Gravity anomalies occurred at Lufeng, Bomei, Nantang, Jiazi, Xixi and other measuring points in the Guangdong Shantou mobile gravity measurement network before and after the Lufeng M_L4.3 earthquake on September 24^th, 2015, and the area was named as Lufeng gravity anomaly area, where the accumulated change of gravity in two years was greater than three times of the RMS errors of the observations and medium-term decline and reverse turning rise appeared. Meanwhile, gravity anomalies also appeared in the measurement points of Chendian and Heping, so the area was named as Chendian gravity anomaly area, where the gravity showed continuous monotonic increase. The two gravity anomaly areas were adjacent to each other, but the nature of the gravity anomalies might be different, there might be seismic gravity anomalies and ground subsidence gravity anomalies. In order to analyze the development trend of later earthquakes, it is necessary to determine the nature of gravity anomaly. The method of data analysis and field verification was used to distinguish the nature of gravity changes in each gravity anomaly area. The results mainly show that: 1)the Lufeng gravity anomaly area is of seismic gravity anomaly, while the Chendian gravity anomaly area is of ground subsidence gravity anomaly. Through the characteristic analysis of seismic gravity anomalies and ground subsidence gravity anomalies, we had a better understanding of the correlation between seismic gravity field evolution and seismic development, which helped to extract the precursor information of seismic gravity changes and predict earthquake. 2)The gravity changes at the observation points of Lufeng, Bomei, Nantang, Jiazi and Xixi in Lufengthe gravity anomalies area began to decline synchronously in August 2014 and rose synchronously in August 2015, forming a positive gravity anomaly area with regional synchronous decline and reverse turning rise. Through anomaly investigation and verification, no interference source to the observation environment was found, and the Lufeng gravity anomaly area was of seismic gravity anomalies. It was the manifestation of the evolution process of the seismic gravity field during the preparation of the Lufeng M_L4.3 earthquake on September 24, 2015. 3)The gravity changes at Heping and Chendian observation points in Chendian gravity anomalies showed a continuous monotonic increase resulting from the pumping of a large amount of groundwater in Chendian and Heping, which led to ground subsidence, house and ground cracking, so the gravity anomaly was related to land subsidence. It is determined that the nature of the anomaly was land subsidence gravity anomaly. The gravity change caused by ground subsidence at Chendian from March 1995 to July 2016 was 292μgal, and the gravity change caused by ground subsidence at Heping from March 2006 to July 2016 was 137μgal. In the analysis of seismic gravity anomalies, the gravity changes at Heping and Chendian should deduct the gravity changes caused by ground subsidence. 4)For obvious gravity anomalies, the background conditions of the anomalies should be understood in detail, such as geological conditions, the use of domestic water and industrial water, ground subsidence, ground fissures, house fissures, etc., and the source of the anomaly should be found. It is necessary to collect water level observation data and hydrological observation data in gravity anomaly area for trend analysis. 5)According to the comparative analysis of the characteristics of time series gravity variation curve of the gravity anomaly points in the gravity anomaly area, the gravity anomalies of the observation points in Chendian gravity anomaly area showed a long-term continuous monotonic increase, while that in the Lufeng gravity anomaly area showed a medium-term decline and reverse turning rise. By analyzing the gravity variation of the adjacent observation points in the abnormal area, and analyzing the difference of the geological conditions and the field survey data, we can basically judge the nature of the gravity anomalies.

The Xijiang Fault is an important NW-trending fault with a length of~200km, located in the western part of the Pearl River Delta. A M4 3/4 earthquake occurred at the northern end of the fault(Sihui)in 1445 and a magnitude 5 earthquake occurred at the southern end of the fault(Modaomen Waters)in 1905. Heshan is the boundary between the southern and the northern segments of this fault. The southern segment which is called Heshan-Modaomen segment is mainly hidden faults. The activity of Heshan-Modaomen segment remains controversial due to the lack of systematic studies for the deep and shallow exploration, which affects the assessment and prevention of earthquake disaster risk. In this paper, we concentrate particularly on the distribution and activity of Heshan-Modaomen segment using seismic geological surveys, shallow seismic exploration, joint borehole profile detection, and Quaternary geochronology.
Field geological surveys show that the fault zone is prominently normal sinistral strike-slip faults, striking about N310°~330°W, with a width of 10~20m. Most of them dip northeast at angles of 60°~80°. Observations on typical outcrop show that cataclasite, breccias and siliceous rocks are developed on the faults. Fault planes often have smooth and polished surfaces and no fault geomorphology has been developed along the fault zone. The overlying eluvial weathered soil materials have not been disturbed or cut. We carried out shallow cross-fault sounding of 7 profiles in the hidden section of the fault zone using longitudinal wave reflection method of multifold coverage observation system. As a result, we obtained the reflection time sections of the target stratum and the main structure. A total of 13 breaking points to be investigated were explained. We also performed cross-fault drilling at the location of the seismic data interpretation profile and catalogued drilling cores. ¹⁴C and OSL samples were collected systematically. The ¹⁴C dating was performed by the BETA Laboratory in the United States and 16 valid age data were obtained. OSL dating was performed by the OSL Laboratory of China University of Geosciences(Wuhan)and 6 age data were obtained.
This paper presents the study results of two representative cross-fault profiles. The shallow exploration survey line XJ1 and the row drill profile P1 are located in the southern section of the fault where six boreholes are arranged. We find the existence of bedrock faults on the joint borehole profile. The grooves developed thereupon are filled with the late Pleistocene paleochannel deposits with no obvious faults observed. The overlying Holocene strata are horizontal and continuous, without cutting and disturbance. Combined with the stratigraphic age, we infer that the fault has been inactive for at least about 11 000 years. The shallow exploration survey line XJ2 and row drill profile P3 are located in the northern section of the fault, where a total of seven boreholes are arranged. The borehole sections reveal the existence of fault crushed zone in the underlying bedrock(Cambrian hornstone). The tectonites are mainly fault breccias and cataclastic rocks with chlorite alteration. Groove landforms are formed along the fault zone with strong erosion at the later stage, and filling and accumulation occurred since the Holocene transgression with no fracture cutting or stratum disturbance. According to the landform, the occurrence of faults and the development of transverse active faults, the Heshan-Modaomen segment of Xijiang Fault can be further divided into two segments with the boundary of Zhupai Island. Both of them have been inactive since the Holocene.

This paper collects 43 225 absolute first arrival P wave arrival times and 422 956 high quality relative P arrival times of 6 390 events occurring in Yangjiang and its adjacent area from Jan. 1990 to Aug, 2019. These seismic data is recorded by 49 stations from Guangdong seismic network, Guangxi seismic network and Hainan seismic network. Based on the seismic data above, we simultaneously determine the crustal 3-D P wave velocity structure and the hypocenter parameters of 6 255 events in Yangjiang and its adjacent area by applying double-difference seismic tomography. The result shows that shallow P wave velocity in Yangjiang area is higher due to the thinner sedimentary layer and widely exposed Yanshanian granite, Indosinian granite and Cambrian metamorphic rocks. There are obvious correspondences between the distribution of shallow velocity and fault structure as well as geological structure. The velocity transfer zone along Mashui-Pubai correlates with the NE strike of Yangchun-Zhilong Fault, and the low velocity anomaly on the west corresponds to the Yanshanian granite system, while the high velocity anomaly on the east corresponds to the Cambrian epimetamorphic rock system. The Yangjiang M6.4 earthquake locates at the high velocity seismogenic body among the low velocity anomalies due to Yanshanian granite system. Besides, there is a low velocity anomaly existing below the high velocity seismogenic body as mentioned above, we speculate the low velocity anomaly is a ductile shear zone due to partial melting of lower and middle crust caused by mantle wedge melting and basaltic underplating. Moreover, a wide range of low velocity anomaly exists in 20km depth, which verifies the low velocity layer in the middle crust at Yangjiang area of South China continent. The velocity image from land to ocean in 30km depth shows low velocity in NW side and high velocity in SE side, which verifies the characteristic of crust thinning in South China coastal continent. The NEE seismic belt from Yangbianhai to Pinggang is speculated to locate in a buried fault in the southwest segment of Pinggang Fault. The buried thrust fault is a N78°E strike fault, dipping to NW with a dip angle of 85°. In addition, the buried fault locates in the abnormal junction of high velocity on the NW side and low velocity on the SE side, which reflects the tectonic activity characteristic of NW plate uplifting and SE plate declining from Miocene period. The characteristic of activity in the buried fault shows thrust movement with a small strike-slip component, which is consistent with the focal mechanism of the M4.9 earthquake occurring in 2004. Finally, there is a large difference of formation occurrence between the southwest buried fault of Pinggang Fault and the northeast segment of Pinggang Fault. We speculate that, the formation occurrence of Pinggang Fault changes near Pinggang area in the form of “dough-twist”, which causes different velocity structure and movement characteristic.

The study is based on the underground fluid observation data in Lijiang area, northwest Yunnan Province. The data include the water level and temperature in Dangxiao well and Jinjia well, and the ion measurements in Ganze spring. Combining with the data of regional hydrogeology, rainfall, well structures, and the geothermal gradient, we analyzed the variations of each measurement item before the Ludian M_S6.5 earthquake on August 3, 2014 and discussed the possible mechanism for the abnormal variations. The water levels of both Dangxiao well and Jinjia well are influenced by local rainfall, but the former shows hysteresis according to rainy seasons and is the long trend influence; while the latter shows synchronization between high water level and rainy season, indicating good connection between well water and shallow aquifer. The recharge water for Dangxiao well is in relatively low temperature, and the temperature sensor is located at the major connecting section between the well water and the aquifer; the water temperature variation is mainly affected by the discharge status and variation of water level. The Jinjia well is always in static level, and the temperature sensor is below the major connecting section between the well water and aquifer, so the water temperature is affected little by water level variations and in smooth fluctuation. The recharge source for Ganze spring can generally increase the contents of calcium and magnesium ions, so does the conductivity. The water level data of Dangxiao well since 2012 are decomposed with wavelet technique. The results, excluding such high-frequency components as the noise and the semidiurnal and daily wave components influenced by earth tide, are further processed with difference method in order to eliminate the trend effect. The results show that the relative change of water level is enhanced and in relatively rapid increase before the Ludian M_S6.5 earthquake; the corresponding water temperature values are high. The tendency of water level in Jinjia well displays descending, while the corresponding water temperature shows ascending. The content of calcium ion, magnesium ion, bicarbonate ion, and conductivity of Ganze spring are descending, while the content of fluoride ion is ascending. The abnormal variations of underground fluid in Lijiang area appeared in turns and were accompanied with minor earthquakes before Ludian M_S6.5 earthquake, which indicates enhancing of regional stress and increasing of fluid activity.

Studies of air relative humidity dynamic process in Longmenshan Mountains area, Sichuan, southwestern China show that, before the 2013 Lushan M_S7.0 earthquake and the 2008 Wenchuan M_S8.0 earthquake, the annual frequency of air relative humidity anomalies appeared to decrease year by year in the initial period, and then increased quickly. The fall time is longer, often more than 10 years, and the rise time is short, about 1～2 years. The epicenters are located in or near the area where the anomaly frequency of air relative humidity declined most and increased more than other regions.
The tendency transition of the annual frequency of air relative humidity anomalies in the middle-term period of earthquake preparation is probably due to the opening and closing of rock fracture, underground fluid movement and geothermal energy release, which cause the change of ground temperature and latent heat exchange rate. In the medium and long-term period of earthquake preparation(10 to several years), the crust rock is under compression deformation, the rock pores and fissures are continuously closed or reduced, the releasing of underground hot water or hot vapor decreases, and the latent heat exchange rate has also decreased, therefore the relative humidity anomaly frequency shows a declining trend. Whereas in the medium and short-term period of the earthquake preparation, as the crustal rock deformation increases further, which may lead to micro fractures expanding, and the underground hot water and hot vapor by releasing will turn from decreasing to increasing, the latent heat exchange rate changes from reducing to rapidly increasing, and the relative humidity anomaly frequency also shows a transition from decreasing year by year to a rapid increase in burst.

Domestic and foreign scholars found that SLHF anomaly occurred before some strong earthquakes in coastal area or in inland region with abundant water. The author noted that SLHF anomaly in some special tectonic position is not only associated with the strong earthquakes nearby but also with the strong earthquakes far away in the same tectonic belt. The dynamic variation rule of SLHF anomaly in the special tectonic position prior to earthquakes might reveal some trend of regional seismic activity and contribute to the research of mechanism of SLHF anomaly prior to earthquakes. Tengchong area in Yunnan Province is such a special tectonic position located in the south of the eastern Himalayan syntaxis and is one of the youngest volcanic regions on the Chinese mainland with frequent magmatic activity and strong hydrothermal activity.The earthquakes (M_S≥6.4) since 2000 in Sichuan-Yunnan region(including the Yao'an M_S6.5 earthquake in 2000, the Ning'er M_S6.4 earthquake in 2007, the Wenchuan M_S 8.0 earthquake in 2008, the M_S7.0 Myanmar earthquake in 2011 and the Lushan M_S7.0 earthquake in 2013)are selected as cases to study the variation characteristics of the SLHF in Tengchong region before and after the above earthquakes.It is found that significant SLHF anomalies occurred in Tengchong area not only before earthquakes happening in the vicinity but also earthquakes far away in the surroundings. The SLHF anomalies occurred mostly within one and half months before the earthquakes, except Wenchuan M_S 8.0 earthquake. The SLHF anomalies occurred two months before Wenchuan earthquake, the earlier occurrence of the anomalies might be due to the relatively long distance between Wenchuan earthquake and Tengchong area. It is also found that the amplitude of anomaly is related to earthquake magnitude, the larger the magnitude of earthquake, the larger the amplitude of the anomalies. For example, the SLHF anomalies in Tengchong region prior to Burma M_S 7 earthquake and Wenchuan M_S 8.0 earthquake are very strong, far more than the reference maximum values, but the SLHF anomalies prior to Yao'an M_S 6.5 earthquake and Ning'er M_S 6.4 earthquake are relatively small.The sensitive response of SLHF in Tengchong region to the surrounding earthquakes on the one hand is related to the development of active faults and the present-day intensive crustal deformation;on the other hand, there are strong volcanic activities in Tengchong area, with springs widely developed and water heat exchanging quickly. So as a result, the SLHF of this region is sensitive to surrounding tectonic movement.

With the Wenchuan and Lushan epicenters as the center, 8 wells were selected to make a comparative analysis of coseismic variations of well-water level caused by the two earthquakes. Emphasis was laid on the three wells, namely, the Huajiang well in Rongchang, the Liuyin well in Beibei, and the Luguhu well in Sichuan Province, where reverse coseismic variations of water-level were observed during the two earthquakes. We further collected the coseismic variations of water-level data of the three wells caused by two remote earthquakes, the Japan M_S9.0 earthquake in March 11, 2011 and the Sumatra northern sea area M_S8.6 earthquake in April 11, 2012. We also computed the coseismic volumetric strain caused by Wenchuan and Lushan earthquakes respectively in order to interpret the mechanisms. During the Wenchuan earthquake, the coseismic variations of water-level of the Huajiang well in Rongchang, the Liuyin well in Beibei, and the Luguhu well coincided with the coseismic volumetric strain at the wells, that is, the wells located in the compressed area of coseismic volumetric strain showed ascending in the coseismic variations of water-level, while the wells located in the dilatational area of the coseismic volumetric strain showed descending in the coseismic variations of water-level. During the Lushan earthquake, the coseismic volumetric strain and the coseismic water-level variations of the three wells were inconsistent, which are similar to the coseismic variations caused by the Japan and Indonesia earthquakes.

Based on ANSYS parallel software platform, and according to active tectonic block region division and distribution of active faults in North China, and combined with GPS data, the range of the numerical model is defined as 99.8°～121.4°E, 27.9°～42.3°N, which contains a majority of the North China active tectonic block region and parts of other block regions including the Tibetan plateau, the Xiyu, South China, and the Northeastern Asia. The model is divided into 416582 elements whose average side length is 25km with 582392 nodes. The main research results are: (1)Simulation of crustal movement velocity and analysis. The results show that the velocity of crustal movement in North China as a whole decreases from east to west and increases from north to south. It is an almost match between simulation results and GPS observed velocity field. (2)Simulation of the slip of faults and analysis. Considering all known late Quaternary active faults in North China in the model and according to the simulation results, the slip of faults obtained from simulation and that from geological survey are almost consistent. (3)Simulation of strain fields and analysis. By numerical simulation, the minimum strain and the maximum principal strain in North China in 1999-2004 and 2004-2007 are calculated. The horizontal strain direction in North China is in accord with the strain direction obtained from inversion of focal mechanism solutions, GPS observations, etc. by previous studies.

The Xiannüshan Fault zone,lying along the western margin of Huangling anticline,is one of the most important fault zones in Three Gorges reservoir area. The fault experienced strong activities during Cenozoic. Whether the fault zone crosses the Yangtze River is one of the key problems in previous studies,as it has significant influence on the assessment of geological hazards and earthquake stability in the reservoir area. Based on tectonic and geomorphic observations along this fault zone around Baixianchi in Changyang County,Huangkou in Zigui County,together with the comparisons between the geology in Guizhou and Quyuan Town in the north bank of Yangtze River and the Xiannüshan Fault zone,it is suggested that the north end of this fault zone locates around Huangkou village and doesnt traverse the Yangtze River northward.
The details are as follows: ① On the basis of field data collection,it is found that the Xiannüshan Fault zone,which stretches 80km,underwent thrust movement in Cenozoic,resulting in ravines and fault scarps,topographically. Whereas,on the northern bank of Yangtze River,faults are rarely found,and most of the faults are developed in the Jurassic strata,without topographical effects. Therefore,the Xiannüshan Fault zone has not stretched to the north bank of Yangtze River.②Fault gouge and tectonite zone were found developed on Xiannüshan Fault zone at Baixianchi village,but only tectonite zone was found at Zhouping village. There are also some branch faults close to the northern end of the fault zone. So,the activity of the fault zone weakened from south to north in Cenozoic. The fault zone extends northward and dies out at Huangkou. It doesnt stretch forward any longer as indicated by continuous strata,sparse joints,and small folds,etc.

With the advances in simulation techniques and understanding of geodynamic processes,numerical simulation is likely to play an increasingly important role in the research of seismic hazard analysis and earthquake prediction.In this paper,on the basis of the paper "A preliminary study on the application of numerical simulation methods to earthquake prediction research(Ⅰ)",the possible application of uncoordinated deformation analysis,Coulomb stress changes and earthquake probability modeling to the study of earthquake prediction is further discussed.When rock deforms from the elastic into the yield stage,the system is in a critical unstable state,the rock movement may deviate from the normal track and become complicated.The study results show that,before Wenan earthquake(M_S 5.1)on July 4,2006,GPS velocity was well consistent with the numerical simulation speed in most areas of North China,while there were some differences in some regions,especially in the northeast of the North China Plain block,where big inconsistency in movement characteristics occurred,resulting perhaps from the preparation of Wenan earthquake.Research on earthquakes triggered by Coulomb stress change is a focus problem now.Numerical simulation may play an important role in the analysis of Coulomb stress changes.By constructing three-dimensional dynamic model,the effect of various factors on the value and distribution of Coulomb stress change can be simulated,and more realistic results can be obtained.By numerical simulation of Coulomb stress changes to seismic activities beneath Sichuan Zipingpu reservoirs,it is found that with the increase of reservoir water storage time,the pore pressure diffusion in the effective additional stress field will be gradually expanded to the range of more than 10km underground.The regional effective additional stress field and seismic activities show different characteristics in several typical regions.The United States Southern California Earthquake Center has tried to study the earthquake probability as research objectives.It is worthy of referencing in China's earthquake research.Computer simulation of synthetic earthquake catalog is an effective way to solve the lack of data.The future direction of development should be a more realistic three-dimensional dynamic model,taking into account the multi-field coupling between heat,fluid and etc. ,improving hardware and software conditions and shortening the calculation time step,obtaining more complete information on fault movement,and simulating the fault activities.

Earthquake preparation and occurrence is a complex physical process.Although the earthquake abnormalities are varied,the strain energy accumulation is requisite before an earthquake.Earthquake prediction analysis must consider the strain energy accumulation process.As hard to go into the Earth's interior,direct measurement of stress and strain in deep focus is very difficulty.The use of numerical analysis,which constructs three-dimensional dynamic models of the crust and upper mantle to simulate the rock deformation process,is currently one of the most effective methods to study the crustal energy transfer and accumulation.The simulation result of current crustal deformation is consistent with the existing GPS data around the Eastern Himalayan Syntaxis and its surrounding areas,in that the crustal horizontal displacement field of the eastern Tibetan Plateau rotates clockwise around the Eastern Himalayan Syntaxis.Current effective stress concentration areas mainly distribute along the block boundary fault belts around the Eastern Himalayan Syntaxis,especially along the southeast section of Jiali Fault,Moto Fault,Apalong Fault,India-Myanmar subduction zone and the Sichuan-Yunnan border region.It should be noted the risk of future strong earthquakes in these areas.In the adjacent interconnected tectonic areas,the blocks and faults are interrelated and interacted each other.When an earthquake occurs in a region,the rapid displacement and deformation of rock will inevitably lead to displacement and deformation of the associated blocks and faults; strain energy will transfer from one region to others.The numerical simulation results of deformation process in the Capital area from 1989 to 1998 clearly show that the high strain energy concentration region shifted from Datong area where 1989 earthquake(M_S 5.8)occurred to Zhangbei area where 1998 earthquake happened.It illustrates that the application of numerical simulation analysis method may help us predict the possible strain energy transfer process,thus,providing the reference target regions for earthquake monitoring.

We select 15 earthquakes with M_S≥6.0 in Xinjiang since 1970 as "source earthquakes",and aftershocks of M_S≥4.0 as target aftershocks.Test analysis has been done on the static stress triggering model.The results show that the static stress triggering model is not so applicable in the Xinjiang region.For 80％of source earthquakes,the number of target aftershocks in positive ΔCFS area is less than that in negative area; for 33％of source earthquakes,the number of target aftershocks in positive ΔCFS area is far less than that in negative area(the former is less than half of the latter); only for 13.3％of the source earthquakes,the number of target aftershocks in positive ΔCFS area is far more than that in negative area(the former is twice of the latter).Even the uncertainties are considered,e.g.the focal depth of the source earthquakes,the fault plane orientation and the slip angle,the results are basically the same.In most shock events,target earthquakes in negative ΔCFS area are more in number than that occurred in the positive area,which does not accord with static stress triggering model.The further inference is that the short-term earthquake prediction based on stress and strain increment changes is limited.

The spatial-temporal variations of atmospheric water vapor in western Sichuan Province and its vicinity during the mid-and long-term earthquake preparation process are studied.The M_S 8.0 Wenchuan earthquake in 2008,the Songpan-Pingwu earthquake sequence in 1976 and the Yushu M_S 7.1 earthquake in 2010 are selected as cases of the study regions.The result shows the frequency of atmospheric water vapor anomalies will decrease at first, then increase quickly during the mid-and long-term process of strong earthquake.The decreasing of the frequency of atmospheric water vapor anomalies begins relatively early,mostly over 10 years before earthquake and will last more than 9 years,which is considered as mid-and long-term precursor.The rapid increase appears 4 years before earthquake,as a kind of mid-and short-term precursor.The frequency of atmospheric water vapor anomalies over epicentral region of Wenchuan earthquake began to decline slowly 18 years before the earthquake and continued for 15 years, then increased rapidly 2 years before the earthquake.The atmospheric water vapor anomalies in the epicenter area changed from lowest frequency to high frequency 1 year before the earthquake,then the event occurred.It continued to increase until 2009 then resumed to normal state.14 years before Songpan-Pingwu earthquake,the frequency of atmospheric water vapor anomalies over the epicenter region began declining and it continued for 9 years,and then increased 4 years before the event.The anomalies in the epicenter region changed from lowest frequency in 1971 to high frequency in 1973,and reached the highest in 1976,and then the main shock happened.After the earthquake,it resumed to normal state.Similarly to the above two cases,there had been atmospheric water vapor anomalies before the M_S 7.1 Yushu earthquake in 2010,the frequency of anomalies declined from 1997 to 2008 when the lowest value was reached.It increased quickly in 2009 till the event occurred in 2010.The atmospheric water vapor anomalies over epicentral region may be due to the opening-closing movement of pores and fractures in the rock layer before the earthquake,resulting in the migration of underground fluid and underground heat energy,and then causing the change of the surface temperature and surface latent heat flux.During the mid-and long-term process of earthquake preparation,the rate of latent heat exchange decreases due to the reduction of the hot water vapor from underground caused by the closing of pore and fracture when the crustal rocks undergo compression deformation,so the frequency of atmospheric water vapor anomalies begins to decline.While during the mid-and short-term process of earthquake preparation,the accelerating of the crustal rocks deformation and the expanding of micro fractures will lead to increasing the hot water vapor from underground,accelerating the latent heat exchange,and quickly increasing the frequency of atmospheric water vapor anomalies.Analyses of the above three cases prove preliminarily that this assumption is reasonable.

The paper summarizes the application and development of finite element numerical simulation methods in the seismo-geological study in North China in the last 30 years,and the discussion is focused on the following three aspects:the finite element model for North China,actual data used in the simulation and numerical simulation results.The study of finite element numerical simulation in North China began in 1980,since then,it has experienced the phases of development from elastic to elastoplastic,viscoelastic and creep simulations,from 2D to 3D,from linear to nonlinear,from continuous deformation to discontinuous deformation,and from the use of individual data to the use of multiple synthetic data.It shows that finite element numerical simulation is an effective means for studying seismology and geology in North China,and distinguished results have been achieved by using simulation methods.The achievement will be the foundation for the future simulation study in North China.The paper also discusses the significance,existing problems and development tendency of finite element numerical simulation in seismo-geology in North China.finite element method,numerical simulation,North China.

In 1585,an earthquake with M_S5³/4 happened in the south of Chaoxian of Anhui Province,and the parameters of this earthquake listed in earthquake catalogues of varied versions are different.According to the detail textual research of the historical earthquake records,the epicenter location of this earthquake is further confirmed by means of field seismo-geological investigation in the Chaohu-Tongling region of the western Changjiang valley.Shallow seismic prospecting and drilling methods are applied to study the buried fault,the possibility of existence of seismogenic fault and fault activity in the western Changjiang valley area are analyzed in depth,and the causative tectonic background of the 1585 M_S5³/4 south Chaoxian earthquake is studied.The result of this study indicates that the Yanjiaqiao-Fengshahu Fault,which was active in early Pleistocene to mid-Pleistocene,is possibly the causative structure of this earthquake.To identify the seismogenic structure of the 1585 south Chaoxian earthquake is helpful to deeply know the tectonic background of the moderate and small earthquake activities in eastern China.

Research shows that anomalous increase of latent heat flux was discovered before several earthquakes.Basing on the Modis data,high-spectral data and ground meteorological data of the working area,we retrieved the sensible heat flux and calculated the latent heat flux from surface energy balance equation using a simplified two-layer model(dualtemperature-difference method).In this paper,we also discussed the characteristics of this model,and retrieved the latent heat flux of the area where the Longmenshan Fault goes through.Finally,we found that there were significantly high values of latent heat flux in the epicenter area of the M_S 8.0 Wenchuan earthquake and the surrounding area of Longmenshan Fault 48 hours before the earthquake,the latent heat flux on May 12 reached the maximum of 300～400W/m²,which might be the anomalous increase of latent heat flux before the M_S 8.0 Wenchuan earthquake.

The paper studies the coseismic changes of water level and water temperature caused by the M_S 8.0 Wenchuan earthquake.The differences of water level and water temperature variations caused by the M_S 8.5 Sumatra earthquake on Sep.12,2007 and the M_S 8.0 Wenchuan earthquake on May 12,2008 are analyzed.The result shows that the well water level change caused by Wenchuan earthquake is dominated by rising.Spatial distribution of wells with water level rising or descending exhibits regional difference.The proportion of wells is higher with water level and water temperature changes in the same directions than in the reverse direction.Water temperature mainly dropped when water level fluctuated.Compared to remote earthquake,near earthquake caused bigger changes in water level and water temperature in wells.Coseismic changes of all the well water level and most of water temperature kept the same direction regardless of distance,magnitude,focal mechanism of earthquakes or epicentral directions,though water temperature direction changes occurred only at some peculiar wells.The water temperature direction changes were caused by changes of artesian condition and water level response from fluctuations to steps.The direction of water level changes might be controlled mainly by both local geological structures and hydro-geological conditions of the well.However,the direction of water temperature relates with mixing of water in the well,the location of the water temperature probe and other factors.The mechanisms of water temperature coseismic change are more complicated.

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.

The relationship between the satellite thermal infrared brightness temperature and ground elevation is very important information that can reflect the thermal state of earth surface.It is also one of the main factors which influence the analysis of satellite thermal infrared anomalies before earthquakes.Previous study is not enough and comprehensive and it needs a deep research.In this paper,the decreasing infrared brightness temperature per 100m increase of the ground elevation is defined as brightness temperature gradient.The mainland of China is divided into 6 regions for separate research:the Northwest,the Qinghai-Tibet,the Sichuan-Yunnan,the Northeast China,the North China and the South China.This paper selects remote sensing data at midnight in order to reduce the influence of solar radiation and preprocesses thermal infrared remote sensing images systematically including cloud and noise removing and so on.On the basis of filtering and processing the satellite thermal infrared remote sensing data,this paper analyses the relationship between the satellite infrared brightness temperature and ground elevation in the 6 regions with the method of regression analysis.Studying on the 6 regions separately,some abecedarian results are developed:1)the general trends of brightness temperature gradient are similar between different areas,but there are some differences in details;2)the brightness temperature gradient of the 6 regions are all less than the air temperature lapse rate(0.64℃/(100m))calculated on the basis of air temperature data,also less than 0.4℃/(100m),from 0.15℃/(100m)to 0.35℃/(100m)in most situations;3)except the Qinghai-Tibet area,the brightness temperature gradient of the other five regions is generally greater in summer than in winter.On the whole,the brightness temperature gradient increases from January to September,reaches the maximum around September,decreases subsequently and gets to the minimum around January.It is considered as the result of the effect of the cold air in winter,the absorption of the atmosphere to ground radiation and the diversity of different geographical environments are the key factors that influence the temporal and spatial variation of brightness temperature gradient.

After Wenchuan Earthquake(M8.0),surface fractures in Beichuan and Yingxiu were investigated.It was found that the earthquake deformation zone strikes in NE-SW.Earthquake fracture is characterized by thrust faulting with small amount of strike-slip movement.The compressional shortening is 3～4m in Beichuan and the left-lateral strike-slip displacement is 0.4～0.5m in Yingxiu.

Some destructions in meizoseismal region of Wenchuan earthquake with magnitude 8,Sichuan,China are shown,and their seismic intensity is determined according to "The Chinese Seismic Intensity Scale".The type of earthquake-generating fault and some features of seismic destruction are discussed briefly.

In the eastern mainland of China there are few cross sections of faults where dislocation of Quaternary strata can be observed. However,recently we found such a profile about 2km away from the Nanling county,Anhui Province(30°55'456″N,118°177'74″E),west to the highway from Nanling to Fanchang. This fault has been identified on the satellite image,but its trace is confined to the southern side of the Nanling Basin. Our field investigation indicates that the northwestern end of this fault lies at the Xiaodanyang-Fangshan Fault. It is only 20km long,striking in NW,dipping to southwest. From observations on the profile,it consists of two small fractures and has two periods of activity at least. The first active period is before middle Pleistocene time,or probably in early Pleistocene. And the second active period is in or after middle Pleistocene. Its latest motion is of thrust with an amount of dislocation of 40cm. This fault cross section shows that the NW-trending faults in the eastern Mainland of China have new activities,though on small scales in general.

The neotectonic movement and characteristics of fault activity in the Anqing-Ma'anshan section of the Changjiang River valley are analyzed on the basis of data obtained from field investigation,shallow seismic prospecting and drilling. The results show that during neotectonic time this river valley section and its both sides as a whole was dominated by weak and intermittent uplift movement. As a consequence,owing to the effect of the activity of NE-and NNE-trending faults,relatively strong vertical differential movement occurred in Wuwei-Anqing area during Paleogene-Neogene,and had continued to early and middle Pleistocene. The NE-NNE-trending and NW-trending faults were developed in the bed rocks of the valley and its both sides. The former was formed during Indo-Chinese epoch,while the later was formed during Yanshan epoch. The most recent active period of the larger faults controlling the development of Cenozoic Basins is middle Pleistocene,while the newest activity of the relatively small faults developed within pre-Cenozoic group is pre-Quaternary. The Quaternary system in the valley is \{10～\}50m thick,consisting mainly of Pleistocene-Holocene deposits. The isopach of these deposits is smoothly distributed,indicating normal valley deposition. Seismic activity along the valley and its both sides is relatively weak,and historically only 4 destructive earthquakes have been recorded. Among these events,the largest one is the M5(3/4) earthquake occurring at Chaohu in 1585,and the other events including one with M5(1/4) and two with M4(3/4). Since the beginning of instrumental records in 1970,the largest magnitude that has been recorded so far is ML 3.7. All these results may provide better constraints on the assessment of the crustal stability for this river valley section.

The Yaoan M_S 6.5 earthquake on Jan.15,2000 and Mani M_S 7.5 earthquake on Nov.8,1997 are taken as cases to study the temporal process of the satellite infrared anomalies on the active faults near the epicenter.The inside-outside temperature relation analysis method(IOTRAM)was used in the middle segment of the Red River Fault.The result shows that the inside temperature is about 0～1.5℃ higher than the outside temperature in the normal periods.But the brightness temperature inside the Red River Fault is 2～3℃ higher than that outside the fault 15 days before the Yaoan M_S 6.5 earthquake on Jan.15,2000.The Mani M_S 7.5 earthquake occurred on Nov.8,1997 in northern Qingzang-Tibetan plateau.Curves of the average brightness temperature in the east section of the Margaichaca fault belt and the difference inside and outside the region of the fault belt in 1996,1997 and 1998 were contrasted.The result shows that temperature difference in Oct.and Nov.in 1997 was2～3℃ higher than that in 1996 and 1998,though the inside temperature in the same term in 1997 was the lowest in the three years because of a heavy snow.It returned to normal after earthquake.IOTRAM was used to study the active faults in Huabei and Chuandian regions where less earthquakes have occurred recently and no infrared anomalies appeared.

The temporal and spatial variations of surface latent heat flux(SLHF)and diagnostic air temperature at 2m before and after the M_S 5.7 earthquake that occurred between Ruichang and Jiujiang in Jiangxi province on the 26th November 2005 are summarized in this paper.It is found that before the earthquake significant SLHF anomalies and air temperature anomalies occurred in the epicentral area and its vicinity.The air temperature anomalies are found from the 2nd to 13th November,2005 and concentrated at the epicentral area and its southern area.Then two days later,that is,from the 4th to 15th November 2005,significant SLHF anomalies occurred in the epicentral area and its northern area where a lot of lakes distribute along the active faults.During the anomalous period the SLHF and air temperature at 2m exceeded the sum of 26 years average and 1.5 times of their standard deviation for the same day.Both anomalies had maintained for 12 days with a peculiar distribution associated with the tectonic active zone.It is considered that both of air temperature anomalies and SLHF anomalies are behaviors of thermal flux from underground prior to earthquake.SLHF anomalies occurred over regions with abundant water,whereas air temperature anomalies occurred over land.

The most remarkable feature of Cenozoic and present-day tectonic deformation of the continental lithosphere of China is that the crust has been cut by huge late Quaternary active faults, forming active crustal blocks of different orders. Various active crustal blocks exhibit different horizontal movement and different deformation styles. The inner part of the active crustal block is relatively stable. Deformation commonly takes places along their boundary structures, and most of the great earthquakes (M≥7) occur along these boundaries. In order to monitor crustal movement in China mainland, China Crustal Movement Observation Network has disposed 25 continuous GPS base stations in the main tectonic units all over the country. These stations had been run for 3 years from March 1999 to December 2001. On 14 Nov. 2001, an earthquake of M_S 8.1 occurred to the west of the Kunlun Mountain Pass. This event has produced a surface rupture zone of more than 350km in length with a general strike of 70°～90°. The rupture zone is dominated by left-lateral strike-slipping, and the largest horizontal displacement is about 6m. The observation data of continuous GPS measurement stations show that various GPS stations in different active blocks around this earthquake site had different responses to the earthquake. The GPS station within the active block where the earthquake occurred, such as the Delingha station, exhibited very obvious displacement. However, no obvious displacement was observed at the GPS stations located in the active blocks that are secluded by one active block from the earthquake site,such as the Lhasa GPS station. If the GPS stations are located on the boundary structures of the active blocks adjacent to the earthquake site, such as the Xiaguan GPS station, then they would record obvious displacements several days after the occurrence of the earthquake. If the stations are located within the active blocks, such as the Xining and Kunming GPS stations, no obvious displacement would be observed. However, no obvious displacements was observed at the Xiaguan GPS station after Burma earthquake (M=7.2) occurred in the north of Burma active block, although the epicentral distance of this earthquake (about 370km) is significantly less than that of the west of Kunlun Mountain Pass earthquake. This can be attributed to the relative small magnitude of the Burma earthquake, which did not cause the compression of the Sichuan-Yunnan active block. This fact may indicate that the deformation on the boundary zone of the active block is obviously stronger than that occurs within the block, and it is independent to the epicentral distance. The difference of the effects of great earthquakes on its adjacent active blocks depends mainly on the mode of action on the adjacent block by the movement of active block where the great earthquake occurs. If the movement of the block results in compression of the adjacent block, then the effect of the earthquake will be obvious, while the movement does not result in compression of the adjacent block, no obvious effect can be recorded by the GPS station in this block, because the effect may rapidly decrease when it passes through the boundary zone of the block. The observation data of the GPS stations in response to great earthquake-demonstrate that more effective monitoring of earthquake related crustal movement can be fulfilled, provided that the GPS stations are reasonably disposed within the active blocks and on their boundary zones.

This paper is aimed at the relationship between satellite infrared anomaly and tectonic activities by analysis of anomaly images of 3 earthquake cases (M>6) occurred in 2000 in China. Yao'an earthquake (M_S6.5) occurred on January 15, 2000 in Yunnan Province, southwest China. The infrared anomaly of the earthquake began 20 days before the mainshock and disappeared after the event. It developed along the Honghe Fault, which is the most important fault near the epicenter. The brightness temperature along the fault is always higher than that away from the fault during the anomalous period. The anomaly of temperature difference along the fault and away from the fault exists not only when weather temperature increases but also when it falls down. It seems that the anomaly is more active along the fault than in the other area. Jingtai earthquake (M_L6.2) occurred on June 6,2000 in Gansu Province, northwest China. Satellite Infrared anomaly appeared in a peculiar geometrical pattern before the earthquake. The monthly average infrared brightness temperature increased abnormally along 3 NE-trending tectonics which intersect with the NW-trending faults. Near the epicenter, the infrared anomaly developed in "Z" configuration that seems to extend along conjugated structures. Xinghai earthquake (M_S6.6) took place on Sept. 12, 2000 in east Qinghai Province, northwest China. The epicenter is located near the NNW-trending Ngola Shan Fault. Before the earthquake, the monthly average infrared brightness temperature was different on both sides of the Ngola Shan Fault and its extension. It was higher than 10℃ on the eastern side, and less then 10℃ on the western side. From the above-mentioned 3 cases, the primary conclusions can be drawn as fellows: (1) The distribution of satellite infrared anomaly is closely related to geologic structures, especially active faults; (2) The earthquake epicenter is often located at the margin of relatively high infrared temperature anomaly area or its vicinity. Infrared anomaly develops asymmetrically, propagating from one side toward the epicenter. (3) The spatial and temporal distribution of infrared anomalies may vary in different areas. In the analysis of infrared anomaly information, therefore, it is necessary to take the tectonic conditions, geographical features and meteorological interferences into consideration.

The Kunlun M_S8.1 earthquake took place on Nov. 14, 2001 in eastern Kunlun Mountains, Northwest China. Two methods are developed to extract the information of the satellite infrared anomalies prior to the earthquake. One is pixel-by-pixel analysis by subtracting the average brightness temperature 1 month before the event from the average brightness temperature in the same period in 2 000 pixel by pixel. The obtained image can be divided into two different regions: temperature increase and non-increase regions. The image shows that most regions in north Tibetan Plateau are temperature non-increase, but some temperature increase belts occur along active faults, and especially along the eastern Kunlun Fault. This temperature increase is considered to be earthquake precursor. The second method is based on the analysis of temperature difference inside and outside the fault zone. The inside region is defined as the buffer region created by GIS within 15km radius to the Hoh Sai Hu segment of the Eastern Kunlun Fault. The outside region is the buffer region in 15～30km radius to the inside region. This method is used to calculate the difference of the average brightness temperature for inside region and outside region in each night(0～8am) in 2001. The result shows that the temperature of the inside region is normally about 2℃ lower than that of the outside region. However, one and half months before the earthquake (beginning from Oct.2001),the brightness temperature of the inside region became 1℃ higher than that of the outside region. It returned to normal value after the event. The results indicate that the satellite infrared anomalies correspond spatially to the seismogenic fault. The temperature increase in the seismogenic fault belt is higher than that in the other regions. The analysis of temperature difference inside and outside the fault zone is an effective method for extracting the precursor information from satellite infrared images.

The spacial relationships of strong earthquakes have been investigated for two catalogues: Chinese historical earthquake catalogue(M≥7) and worldwide earthquake cat-alogue(M≥7) from 1900 to 1980. Three strong earthquakes taking place one after another are labeled as A,B and C. The spacial relationship between A,B and C are revealed through systematical statisical analyses. (1) The preferential distribution of ABC planes is to intersect the upper part of the asthenosphere or the lower part of the lithosphere. (2) After earthquakes A and B,both extending regions from A to B or from B to A,and the place between A and B ((ΔCA-ΔCB)/ΔAB≈-0.2),are more risk regions then the other regions. Based on these spacial relationships,strong earthquakes can be predicted in probability. The 56 groups of Chinese historical strong earthquakes (M≥7) have been calculated. Most of earthquakes (C) in each group took place in the high probability regions.