The neotectonics in Zhanjiang Bay area is almost the inferred faults and there are not any active faults seen on the ground surface. So it is difficult for research on the seismogenic structure. This paper analyzes and interpretes the gravity data that can reflect the feature of deep faults and then discusses the seismogenic structure of Zhanjiang Bay area in combination with its geology and earthquake activity. There is a huge NEE-trending high gravity gradient belt lying in the coastal region among Guangdong, Guangxi, and Hainan, and Zhanjiang Bay is located in this gravity gradient belt. We analyzed and interpreted more than eighty images obtained with many different methods one by one, then, got the result that Zhanjiang Bay area is embraced by two giant fault belts trending in the NEE and NW direction respectively, and its interior is crossed over by the NE-trending fault belt. These three fault belts are well shown in the gravity images, especially the NEE-trending fault belt and NW one. The gravity isolines and gradient belts or the thick black stripes of the NEE-and NW-trending fault belts are displayed apparently. Also, these gravity structures are good in continuity, extend vastly and cut deeply. What is more, the NEE-trending fault belt plays a leading and region-controlling part. It shows good continuity, and cuts off the NW-and NE-trending faults frequently and intensively. The NW-trending fault belt also is good in continuity and cuts the NEE-and NE-trending faults relatively frequently and strongly, but it is restricted by the NEE-trending one. Last, the continuity of the NE-trending fault is worse and the strength cutting off NE-and NW-trending faults is significantly weak, just in some segments and in the shallow positions. According to the characteristics above and combined with the analyses of geological structure and earthquake activity, the conclusion can be drawn that the NEE-trending fault is the controlling structure and the main seismogenic structure in Zhanjiang Bay area, and the NW-trending fault is the second one. They conjugate and act together. Therefore, Zhanjiang Bay has the tectonic condition for generating M_S=6.5 earthquakes.

In this paper,we present a method which allows to calculate the mean stress field according to the total seismic moment released by earthquakes.The exact method is as follows: First,we calculate the scalar seismic moment released by each earthquake according to the statistical relationship between earthquake magnitude and its seismic moment; Second,we calculate the seismic moment tensor released by each earthquake according to the relationship between focal mechanism solution and seismic moment tensor; Then,we can get the total seismic moment tensor released in a specific time period of the study area; Finally,we calculate the eigenvector and eigenvalue of the total seismic moment tensor,the obtained eigenvector is corresponding to the mean stress field direction released by the study area. We tested the method by using the synthetic focal mechanism to which random error was added and with the focal mechanism data of Tangshan aftershock zone.The testing results show that,the released stress field of the study area obtained by our method is in consistency with the regional stress field. So our method can be applied to solve regional stress field.The more focal mechanism data used,the more stable the result would be,and closer to the real regional stress field. One of the advantages of this method is that it uses magnitude as the weight of each earthquake,so the contribution difference of the earthquake size in the stress field inversion can be better reflected. Another advantage is that it does not need to know which nodal plane of the focal mechanism is the real fault plane when we calculate stress field.

In this research,we made a primary research on the Coulomb stress triggering of the March 11,2011 M_W=9.0 Tohoku earthquake sequence by using the software of Coulomb 3.2,the earthquake rupture models obtained by Hayes and Guangfu Shao et al. , and the aftershocks data from Harvard CMT catalogue(Centroid Moment Tensor)and Japan F-net catalogue. Our results suggest that: Firstly,the M_W7.2 foreshock,which occurred on March 9,has stress triggering effect on the M_W9.0 main shock; Secondly,the statistical result of the aftershocks triggered by the main shock shows that different statistical results would be obtained when using different main shock rupture model,aftershock catalogue,equivalent friction coefficient and different nodal plane of focal mechanism as the receive plane. The minimum and maximum triggering rates of main shock to aftershock are 56.8%and 75.3%,respectively; and thirdly,when calculating the Coulomb stress by using earthquake focal mechanism,the shear stress on the two nodal planes is the same theoretically. However,in actual calculation,the shear stress would be different between the two nodal planes,due to the non-orthogonality of the two nodal planes or rounding off decimal places in the focal mechanism results. But,the difference of the shear stress on the two nodal planes is relatively small. More attention should be paid on the selection of receiver fault plane from the two nodal planes,when discussing the stress triggering of one specific earthquake or making statistics of aftershock triggering rate,because the selection of receiver fault plane would have a certain effect on the shear stress,and have greater effect on the normal stress,thus the Coulomb stress would be affected.

The precisely located earthquake catalogue is important to seismicity, seismic tomography and crustal stress inversion studies. It also has great application value in rapid report of an earthquake that just occurred. By considering the arrival time uncertainty, and the constraints on station elevation and seismic depth, we propose a relatively accurate method to estimate hypocentral location and its uncertainty based on inversion theory. Our method can combine the arrival times of Pg wave, Sg wave, Pn wave and Sn wave in hypocenter location, so it increases the location accuracy by involving more data; and it can be also used in local and regional earthquake location simultaneously. In order to test our location method, we located earthquakes by using the simulated data with different uncertainty of Pg,Sg,Pn,Sn arrivals. The result shows that the location determined by using our method is more accurate than that by using other method. We apply it to earthquakes occurring in the period from 2001 to 2008 in Sichuan area, and obtained a more clustered hypocentral distribution convergent to the fault zones. The result provides a solid foundation for studies of seismicity, geometry of the active faults and seismic tomography in Sichuan region. It is also helpful to study the seismicity precursors before the Wenchuan earthquake.

In 1556, seismic shocks occurred in Huaxian, Shanxi Province, and in Yueyang, Hunan Province. Whether the seismic shocks belong to one earthquake? For this purpose, we studied the attenuation of seismic intensity, the similarity of the isoseismal forms of several strong earthquakes in and near Huaxian area, and the W-E-trending Qinling tectonic belt acting as a barrier for the seismic wave propagation. The results show that the boundary of seismic intensity IV of the great Huaxian earthquake lies in Xiangyang-Guangshan zone, near the 23℃ N. That is to say, there is an aseismic belt between Huaxian and Yueyang, and south of the aseismic belt there is another high-seismic intensity area up to VII. For that the epicenter must be near Yueyang, From where the seismic intensity attenuated outward and a closed boundary of intensity contour was formed. It indicates that the earthquake occurring in Yueyang is simultaneous but independent from the Huaxian earthquake in 1556. After then the high seismic frequency in Yueyang area lasted about 20 years. It means a period during which the regional stress was adjusted after the earthquake, whereas the anticlockwise rotational deformation in the upper part of Cisi Pagoda in Yueyang might result from the strong earthquake.