 5 days to eliminate the jamming signals.Generally,in natural calcite,there are many ESR signal centers,among them,the ones,g=2.0040 and g=2.0023,respond well to absorbed dose,and can be used in the dating.Growth curves of these two signals indicate a linear growth at least in the range of 1500Gy.But it is indicated with the artificially fixed known dose method that different microwave powers have to be taken for g=2.0040 and g=2.0023 signals.If we consider the deviation of ED value is smaller than 5%,then the microwave power should be 0.8 or 2mW for the g=2.0040 signal,and the microwave power be 2mW or 5mW for the g=2.0023 signal.

As a fault depression basin developed in the middle and late Quaternary,the Fuzhou basin consists of complicated deposit structures,including deposits of marine facies,terrestrial facies and marine and terrestrial alternating facies. The thick source alpha counter and flame-photometer are used to measure the content of radioactive elements (uranium,thorium and Potassium) of two drill holes (SZK1 and SZK2) strata in Fuzhou basin. Comparison of the content of radioactive elements with the result of spore-pollen analysis of one drill hole (SZK1) strata shows that the former changes with the variations of lithology and depositional environment. The content of radioactive elements in sediment is related to the scale of sediment granularity and lithology,being high in argillaceous sediments (e.g. silt and clay),low in sandy sediments and median in gravel strata. On the other hand,the contents of radioactive elements are relative to paleoenvironment,the warm and humid environment is propitious for high concentration of radioactive elements; while cool and dry climate is just the opposite.

Fuzhou basin is tectonically located in the eastern part of South China fold system,within the Fuding-Yunxiao Fault depression area of East Fujian volcanic depression zone. Based on detailed logging,and combined with results of other researchers,the preliminary summarization on sedimentary characteristics of late Pleistocene of Fuzhou basin is done. The depositing of Fuzhou basin began from about 56.5ka BP,sediments comprise gravel,sand,clay and silt,showing a depositional sequence with the granularity becoming small from lower to upper,and a bigger variation of lithologic properties in horizontal direction. There are three silt layers in late Quaternary strata of Fuzhou basin. The first and second layers developed during the middle and late Holocene,the depositing time of them is about from 7.86ka BP to 3ka BP to 3.08ka BP and they are the results of “Changle Transgression”; the third developed during the late of late Pleistocene and the depositing time is about from 44ka BP to 20ka BP and it is the result of “Fuzhou Transgression”. The buried late Quaternary sediments in Fuzhou basin can be divided into upper Pleistocene series and Holocene series,which include four formations from old to new. They are Longhai formation(Q³_pl),Dongshan formation (Qhd),Changle formation (Qhc)and Jiantian formation (Qhj) respectively. The Longhai formation can be divided into three members,the sediments of the lower member are celadon,yellow gravel and moderate coarse sand; the middle member is of gray clay,silt,sand and gravel,or interbed of silty clay and fine sand; the upper member is of yellow moderate coarse sand. Dongshan formation comprises gray silty sand and clay. Changle formation can be divided into two parts,the lower part is of interbed of silt and flour sand; the upper is of gray and plumbeous silt or douke silt. The Jiantian formation comprises light gray,brown clay.

Several dark-green or yellow-brown layers with different thickness have been found in the late Quaternary strata in the area of the Changjiang River and Jiantangjiang River. They are very hard,called “hard clay”. The top hard clay layer is called the first hard clay. There are the marine sediments on the first hard clay. The boundary of Holocene has been discussed based on the first hard clay. The first hard clay is found in Ningbo region too. 19 samples from the first hard clay,the marine sediments on the first hard clay and the layer under the first hard clay,were dated using optical luminescence and 14 C dating method to provide age control of its development. According to our dating results,the age of the original material of the first hard clay is 45～55ka BP. The first hard clay is uncontinuously distributed,being eroded by the later stages transgression in some places,and unconformable with the upper marine sediments. The lower limit of the marine bed may not be simply delineated as the Holocene boundary. It is possible that the marine sediments occurred at the end of late Pleistocene.

The information of tectonic movement in the earthquake-stricken areas of the 1979 M_S 6.0 Wuyuan earthquake,the 1989 M_S 6.1 Datong-Yanggao earthquake and the 1996 M_S 6.4 west Baotou earthquake are studied by using multi-phase and multi-channel satellite images. Furthermore,the geologic-tectonic background of these earthquakes is discussed by combining the obtained results with the existing seismic intensity investigation data. The results of this study show that the 1979 M_S 6.0 Wuyuan earthquake occurred at the convergence of the NE-trending Haiziyan Fault and the NW-trending west Wuyuan Fault,as indicated by tone difference and micro-morphologic features on satellite images. The long axis of the high intensity isoseism of this earthquake is NE-trending,consistent with the Haiziyan Fault,which is therefore assumed to be the seismogenic fault of this earthquake. The long axis of the low intensity isoseism of this earthquake is NW-trending,indicating the effect of the NW-trending west Wuyuan Fault. The 1989 M_S 6.1 Datong-Yanggao earthquake occurred on the NEE-trending Dawangcun-Xi'anquan Fault which extends from the hinterland of the Liulengshan Mountains to Datong-Yanggao basin,as indicated by micro-morphologic and tone differences on the satellite images. The fault is consistent with the long axis of the isoseism of this earthquake as proposed by Su Zong-zheng et al. The 1996 M_S 6.4 west Baotou earthquake occurred at the convergence of the Wulashan north marginal fault and the NW-trending Xinsheng Sand Yard Fault,as indicated by distinct scarp and groove landforms. The long axis of high intensity isoseism of this earthquake is NE-striking,sub-parallel to the NEE-trending Wulashan north marginal fault that is assumed to be the seismogenic structure of the west Baotou earthquake. The long axis of low intensity isoseism of this earthquake is NW-striking,which is considered to be the effect of the NW-trending Xinsheng Sand Yard Fault.

Among the Altun Tectonic system,there are thrust faults and thrusting components in the Altun strike-slip fault zone. Being similar to the Altun strike-slip fault zone,the thrust fault can be divided into three segments. They are the western segment,middle segment and eastern segment. The western segment is from Ayinaike to the debouchment of the Cheerchenhe River. There is thrusting component on the southern marginal fault of Altun and there are small thrust faults in piedmont in the western segment and evidences are discovered that show the latest terraces and alluvial fan being affected by thrust activity. The middle segment is from the debouchment of Cheerchenhe River to Lapeiquan. There are large-scale thrust faults in the northern margin of Altun Mountains,and there is evidence to show the younger surface of landform displaced by the thrust activity of the middle segment. The eastern segment is from Lapeiquan to Kuantanshan. There is thrusting component on the northern marginal fault of Altun,and thrust faults are developed on the northern margin of Altun Mountains and pediment alluvial fan. From the middle and late Pleistocene to present,the thrust rate on the middle and eastern segments of thrust fault is less than 2mm/a. Since 16ka BP,the thrust rate in Jianggalasayi on the west of the middle segment is 0.33mm/ a,the thrust rate in Milanqiao on the middle part of fault is 1.42mm/a since 32ka BP. The thrust rate of the middle part is bigger than that of the western part. The biggest thrust rate of the eastern segment is at Tuanjiexiang near to the middle part,being 1.81mm/a. From the middle part to the eastern or western part,the thrust rate becomes smaller. At Liuchengzi of the western part,the thrust rate is 0.57mm/a since 72.36ka BP. While at Hongliugou of the eastern part,the thrust rate is only 0.05mm/a since 8.99ka BP.

On November 14,2001,an M_s 8.1 earthquake occurred in the west of Kunlunshan Pass. This event is the greatest earthquake that occurred in China continent following the Zayu-Medog,Xizang M_s8.6 earthquake of August 15,1950 and the Damxung,Xizang M_s 8.0 earthquake of November 18,1951. The earthquake occurred in Hoh Xil,a depopulated zone in the northern Qinghai-Xizang Plateau,where field investigation is very difficult to be carried out due to harsh climate and thin air. High spatial resolution satellite images are applicable to the interpretation of earthquake surface rupture zone. The 10m spatial resolution SPOT images may reveal distinctly the major rupture zone,while the 1m spatial resolution IKONOS images may display the fine structures and kinematical characteristics of the surface rupture. This paper presents the results of the study on the surface rupture of the west of Kunlunshan Pass earthquake by using SPOT and IKONOS images. The study shows that the surface rupture zone is located mainly along the abrupt variation zone of landforms on the alluvial-pluvial platform or the rear edge of the platform at the southern foot of the Kunlunshan Mountains. The surface rupture to the east of Buka Daban Peak is about 350km in length and striking 100°,consisting of 3 sub-rupture zones. This surface rupture is superimposed on the preexisting rupture zone. A series of gullies crossing the rupture zone have been offset left-laterally,and the average slip rate is estimated to be 13.4～16.8mm/a. The interpretation of the SPOT and IKONOS images demonstrate that the macroscopic epicenter of this earthquake should be at 93°17′E,35°47′N,in the vicinity of the Yuxifeng Peak,where the maximum left-lateral horizontal dislocation is 7.8m and the width of the surface rupture zone reaches up to 1250m. This conclusion is in good agreement with the location of macroscopic epicenter (93.3°E,35.8°N ) deduced by China Earthquake Administration for this earthquake.

Strath terrace is tectogenetic stream terrace.In equilibrium state,lateral erosion of the stream bevels bedrocks beneath the active channel,giving rise to widening of the channel and the formation of strath.Downcutting of the stream due to tectonic movement causes the rise of strath above the channel,resulting in strath terrace.Strath includes major strath and minor strath,and accordingly strath terrace includes major strath terrace and minor strath terrace.The existence of strath terrace may indicate that stream has once reached the equilibrium state,and that tectonic uplift has occurred.The investigation of strath terrace,therefore,may reconstruct the history of tectonic uplift of the studied area.Yandantu and Changcaogou are located at the middle segment of the Altun Northern Marginal Fault,where multistage strath terraces are well developed.Basing on field investigation of strath terraces and thermoluminescent(TL)dating of the ages of treads of terraces,and combining with the data of paleoclimate of Qinghai-Xizang Plateau,we discuss the tectonic uplift since late Epipleistocene as reflected by strath terraces in these two areas.At Yandantu,three levels of stream terraces(T₁,T₂and T₃)have been developed since16ka BP,among which T₁and T₃are strath terraces,while T₂is fill terrace.A buried major strath is exposed in the area as well.The ages of three treads are dated to be about16.1ka BP,12.8ka BP and 6.2ka BP,respectively.The three terraces reflect three tectonic uplift events,while the three ages of treads represent the occurrence time of these events.The stream is still beveling the bedrock and widening channel at pre-sent,and modern strath is being generated.The uplift rate is 4.8～4.5mm/yr since 16.1ka BP in this area.From 12.8ka BP to 62ka BP,the uplift rate was 6.4mm/yr,while since 6.2ka BP till nowthe uplift rate is 3.1mm/yr. At Changcaogou,four levels of stream terraces(T₁,T₂,T₃and T′₁)have been developed since 7ka BP.All of them are fill terraces,and there are buried straths under deposits of every terrace.The buried straths of T₃and T₂are major strath,and those of T′₁and T₁are minor strath.The ages of treads of three terraces(T₃,T₂and T′₁)are 7ka BP,3ka BP and 2.5ka BP,respectively.The four terraces reflect two uplift events induced by tectonic activities.One of them occurred in about 7ka BP,and the other in 3ka BP.The uplift rate is 5.9mm/yr since 7.0ka BP at Changcaogou.From7ka BP to 3ka BP,the uplift rate was 7.0mm/yr,and since 3ka BP till now the uplift rate is 4.7mm/a.

It is well known that weathering process initiates as soon as the volcanic products reach the earth surface. As a result, various kinds of secondary clay minerals and their polytypes are formed in different periods of weathering process. Based on this fact, the time sequence and succession of volcanic eruption can be inferred. It may provide, therefore, a convenient dating method applicable to the prediction and assessment of volcanic hazards. In this paper, the weathering features of clay minerals from 3 profiles of volcanic ejecta in Leihuling-Ma'anshan area, northern Hainan Island are studied in detail by using various techniques including scanning electron microscopy, energy spectroscopy and x-ray diffraction analysis. The results of this study have led to some new insights about the clay minerals and their weathering features, as well as the time sequence of volcanic eruption in this area. (1) The spheroidal halloysite is predominant in the secondary clay minerals formed by the weathering of volcanic products in this area. The interlayer water within halloysite decreases with the development of weathering, and as a result halloysite changes its form in a sequence of 10Å halloysite→(7Å+10Å)halloysite→7Å halloysite. (2) The Fe³⁺ content in halloysite is within the range of 9.9%～28.8%, and the average SiO₂/Al₂O₃ ratio is 1.62, belonging to iron rich halloysite. (3) Based on final products of weathering, it is postulated that the succession of volcanic eruption in this area was southeast Leihuling→north Ruhongcun village→east Ruhongcun village. (4) Taken the study of weathering products of volcanic ejecta in Japan as a reference, the eruption time of volcano in this area can be inferred to be 6～12ka BP.

Fuzhou basin is tectonically located in the eastern part of South China fold system, within the Fuding-Yunxiao Fault depression zone of East Fujian volcanic fault depression zone. It is a fault depression basin developed in the middle and late Quaternary, filled with marine, terrestrial and alternating marine-terrestrial facies sediments with complicated structures. Two test holes(SZK-1 and SZK-2)were drilled in the Gutian Street, Fuzhou City, with the purpose of setting up a standard section of Quaternary strata in Fuzhou basin. The SZK-1 test hole is 56.3m deep, in which the thickness of the Quaternary sediments is 51.1m. From the cores of this borehole, 25 samples were collected for thermoluminescent(TL)dating and 73 samples were collected for sporopollen analysis. The SZK-2 test hole is 53.2m deep, in which the thickness of the Quaternary sediments is 50.9m. From the cores of this test hole 18 samples were collected for TL dating. The distance of these two test holes is 113.5m. Basically, these two boreholes reveal the same features of deposition. The TL dating results of samples from these two test holes show good vertical time sequence and lateral correlation. The relative ages of sediments dated by sporopollen analysis in SZK-1 test hole show a good consistency with the ages dated by TL method. This study has led to some new insights about the sedimentary facies and succession, as well as the paleo-environment of the Fuzhou basin:(1)The sedimentation in Fuzhou basin initiated at middle Epipleistocene, at least at 54.3ka B.P., and terminated after 1ka B.P.(2)The ages of sediments from the two test holes can be assigned to Epipleistocene and Holocene. The Epipleistocene sediments can be subdivided into the sediments of middle and late Epipleistocene, i.e. 54.3ka B.P.～24 ka B.P. and 24 ka B.P.～ 11.3 ka B.P. Moreover, the Holocene sediments include the sediments of early, middle and late Holocene, i.e. 9ka B.P. to later than 1ka B.P.(3)The strata in the two test holes can be divided into three sections. The lower section corresponds to middle Epipleistocene, the middle section to late Epipleistocene, and the upper section to Holocene.(4)The sediments of middle Epipleistocene are composed of celadon gravel, coarse sand and clay, indicating a temperate and dry climate, while the sedimentary facies of the sediments can be assigned to lacustrine and lakeshore facies. The sediments of late Epipleistocene are composed of gray and yellow, coarse and medium sand, indicating a temperate and dry climate, while the sedimentary facies of the sediments can be assigned to fluvial facies. The early Holocene sediments consist of gray medium and coarse sand with a small amount of sludge, indicating a temperate and relatively wet climate, and littoral deposition environment. The middle Holocene sediments are composed of gray interbedded sludge and fine silt, indicating a warm and wet climate, and shallow-sea deposition environment. The late Holocene sediments consist of gray medium-coarse sand, sludge and clay, indicating a transition from littoral, neritic to lacustrine environment, and a warm and wet climate.

The results of physical modeling for plastic-flow waves in the lower lithosphere indicate that in addition to "fast" waves, which roughly correspond to the plastic-flow waves with velocities of 10⁰～10²km/a in the prototype, the "slow" waves with significantly lower velocities also exist in the model compressed at its driving boundary. The experimental procedure and similarity principle for the "slow" waves are essentially the same with those for the "fast" waves (Wang et al., 2001). The measuring markers placed on the model's surface are arranged along a longitudinal line with constant initial spacing of about 10mm. The "slow" waves are also decomposed into major and subsidiary waves, and both of them are viscous gravity waves: the major wave is similar to solitary wave or surge, and the subsidiary wave is manifested as wave group. Note that the "slow" waves described here are mainly induced by the "fast" waves when they arrive somewhere with a distance X₀ away from the boundary, where X₀ is called wave-generating distance. The value of X₀, as shown in Fig. 6, tends to get larger with increase of the nominal relaxation number , a ratio of the duration of test to the relaxation time of media in the ductile layer. As a matter of fact, it is impossible for "fast" waves to induce the "slow" waves when the distance X₀ required is too large and the wave-propagation range is limited. On the contrary, the "slow" waves can be considered as the waves originated at plate boundary as the distance X₀ is approximate to zero. As inferred from the models according to the similarity principle, the "slow" waves are associated with some slow processes of tectonic deformation in the geological history of the prototype, which are characterized in orders of magnitude by wave velocities of 10^-1～10⁰m/a, time intervals of 10⁰Ma and spatial spans of 10²～10³km. They control the long-term fluctuation and migration of seismic activities and influence the evolution of tectonic movements. The horizontal strains (along longitudinal axis) Δε and vertical strains Δε_z are estimated in terms of the measurements of horizontal and vertical displacements of the measuring markers, while the Poisson's ratio and volumetric strain increment of each marker pair can be calculated for each time step. As a result, the average Poisson's ratio υ =0.465 (approximate to 0.5) and the average volumetric strain increment Δε_v=-0.002 7 (its absolute value is less than the strain measuring error of 0.002 9) are obtained, indicating the volume of the model without considerable variation. It means that the strain response along the horizontal direction is mainly transformed into the variations of the thickness of the layer and the elevation of the model's surface, and it is therefore confirmed that the "slow" waves are similar to viscous gravity waves. The strain rates in the lower lithosphere inferred from the models are 10^-15～10^-14/s in orders of magnitude, which are comparable to those in tectonically active regions. However, the pushing velocities of the driving boundary inferred from the models, 4.1～12.9m/a, are significantly greater than those in the prototype, for instance, about 0.05 m/a for the Himalayan driving boundary. It implies that the physical modeling stated in this paper is qualitative or semi-quantitative.

In order to study the propagation processes of plastic-flow waves in the lithosphere, plasticized rosin, i.e. the mixture of rosin (P) and plasticizer (P) with proper ratio of P/R, is used as analogue material for modeling ductile layer in the lithosphere. The mixture is poured into a rectangular shallow trough (280～300mm long and 190～200mm wide), forming ductile one layer model (1) with a thickness of 8～9mm, as shown in Fig. 1. The bottom (2) and walls (3 and 4) of the trough are made up of glass plates. Some models are covered by a 0.1～0.2mm-thick brittle layer composed of dried consolidated talc-powder slurry, forming brittle/ductile two-layer models. The driving boundary of the model, a movable wall (4) of the trough, is pushed by the springs (6), which are compressed by screw-pushers (7) (micrometer screws) or adding spacers. A number of measuring points (8) are placed on the models surface and their displacements are measured by using a coordinate-system micrometer with the horizontal accuracy of 0.002mm and vertical accuracy of 0.01mm. The tests are carried out at constant temperature with the errors within the range of ±0.5℃ for most of the models. According to the similarity criterion and considering the Stokes number, S_t(the ratio of viscous force to gravitational force) for viscous flow, the time ratio of prototype to model, K_t, is calculated from K_t=K_η/(K_ρ,K_g,K_L), and hence the velocity ratio K_V=K_L/K_t, where K_ηK_ρK_g and K_L are the ratios for viscosity, density, gravitational acceleration and length, respectively. The results of the experiments indicate that plastic-flow waves, being similar to gravity waves in viscous media, include "fast" and "slow" waves and both of them are the superposition of major and subsidiary waves. The major wave is similar to solitary wave or surge. The "fast" waves, including the major and subsidiary ones, are originated from the boundary of the model. The periods of them depend mainly on the pulsative driving period at the boundary, while the wave velocities and strain rates of them are not associated with the pushing velocity of the boundary. In terms of the theory of similarity, the velocities of the major "fast" waves inferred from the models are about 0.12～2.5km/a, corresponding or close in the orders of magnitude to those of some plastic-flow waves in the lower lithosphere, which control the migration of seismic activities (Wang et al., 1994). The major and subsidiary "fast" waves decay in general tendency with their propagation, showing the descend of strain rates, whereas they ascend again locally in a certain range of distances. As is inferred from the models, for instance, the strain rates ascend again in the range of 1 500～2 250km away from the driving boundary for the situation of the waves propagating in the lower lithosphere. Its upper limit corresponds roughly to the distance from the Himalayan collision boundary to the North China Plain. The strong seismic activities in the North China region may be associated, as one of the important factors, with the local ascent of strain rates in the propagation process of plastic-flow waves. Although what has been done in the experimental studies so far stays with qualitative or semi-quantitative simulation, the results of the physical modeling stated above have provided the powerful experimental evidences for understanding the generation and propagation of the plastic-flow waves in continental plate, which control the migration and fluctuation of seismic activities.

The interpretaition of satellite images reveals distinctly the existence of 4 arcuate fault zones (the Maomaoshan-Nanhuashan-Liupanshan,Xiangshan-Tianjingshan,Yantongshan and Niushou~shan Luoshan fault zones ) at the northeastern corner of the Tibetan plateau ,as well as the existence of annular tectonics in southern Ordos. Both the kinetic features of the arcuate structures and the configuration of annular tectonics in southern Ordos are analysed to explore the origin of these fault zones. It is suggested that the subduction of India plate in NNE direction has given rise to the extrusion of the Tibetan plateau in NEE direction. As this NEE-trending extrusion is blocked by the annular tectonics in southern Ordos, 4 arcuate fault zones, converging toward Guyuan and Jingyuan towns and diverging in NWW and NW directions, are fomed at the northeastern corner of the Tibetan plateau. This condition has caused the sinistral strike slip motion along the NWW-and NNW-trending faults,and the dextral strike-slip motion along the NW-or S-N-trending faults at the diverging portion,as well as the compressional thrusting along the faults at the converging portion.