The current and conventional fault-crossing short baseline measurement has a relatively high precision, but its measurement arrays usually fail to or cannot completely span major active fault zones due to the short length of the baselines, which are only tens to 100 meters. GNSS measurement has relatively low resolution on near-fault deformation and hence is not suitable for monitoring those faults with low motion and deformation rates, due to sparse stations and relatively low accuracy of the GNSS observation. We recently built up two experimental sites on the eastern boundary of the active Sichuan-Yunnan block, one crossing the Daqing section of the Zemuhe Fault and the other crossing the Longshu section of the Zhaotong Fault, aiming to test the measurement of near-fault motion and deformation by using fault-crossing arrays of one-kilometer-long baselines. In this paper, from a three-year-long data set we firstly introduce the selection of the sites and the methods of the measurement. We then calculate and analyze the near-field displacement and strain of the two sites by using three hypothetical models, the rigid body, elastic and composed models, proposed by previous researchers. In the rigid body model, we assume that an observed fault is located between two rigid blocks and the observed variances in baseline lengths result from the relative motion of the blocks. In the elastic model, we assume that a fault deforms uniformly within the fault zone over which a baseline array spans, and in the array baselines in different directions may play roles as strainmeters whose observations allow us to calculate three components of near-fault horizontal strain. In the composed model, we assume that both displacement and strain are accumulated within the fault zone that a baseline array spans, and both contribute to the observed variances in baseline lengths. Our results show that, from the rigid body model, variations in horizontal fault-parallel displacement component of the Zemuhe Fault at the Daqing site fluctuate within 3mm without obvious tendencies. The displacement variation in the fault-normal component keeps dropping in 2015 and 2016 with a cumulative decrease of 6mm, reflecting transverse horizontal compression, and it turns to rise slightly(suggesting extension)in 2017. From the elastic model, the variation in horizontal fault-normal strain component of the fault at Daqing shows mainly compression, with an annual variation close to 10^-5, and variations in the other two strain components are at the order of 10^-6. For the Longshu Fault, the rigid-body displacement of the fault varies totally within a few millimeters, but shows a dextral strike-slip tendency that is consistent with the fault motion known from geological investigation, and the observed dextral-slip rate is about 0.7mm/a on average. The fault-parallel strain component of the Longshu Fault is compressional within 2×10^-6, and the fault-normal strain component is mainly extensional. Restricted by the assumption of rigid-body model, we have to ignore homolateral deformation on either side of an observed fault and attribute such deformation to the fault displacement, resulting in an upper limit estimate of the fault displacement. The elastic model emphasizes more the deformation on an observed fault zone and may give us information about localizations of near-fault strain. The results of the two sites from the composed model suggest that it needs caution when using this model due to that big uncertainty would be introduced in solving relevant equations. Level surveying has also been carried out at the meantime at the two sites. The leveling series of the Daqing site fluctuates within 4mm and shows no tendency, meaning little vertical component of fault motion has been observed at this site; while, from the rigid-body model, the fault-normal motion shows transverse-horizontal compression of up to 6mm, indicating that the motion of the Zemuhe Fault at Daqing is dominantly horizontal. The leveling series of the Longshu site shows a variation with amplitude comparable with that observed from the baseline series here, suggesting a minor component of thrust faulting; while the baseline series of the same site do not present tendencies of fault-normal displacement. Since the steep-dip faults at the two sites are dominantly strike-slip in geological time scale, we ignore probable vertical movement temporarily. In addition, lengths of homolateral baselines on either side of the faults change somewhat over time, and this makes us consider the existence of minor faults on either side of the main faults. These probable minor faults may not reach to the surface and have not been identified through geological mapping; they might result in the observed variances in lengths of homolateral baselines, fortunately such variations are small relative to those in fault-crossing baselines. In summary, the fault-crossing measurement using arrays with one-kilometer-long baselines provides us information about near-fault movement and strain, and has a slightly higher resolution relative to current GNSS observation at similar time and space scales, and therefore this geodetic technology will be used until GNSS networks with dense near-fault stations are available in the future.

More than 80 percent of strong earthquakes(M≥7.0)occur in active-tectonic block boundaries in mainland China, and 95 percent of strong earthquake disasters also occur in these boundaries. In recent years, all strong earthquakes(M≥7.0)happened in active-tectonic block boundaries. For instance, 8 strong earthquakes(M≥7.0)occurred on the eastern, western, southern and northern boundaries of the Bayan Har block since 1997. In order to carry out the earthquake prediction research better, especially for the long-term earthquake prediction, the active-tectonic block boundaries have gradually become the key research objects of seismo-geology, geophysics, geodesy and other disciplines. This paper reviews the research results related to seismic activities in mainland China, as well as the main existing recognitions and problems as follows: 1)Most studies on seismic activities in active-tectonic block boundaries still remain at the statistical analysis level at present. However, the analysis of their working foundations or actual working conditions can help investigate deeply the seismic activities in the active-tectonic block boundaries; 2)Seismic strain release rates are determined by tectonic movement rates in active-tectonic block boundaries. Analysis of relations between seismic strain release rates and tectonic movement rates in mainland China shows that the tectonic movement rates in active-tectonic block boundaries of the eastern region are relatively slow, and the seismic strain release rates are with the smaller values too; the tectonic movement rates in active-tectonic block boundaries of the western region reveal higher values, and their seismic strain rates are larger than that of the eastern region. Earthquake recurrence periods of all 26 active-tectonic block boundaries are presented, and the reciprocals of recurrence periods represent high and low frequency of seismic activities. The research results point out that the tectonic movement rates and the reciprocals of recurrence periods for most faults in active-tectonic block boundaries exhibit linear relations. But due to the complexities of fault systems in active tectonic block boundaries, several faults obviously deviate from the linear relationship, and the relations between average earthquake recurrence periods and tectonic movement rates show larger uncertainties. The major reason is attributed to the differences existing in the results of the current earthquake recurrence studies. Furthermore, faults in active-tectonic boundaries exhibit complexities in many aspects, including different movement rates among various segments of the same fault and a certain active-tectonic block boundary contains some parallel faults with the same earthquake magnitude level. Consequently, complexities of these fault systems need to be further explored; 3)seismic activity processes in active-tectonic block boundaries present obvious regional characteristics. Active-tectonic block boundaries of the eastern mainland China except the western edge of Ordos block possess clustering features which indicate that due to the relatively low rate of crustal deformation in these areas, a long-time span is needed for fault stress-strain accumulation to show earthquake cluster activities. In addition, active-tectonic block boundaries in specific areas with low fault stress-strain accumulation rates also show seismic clustering properties, such as the clustering characteristics of strong seismic activities in Longmenshan fault zone, where a series of strong earthquakes have occurred successively, including the 2008 M8.0 Wenchuan, the 2013 M7.0 Lushan and the 2017 M7.0 Jiuzhaigou earthquakes. The north central regions of Qinghai-Tibet Plateau, regarded as the second-grade active-tectonic block boundaries, are the concentration areas of large-scale strike-slip faults in mainland China, and most of seismicity sequences show quasi-period features. Besides, most regions around the first-grade active-tectonic block boundary of Qinghai-Tibet Plateau display Poisson seismic processes. On one hand, it is still necessary to investigate the physical mechanisms and dynamics of regional structures, on the other hand, most of the active-tectonic block boundaries can be considered as fault systems. However, seismic activities involved in fault systems have the characteristic of in situ recurrence of strong earthquakes in main fault segments, the possibilities of cascading rupturing for adjacent fault segments, and space-time evolution characteristics of strong earthquakes in fault systems. 4)The dynamic environment of strong earthquakes in mainland China is characterized by “layering vertically and blocking horizontally”. With the progresses in the studies of geophysics, geochemistry, geodesy, seismology and geology, the physical models of different time/space scales have guiding significance for the interpretations of preparation and occurrence of continental strong earthquakes under the active-tectonic block frame. However, since the movement and deformation of the active-tectonic blocks contain not only the rigid motion and the horizontal differences of physical properties of crust-mantle medium are universal, there is still need for improving the understanding of the dynamic processes of continental strong earthquakes. So it is necessary to conduct in-depth studies on the physical mechanism of strong earthquake preparation process under the framework of active-tectonic block theory and establish various foundation models which are similar to seismic source physical models in California of the United States, and then provide technological scientific support for earthquake prevention and disaster mitigation. Through all kinds of studies of the physical mechanisms for space-time evolution of continental strong earthquakes, it can not only promote the transition of the study of seismic activities from statistics to physics, but also persistently push the development of active-tectonic block theory.

Aketao M6.7 earthquake on November 25, 2016 occurred in Aketao Country, Kizilsu Kirghiz Autonomous Prefecture, Xinjiang Autonomous Region. In this paper, according to the focal mechanism solution, the geometrical parameters of the fault are used as input to simulate the acceleration records of the surrounding area of Aketao M6.7 earthquake on November 25, 2016 with the QSGRN/QSCMP program. The peak ground acceleration distribution is plotted by extracted peak ground accelerations and site condition correction. Comparison is carried out between the observed waveforms and the simulated waveforms in the time domain and the frequency domain of two strong motion stations. It is verified that the amplitudes of simulated and observed data are equal in magnitude, their calculated residuals are smaller in low frequency range and the spectral characteristics of simulated and observed data are consistent. Furthermore, the simulated peak ground acceleration corresponding to the coordinates of the actual survey points is extracted and compared with the survey spot intensity, and both have good correspondence. This paper attempts to provide a method for rapid output of peak ground acceleration distribution after an earthquake, so as to provide a method for estimating the seismic impact field in areas where the monitoring station is scarce, the topography is complex or it is difficult to implement the rapid earthquake disaster investigation. And it also provides supporting information for rapid assessment of earthquake disaster and emergency decision-making.

Ludian M6.5 earthquake on August 3, 2014 occurred in Ludian country, Zhaotong City, Yunnan Province. The epicenter is located at longitude 103.3°E and latitude 27.1°N with focal depth of 12km. Macro epicenter is located in Longtoushan Town, Ludian Country. The seismogenic fault is the NW-striking Baogunao-Xiaohe Fault, which is a second-order strike-slip fault of the NE-striking Zhaotong-Ludian Fault system. In this paper, we use observation records of 50 strong motion stations in the Ludian earthquake to calculate the distributions of peak ground acceleration and horizontal peak ground acceleration. According to the records, we find that the strong motion records decay slowly along the northwestern direction, thus we infer that the seismogenic fault is in NW direction. Meanwhile, the horizontal peak acceleration distribution map is similar with the synthetic three-component peak acceleration distribution map, which illustrates the horizontal component is the main part. Normalized S-wave radiation patterns are drawn using the observation records of 135 seismic stations. In contrast to single couple source radiation pattern, we infer that the seismogenic fault is in NW direction. The amplitude in the southeast direction is slightly larger than the amplitude in the northwest and the difference is not big, showing that the rupture trends slightly to the southeast. The above results show that, the conclusion of Ludian M6.5 earthquake rupture along the northwest direction is in consistency with the observation results of the strong motion stations and the seismic stations. In addition, the distribution of earthquake disaster also supports this conclusion.

We employ multistep inversions in the frequency or time domain to infer the kinematic characteristics of the M_S6.5 Ludian, Yunnan earthquake in 2014, mainly using regional broadband waves recorded by the China Digital Seismic Network. In this paper, we firstly invert the focal mechanism solution and the centroid depth of the Ludian earthquake, and then determine the best-fitting finite-fault model and the dominant rupture direction. According to the above results, we further analyze and discuss the kinematic characteristics of the Ludian earthquake, and explore preliminarily the reason for the serious disaster caused by this event.
We take into account some factors which could have effects on the inversion results, e.g. the use of different waves and simplified 1-D velocity models. Several test results indicate that the misfit between observed waves and synthetics is better, if we use the full waves and the 1-D velocity model (Model M2) in this study area. According to the point-source model (focal mechanism solution), this event occurs on a true rupture plane (strike=342°, dip=83°, and rake=-34°), which shows a left-lateral strike-slip faulting with a minor normal oblique component. The centroid in the horizontal direction is located at nearly 5.4km southeast of the epicenter(27.109°N/103.354°E), and the best-fitting centroid depth is around 4.4km. The total scalar moment, M₀, is retrieved with an average value of 2.1×10¹⁸N·m (or moment magnitude M_W6.1).
The rupture history indicates the event can be considered to have an asymmetric bilateral rupture source with a radius of 10km. The total rupture area is about 227.6 km² with an average slip of nearly 0.16m. Most of the energy releases within about 6s. From 0s to 2s, the energetic rupture starts at the nucleation center, then propagates bilaterally along the fault plane. After 2s, the rupture mainly extends south-east, showing an obvious rupture directivity. Finally, the rupture ends at nearly 6s.
In order to investigate rupture directivity of the Ludian earthquake, we retrieve the apparent source duration at different stations, using the method developed by Cesca et al.(2011). Rupture directivity of the Ludian earthquake is detected on the basis of a frequency domain inversion of the apparent duration at each station and the further interpretation of its azimuthal variation. The result indicates that it is obvious that the apparent source durations at majority of stations located southeast of the epicenter of the Ludian earthquake are relatively shorter, and the shortest one is round 2s. However, the apparent source durations of majority of stations distributed northwest of the epicenter are much longer, and the longest one is up to 9s. The mean value of apparent source durations of all stations is about 4.96s. Based on the principle of the Doppler effect, this result provides a clear indication for the rupture propagating towards southeast, and thus can be used to discriminate the true fault plane (NW-SE).
In the end, we speculate that one of the most important reasons why the Ludian earthquake caused the devastating damage is that most of the energy is instantly released within relatively short duration in the shallow layer of the upper crust.

Lushan M7.0 earthquake on April 20, 2013 occurred on the southern segment of Longmen Shan Fault zone. It is another strong earthquake after the Wenchuan M8.0 earthquake, rather than an aftershock. A total of 196 people died in Lushan M7.0 earthquake, 21 people missing, and 13019 injured. In this paper, we use the April 20, 2013 Lushan M7.0 earthquake strong motion records to calculate equivalent peak acceleration A_0.5 and convert it into the instrumental intensity. Then, we carry out the comparative analysis of instrumental intensity and surveyed intensity at the spots in the range of 5 kilometers around the stations. The results show a coincidence degree of 58.6% between the instrumental intensity and surveyed intensity, with the deviation less than ±1 degree. Meanwhile, we also recognize that there is a big dispersion of the observed ground motion parameters, even in the same macro-intensity. It is uncertain about the corresponding relationship between the individual station measured ground motion parameter values and the seismic intensity. The damage may be caused by earthquake ground motion characteristics, building types and site conditions and other factors, and therefore the difference between instrument intensity and actual seismic intensity cannot be ignored. On the whole, instrumental seismic intensity has potential for determining the range of seismic intensity, thus to provide quantitative reference, especially, in the absence of earthquake site investigation initially.

This paper discusses ultrasonic mechanism of the sealed-H₂ release from basalt rock on the basis of experiments.A small amount of released H₂ by ultrasonic vibration is escaping H₂,whereas most H₂ is absorbed H₂.Both phenomena can be regarded as a precursory H₂.So,it is possible to use H₂.to.forcast earthquake.Seeing that the intensity of u.v.and the its duration were constant in the experiments and the amount of released H₂ gas increased during each vibration,abnormal H₂ seems not to be representative of accoustic emission intensity,thus suggesting that the intensity of the failure of rock may be representative of the earthquake magnitude.

This paper illustrated experimental study on the influence of ultrasonic vibration upon microstructure and emanation process of saturated rocks. It was found by scanning electron microscope observation that specimens under ultrasonic vibration, whether on the ultrasonically irradiated sides or on the non-ultrasonically irradiated sides, all show a large number of fissured traces, i. e. microcracks. So it can be inferred that under the ultrasonic vibration the development of microcracks provided an outward-diffusing "passage" for the sealed radon to take part in the emanation process.The results of the ultrasonically vibrating experiments indicate that the ultrasonic vibration made the amount of radon separating-out of rock obviously increased, and the amount would be as much as it was before vibration when the vibration stopped. The increment of separating radon depends on the times of ultrasonic vibration. These results obtained are consistent with the scanning electron microscope observation showing that the primary rock structures were certainly destroyed by ultrasonic vibration. During each vibration, new cracks appeared and the separating amount of radon was also increased correspondingly. For the first vibration, the amount of radon increased by a factor of 4.5 under the temperature-control condition (it will be 4.3 under temperature-uncontroi event). For the fourth vibration, it increased by a factor of 9.1 under temperature-control condition (it will be 11.2 under temperature-uncontroi condition). Among the increment of radon, absorbed radon accounts for 64-76 percent, sealed radon accounts for 24-36 percent. The contribution of the ultrasonic thermal effect to the separating radon accounts merely for 10-30 percent.The experiment results show that the rock emanation process was clearly related to the ultrasonic vibration. Its mechanism is rather complicated. It is reasonably considered that prior to the failure of rock the expensive microcracking produced accoustic emission, i. e. ultrasonic vibration, leading to a constant emanation from rocks in a significant amount which may be probably regarded as a precursory information of radon.