A crucial question in earthquake science is how earthquakes start. Field and experimental observations show a short period exists between the fault reaching peak stress and the coseismic event. Therefore, it is of fundamental significance to capture the spatio-temporal evolution of a fault’s deformation during this premonitory stage. It can help us understand how the rupture of an earthquake initiates and also provide precursory information. Stick-slip events or lab quakes can be produced in controlled conditions to mimic earthquakes in nature. In previous studies, we proposed the fault meta-instability model focusing on depicting this stage(hereinafter referred to as the meta-instability stage)and interpreting the transition from energy/stress accumulation to energy/stress release. We further divided the meta-instability stage into two substages, i.e., the quasi-static phase and a quasi-dynamic phase, corresponding to slow energy release and irreversible energy release elevated rate.
However, how the meta-instability stage can facilitate the final failure remains puzzled. In contrast, the meta-instability stage exhibits slow and mild deformation, while the coseismic stage is fast and violent. In order to bridge these two processes, it is essential to record the complete dynamic process of stick-slip events, including the premonitory and coseismic stage. Thus, the data acquisition system required must feature a high signal-noise ratio, high frequency, continuous recording, and dense instrumentation. In 2016, we developed an ultra-high-speed, multi-channel and continuous recording data acquisition system for deformation measurement(UltraHiDAM). UltraHiDAM has 64 channels, 16-bit resolution, and 4MHz sampling frequency, and can perform parallel continuous data acquisition. It is able to record strain signals and acoustic emissions continuously and synchronously at a high sampling frequency up to 4MHz for as long as a few hours. To our best knowledge, it is the first system that is capable of doing so.
Based on this system, we conducted a series of stick-slip experiments. We recorded the entire deformation process of the laboratory earthquake cycles, including the relatively slow deformation in the quasi-static phase(several seconds before the stress drop), the relative fast deformation in the quasi-dynamic phase(a few microns before the stress drop), and the complete process of the transient coseismic slip. High frequency continuous synchronous sampling allows us to reveal as many details as possible of unstable sliding transient processes, and analyze mechanical problems related to the seismic source.
We report results of stick-slip experiments using saw-cut bare-surface granodiorite samples. The main findings of this paper are summarized as follows: First, the substages can be further recognized based on the local deformation characteristics(Table 2). Second, strain and stress start to localize before the quasi-static phase; such localization’s acceleration indicates the whole fault has entered the quasi-static phase. Third, the strain field during the quasi-dynamic phase is characterized by a wave-like acceleration and reciprocating propagation(Fig. 9). Fourth, there is a short preparation period for each sub-stage of the quasi-dynamic process(Fig. 6). The existence of such preparation periods may help the imminent earthquake prediction. Finally, even for the stick-slip events captured on a simplified plane laboratory fault, the coseismic process can be multiple rupture events, each event has its own AE waveform that is distinguishable in time(Fig. 8).
The implications are that there is indeed precursory information during the different substages before the coseismic event, most of which are associated with the localization and propagation of strain and stress. An earthquake source’s actual mechanical process can be complex in terms of multiple stress drops and ruptures.

Earthquakes commonly occur in the sliding surface of the fault zone. The morphology of the sliding surface is the result of fault activities, and also it evolves with the activities. The irregular geometry of the fault plane affects the sliding resistance, the concentration and anisotropy of the stress distribution within the fault plane and the fault shear strength. So, the acquisition of high-precision morphological features is of great significance for studying the correlation between fault surface morphology and seismic nucleation, fracture propagation and termination. Due to the lack of reliable micron-scale morphological measurement apparatus, the study of the coherence of the fault surface morphology from large scale(unit: m-cm)to small scale(unit: μm)is subject to restrictions, as well as the study of the relationship between the micro-morphology of the experimental frictional surface and the rupture process. In order to improve the measurement accuracy of the fault plane and overcome the shortcomings of existing measurement methods, we have invented a morphology measurement system with independent intellectual property rights.
The measuring principle of this morphology measurement system is based on the laser rangefinder theory. The frame of this system consists of four parts: Braced Frame, Moving Scanner Unit, System-Controlling Unit and Data Collection Unit. Braced Frame is made up of high-adjustable frame, loading stage, dust-proof box and isolation platform, which is used to provide a vibration isolation, light proof and dust-proof measuring environment. Moving Scanner Unit contains a laser head and a two-dimensional translation stag, the laser head is used to measure vertical distance and a two-dimensional translation stage carrying a laser head moving in X-axis and Y-axis orientation to provide X, Y coordinate values. System-Controlling Unit includes two-dimensional translation stage controller, laser head controller and signal convertor. The function of this part is mainly to control operation of other parts. The Data Collection Unit is composed of computer system and software module. This part connects other parts for receiving and storing data. In order to improve the scan efficiency, we developed new software by which we can precisely control the measuring process and efficiently process the acquired data. The software is comprised of five modules: 1)Move Module, this module is used to control the original moving of the laser head relative to the two-dimension translation stage and display the 3-dimensional coordinate information in real time; 2)Set Parameters of Scan Area, the function of this module is to obtain the XY coordinate values of four corner points of the target area to scan; 3)Scan Method Module, though this part, we can control the point spacing in the X-axis orientation by inputting velocity of laser header, as well as the point spacing in X-axis orientation by inputting the Y-step parameter; 4)Pre-Scan Module, there are three functions in this module to inspect whether the z-value of the target area is beyond the range of the laser head or not, estimate consuming time for scanning the object area under the predefined parameters and to estimate the size of the result file; and 5)Scan Module, the function of this module is to store the scanning data.
We scanned the camera lens and the standard plate whose standard deviations are lower than 5μm to acquire the precision of the measurement system, and the results show that the precision of the plane positioning (X-axis and Y-axis direction)is better than 3.5μm; the vertical measurement precision is better than 4.5μm. The highest resolution of the measurement system is constrained by the performance of the laser head and two-dimension translation stage, and the horizontal resolution can reach 0.62μm, vertical resolution 0.25μm. When the needed resolution is lower than the highest, we can achieve it through adjusting the parameter of the velocity in the X-axis orientation and steps in the Y-axis orientation. To test the practical effect of the measurement system, we scanned an area of frictional surface of experimental rock using this system and obtained a high-resolution topography data. From the DEM interpolated from the cloud data, we can observe the striation on the fault plane and the variation of the roughness distribution. The roughness and slope distribution results show that the topography measurement system can meet our requirements for analyzing the microscopic morphology on the micrometer scale.
Compared with traditional measurement devices, the morphology measurement system has the following advantages: 1)The measurement system can obtain the data even in a valley region with a large dip angle on the surface because the vertically emitted beam by the laser head is practically perpendicular to the surface. So compared with other means, it can avoid producing a blank area of measurements and get a complete area; 2)the measurement system has a larger measurement range of 30cm×30cm. When the high-resolution measurement is performed on a large scale, the error caused by the registration of multiple measurement results can also be avoided.

In the simulation experiments of earthquakes in laboratory, the instability slip or rupture events are obtained through steady state loading to simulate earthquake processes. In the experiments, steady-state deformation and unstable sliding occur alternately. It is hard to determine the origin time and duration of the instability event of fault, and there may be many instability events in one experiment. Therefore, in order to ensure that sufficient data is obtained at the extremely short instability moment to analyze the mechanical process of the earthquake source, the data acquisition system is required to continuously collect data at high-speed from the beginning of the experiment until the end, and the lasting time can be more than ten hours, so it requires huge storage space. Although the upper frequency limit of the instability signal is unknown exactly, but the previous experiments have shown that the frequency of the signal will reach hundred to several kilohertz, so the sampling frequency of the data acquisition should be above megahertz. In this case of long-time continuous high-frequency data acquisition, it is still necessary to maintain a high signal-to-noise ratio. Furthermore, previous studies have proved that the source mechanics field has a complex spatial structure, which is difficult to describe with a few measuring points, and it is necessary to perform simultaneous measurements of the source mechanics field for dozens or more measuring points. The combination of long-term continuous recording, high-resolution high-frequency sampling and multi-point simultaneous measurement poses a huge challenge to the technical indicators of the observing system. With the method for composing distributed synchronous acquisition machine group by using multiple high-resolution high-frequency sampling computers, a super dynamic deformation measurement system of high signal-to-noise ratio, which features 64-channel, 16-bit resolution, 4MHz sampling frequency, and parallel continuous acquisition with tens of hours was developed. This system can realize the synchronous acquisition of various signals, such as strain, acoustic emission, electromagnetic waves and displacement, so it is convenient for analyzing the conversion relationship between various physical quantities.

As the rock samples will produce abnormal signals of acoustic emission, microseismic and charge signals under external loading, the waveform comprehensive monitoring devices are used to synchronously monitor acoustic emission, microseismic and charge signals during the deformation and failure process of granite with fault zone under uniaxial compression. The results show that, the granite with fault zone has obvious synchronous precursory signals of acoustic emission, microseism and charge induction in the elastic deformation stage, and has high amplitude synchronous precursory signals in the instability destruction stage. The influence of fault zone on granite samples strength is remarkable, and the uniaxial compressive strength of samples with the fault zone is greatly reduced. With the angle of the fault zone decreasing, the uniaxial compressive strength of the specimens is reduced, the samples are more liable to instability and the energy of instability destruction is greater. With the fault zone angle of granite samples decreasing, the acoustic emission, microseismic and charge induction signals increase in the deformation and failure process of samples. The samples stress decreases when the acoustic emission, microseismic and charge induction precursory signals appear synchronously. The duration of acoustic emission, microseismic and charge induction precursory signals is increasing in the instability destruction stage. When the angle of the fault zone reaches 30°, the mutability of acoustic emission, microseismic and charge induction signal increases, the time to enter the dangerous stage is much earlier, and the acoustic emission events of large magnitude increase significantly, and the large angle faults of coal mine are more dangerous. The intensive and high amplitude synchronous precursory signals of acoustic emission, microseism and charge induction are produced before the instability destruction, and the signals duration is shorter. The intensive and strongest synchronous precursory signals of acoustic emission, microseism and charge induction are produced in the instability destruction, and the signals duration is longer. Acoustic emission monitoring data can better reflect the micro rupture of rock. And combined with the acoustic emission, microseismic and charge induction precursory signals, the precursory information of rock instability destruction can be obtained more accurately.

In order to study the complex strain field during the instability process of rapid deformation adjustment stage and the rock failure process in rock experiments, high-speed, high-resolution, multi-channel strain observation technology is applied to physical simulation experiment and developed as Multi-channel Dynamic Strain Observation System for monitoring the transient deformation field. The specially designed system has 96 channels with a sampling rate up to 3, 400 samples/s/ch, and a resolution of less than 1με(micro strain)and continuous recording. We found some important phenomena by observing the temporal and spatial variation during the fault instability process and crack propagation process. The experiment on three typical stick-slip models shows that the evolution of the unstable fault stick-slip process has a relatively stable feature characterized by three typical phases(precursory slip, rapid slip incorporated with high-frequency strain vibration and terminal adjustment). Each phase has its own characteristics of duration, strain rate, frequency, amplitude and energy. The experimental results on the three-dimensional fault propagation model show that the coalescence of the bridge area occurs at the last stage and is a rapid process. The crack initiation occurs at several points which combine each other randomly. There is a stable period existing between the full propagation and coalescence and sample failure, and the duration time is several tens milliseconds. Multi-channel dynamic strain observation system fills the band gap of strain observation in rock mechanics experiments and earthquake simulation experiments. High-density and high-precision strain data can be obtained by this system, which can further promote the studies on the transient strain field evolution and the temporal and spatial process of strain wave and provide technical support for understanding the process from a slowly progressive deformation releasing to the sudden instability.

After the great 1906 San Francisco earthquake, in 1910, Harry Fielding Reid published the article "The Mechanics of the Earthquake". From an investigation of the deformation of the ground surface which accompanied the 1906 earthquake and the seismological data from USGS Reid expounded explicitly the correlation between the faults and the shallow earthquake. He put forward firstly the elastic rebound model to explain the mechanical mechanism of the earthquake. The model consists of three basic points:
1)The earthquake originates from the fault movement; 2)The movement leads to the inhomogeneous elastic deformation on both sides of fault, accumulating vast amounts of energy; 3)Part of the elastic energy is released in the earthquake.
Brace(1966)proposed that the physical mechanism of elastic rebound should be explained with jerky sliding motion or stick-slip which is well-known in engineering. The introduction of the stick-slip concept combines the physical interpretation of the earthquakes between the shallow and deep source one and converts the research of deformation into the problem of friction, thus, resulting in a great upsurge in the study of friction in the 1970s to 90s. Twelve years later, Byerlee published the research paper-Friction of rock(1978), regarded as the Byerlee's Law, that is, when the normal stress σ_n is smaller than 2kb, the shear stress τ=0.85σ_n approximately; when σ_n is larger than 2kb, τ=0.5+0.65σ_n.
Unfortunately, although Byerlee's paper referred to the variation of data for many times, the empirical formula of his own did not give the range of variation, so the reader could not calculate or evaluate its reliability. Furthermore, in the friction constitutive relation expressed by a piecewise function, the cut-off point is set at 2kb, of which there have been no explanations about its physical meaning or statistical basis so far. In Byerlee's frame, stick-slip is assimilated as a spring-block model. Fault displacement is set for the rigid block friction movement and deformation is set for spring extension. The spring extension is imputed optionally to the mechanical frame deformation of loading machine.
The friction constitutive relation can be described only by one constant and the stress field along the fault plane be gotten directly with projection transformation of loading force. This simplified mechanical model is so exciting that it seems to be paving the understanding avenue in the process of earthquake. Only after a year when Byerlee published the paper about rock friction, a mathematical model was deduced based on the simplified assumptions(Dieterich, 1979), and soon it was further simplified as the so-called velocity-dependence equation:τ[μ₀+(a-b)ln (V/V₀)]σ
Where, τ is shear stress and σ is effective normal stress; a and b represent the material properties; V is sliding velocity, where V₀ is the reference velocity, and μ₀ is steady-state friction coefficient when V=V₀. For a specified fault, the sliding friction behavior or the instability depends only on the plus or minus of(a-b)
.But if μ₀ is not constant, the situation will become very complicated. The experimental results show that the numerical scale(a-b)is often on the parts per thousand(0.001～0.004). Then even the same rocks in different conditions or different sliding stages the variation of friction coefficient is also on decimal point first(such as granite: 0.5～0.7, gabbro: 0.2～0.7). Therefore, if the variation of friction coefficient μ₀ is taken into account, the contribution or effect of (a-b) to the friction angle changes is almost negligible.
In addition, the stress σ and τ on fault surface are taken from the axial projection. The projection must assume the rock slide block is rigid. It means along the slip surface of rock without any deformation. This has violated the elastic rebound model in essence described by the basic facts and does not conform to earthquake field investigation results.
Actual measurements have proved there is a complex deformation mode on smooth fault slip zone. Different parts of the fault have different deformation processes. Even the average stress state of the near fault parts is different from both in the direction and the value projected with loading force. Other experimental results show that, during the stick-slip, fault is not only to complete a simple smooth one-way movement, but experiences complicated multi-times and multi-point tremors to release the energy accumulated in the fault zone. These new experimental results agree with the basic model proposed from Reid by earthquake field investigation in 100 years ago.
Back to the classic is to change the research direction back to reality of earthquake, this is the right way of the future.

The objective of this paper is to explore the current tectonic activity with satellite remote sensing thermal information by taking a case of the Wenchuan earthquake. Three items are accomplished as follows: 1)the process of evaluation of thermal field before and after Wenchuan earthquake is analyzed. The results indicate that there exist plenty of thermal information associated with the distribution of tectonics in the in-situ land surface temperature field, which is extracted from land surface temperature in which the effects of non-tectonic factors, such as topography, atmosphere and solar radiation are gotten rid of. 2)Combining with measurement of the shallow atmosphere temperature, the quantitative relationship between increment of land surface temperature and crustal stress-strain is preliminarily investigated. Results indicate that the increment generated by the crustal stress may obviously affect the shallow atmosphere temperature. Especially, the temperature-lowering zone has more reliability. There exist some temperature lowering zones along the boundaries of large geological blocks before and after Wenchuan earthquake, which reflects the relatively extensional movement (or stress relaxation) among these corresponding blocks. 3)Based on the co-seismic deformation, the comparative analysis is done between co-seismic deformation and thermal information. Results indicate that the tectonically adjusted area obtained from temperature field is largely accordant with that of co-seismic deformation. This shows that the variation of temperature along the boundaries of large geological blocks within the Tibet Plateau reflects the process of the adjustment of crustal deformation of the Tibet Plateau before and after Wenchuan earthquake. In summary, it is a possible approach to obtain the change of state of crustal stress by using the thermal method.

In the present paper, double shear stick-slip model experiments were conducted for simulating the regional loading process of spontaneous earthquake and induced earthquake, and a strain observation system was employed to acquire dada continuously to monitor the local strain changes near fault under the loading process. Regional stress path and local strain path during the whole earthquake process have been described on the stress space and strain space. Though the morphology of local strain path is different greatly from the macroscopical stress path, there is a certain mapping relation on the corresponding transformation stages between stress path and strain path. The evolution trend of local strain path indicates the possible stage of the fault deformation. The strain path of spontaneous earthquake can be divided into three parts: Strain accumulation, linear deviation of shear strain and unstable slip. The strain path of induced earthquake includes four stages: Strain accumulation with positive slope, steady state slip with negative slope, metastable strain stalemate and unstable slip under a disturbance. Spontaneous earthquake and induced earthquake have their own inherent and steady path model, so the fault stability and the possible earthquake type could be judged according to the special strain path.

Earthquake is a form of tectonic activity, an earthquake contains a lot of information, its frequency range is very wide. So far, the main frequency range of seismic records is below 100Hz, and the ultrahigh frequency seismic information got less attention by researchers. However, high frequency microseismicity can provide important information on micro-activities of fault. This paper introduces a new 100kHz ultrahigh frequency seismometer and its first application on high frequency microseismicity observation. The telemeter has been applied in Dali, Yunnan. Several high frequency microearthquakes were recorded and the corresponding earthquake location and spectral analysis were done. The results show that high frequency microearthquakes caused by crustal movement do exist and can be recorded by our telemeter. The recorded microearthquakes are of magnitudes between M_W=-3.0 and M_W=-1.0, maximum hypocentral distance up to 4.87km, dominant frequency ranging from 100Hz to 300Hz, and maximum frequency as high as 800Hz. It is possible that the ultrahigh frequency microseismicity observation(UFMO) will become a new approach to the study on the tectonic movement.

Recently,the strong earthquakes in China mainland occurred mainly around the Bayanhar block. It is important to monitor the information of ongoing crustal activity at the key tectonic positions. We have developed a set of wireless equipment for measuring the ground temperature in field,and have established a network of measurement of the ground temperature along Xianshuihe Fault. Some changes of temperature were observed before and after the Lushan earthquake on April 20,2013.First of all,an apparent and persistent change of the ground temperature in Kangding appeared,starting from January 31,2013.This temperature variation corresponded with the occurrence of the small earthquakes around the observation station. According to the relationship between the temperature and stress,the abrupt change of ground temperature is essentially the geological stress adjustment. From the viewpoint of geological structures,both Longmengshan Fault and Xianshuihe Fault are the boundary faults of the Bayanhar block,but located at different boundaries,so,Kangding in Xianshuihe Fault is tectonically related to Lushan in Longmengshan Fault. Thus,the temperature change described above would possibly be the precursor of the Lushuan earthquake.

By using a newly developed distributed multi-channel transient signal acquisition system,the spatial distribution of arrival time and direction of the first motion of the acoustic emissions (AE) along the fault under near field conditions was well studied. With the high frequency strain recording system,the stress distributions along the fault were well characterized and the focal dislocation behavior during stick-slip was studied as well.The experiment was conducted in a biaxial servo-control loading system. Analysis shows that for each sudden stress drop event,the corresponding focal dislocation consists of a few smaller sub-dislocations and each sub-dislocation results in its own AE event and stress change. Besides,it can be proved that each sub-dislocation may correspond to multiple sources rather than point source,the evidences of which can be found in the heterogeneity of the spatial distribution of the first motion and the strain field along fault. Our study indicates that the multi-dislocation during stick-slip under near field conditions is very complicated. By the result of our study,it is more convenient to understand the complication of focal process. Besides,it could provide an evidence for the uncertainty of in-situ earthquake focus and the high contradiction rate of earthquake mechanism solutions.

The relationship between the thermal field and strain field during deformation of fault is the physical basis to clarify whether satellite infrared information or the ground temperature field can be used to study fault activity.This study attempts to discuss these problems by laboratory experiments.A two-direction servo-control system was used to apply load on the samples with compressive and tensile en echelon faults.An infrared thermal image system and a contact-type thermometer recorded synchronously variations of the brightness temperature field of infrared radiation and temperature field during deformation of the rock specimens.A digital CCD camera and the digital speckle correlation method(DSCM)was used to collect and analyze images to obtain the evolution processes of displacement and strain fields.The experimental result shows as follows:1)the temperature is the highest of the jog area of the compressive en echelon faults,whereas that is the lowest of the extensional en echelon faults prior to failure of the jog area.The record by DSCM displays that the mean strain of the jog area is the largest for the compressive en echelon faults,while that is the smallest for the extensional en echelon faults.These mean that the temperature field has clear responses to the opposite stress states at the jog areas with two kinds of en echelon faults,and thus provides an indicator for determining whether the fault has slid;2)the en echelon faults experience two deformation stages from the stress building up and fault propagating in the jog area to the unstable sliding along the fault.Correspondingly,the mechanism of heating-up turns from strain heating into frictional heating.Three kinds of phenomena have been observed at the jog area and its vicinity during the stage of transformation.They are temperature drop,fast fluctuation of temperature,and pulses of temperature rising.Mechanism of these phenomena is discussed;3)these variations of thermal field at the jog area are followed by swift rise of temperature along the fault.The onset of temperature rise along fault occurs 2～3 seconds prior to its unstable slip.However,the temperature drop of the jog area happens about 20 seconds before the unstable slip of the fault and the appearance of temperature rising pulses is 10～20 seconds earlier than that of the fault slip.They are precursors to unstable slip of the fault.These experimental analyses demonstrate that observations and studies on the thermal variations at the sensitive portion of a fault comparing with other relevant data are of great importance for detecting precursors ahead of unstable slip of active faults.It is planned to design an observing project along a suitable fault.

New software for acoustic emission location has been developed and the main functions,methods for location are introduced briefly.The tested location result with the software is presented also.The multi common Data Base interfaces are built in the software,so it is device-independent.Besides the three-dimension location,it can be employed for surface or linear location.In locating,the velocity of sample material will be an appointed constant or an unknown variable.There are two improvements in the algorithm.The first is that the least absolute deviation based on robust estimation is taken as objective function for location.The second is a simulated annealing method that is applied to search the global minimum of objective function in optimization procedure.The location accuracy of acoustic emission focus is much better with the software and it is a flexible tool for data processing.

Compared with the strong noise background,intensity of efficient signal is very small when using thermal radiation to obtain the current tectonic activity. Increasing range of thermal radiation resulting from fault activity accords with dimension of inversion precision of thermal radiation of land surface,and information of fault activity is submerged by inversion error of thermal radiation of land surface. Besides,it is difficult to discriminate the information of fault activity when the data precision is high. This paper develops a new method,i.e. the Mean Grad Method (MGM),to solve the above-mentioned problems. First,we average the long-time data to gain a higher data precision at an expense of losing time information,then,distinguish the space distribution of the current tectonic activity from thermal radiation field of land surface by space grad of thermal radiation according to the difference between influence of atmosphere and tectonic activity on land surface. This method can offer some features on variation of current tectonic activity with time and space and is helpful for plotting out earthquake danger area.

Acoustic emission (AE) technique is one of the important methods in rock mechanics and experimental seismology. Improving AE data acquiring and processing technology is a basic work in research of the features of AE during the process of rock deformation. Following the progress of the AE systems in State Key Laboratory of Earthquake Dynamics,more and more AE data were obtained with higher recording speed and higher resolution for every experiment. It is a challenge to store such growing data reasonably and deal with it efficiently up to now. Researchers who work in the stand-alone mode have to copy a lot of raw AE data files with different format from remote computer to local PC. Storage space and time are thus squandered away. In addition,every user has to rewrite some processing programs using his own language for the row data,as a result,human resource is wasted. The commercial data service systems are not designed for laboratory study. According to the special demands,AE data service system based on web is designed and realized in the laboratory. Experiment AE data and processing method can be accumulated with the system. On the AE data access interface which is independent of equipments,researchers can acquire and process data easily. The system enhances the frequency of data use and the efficiency of data processing because it allows more and more people to participate in processing and analyzing the AE data. The system not only satisfies the needs of researcher and programmer,but also overcomes the previous shortcoming in the operating mode of AE data management. The most important thing is that the resolvent of AE data service system can be extended and applied to other experiment data to improve work efficiency.

The crustal movement shows obviously non-linear characteristics in time and space,and the tectonic deformation and related phenomena are undoubtedly affected by such non-linear mechanical process. Experimental study on frictional behavior under non-linear loading is helpful for understanding the effect of stress change on earthquake mechanism. In this paper,the frictional experiments on medium-scale granodiorite samples containing 45? precut surface have been conducted by using a servo-controlled biaxial rig,and the effect of lateral stress (σ₂) on friction has been studied. The experimental results indicate that the interval and stress drop of stick-slip events increase with increasing normal stress on sliding surface under constant loading velocity in the direction of maximum compression stress σ₁ and constant lateral stress σ₂.However,the relation is greatly changed when stress perturbation of sine wave with high frequency and small amplitude is added to σ₂. Stress drop and interval of stick-slip events are obviously scattered. In particular,the magnitude of stress drop may tremendously increase. Strain measurement indicates that the perturbation of lateral stress increases the heterogeneity of strain distribution and release along fault,and stick-slip events with large stress drop and that with small stress drop are corresponding to strain release along whole fault and that along parts of fault,respectively. Compared to the previous experiment result of stress perturbation in σ₁,the effect of perturbation in σ₂ on stick-slip is more obvious,meaning that change in normal stress may affect fault instability more greatly than change in shear stress. The experimental results imply that not only the effect of change in coseismic Coulomb stress on seismicity should be analyzed,but more attention should be paid to the possible effect of change in normal stress on fault when the fault interaction is considered.

For extracting the information about fault activity and seismicity from thermal infrared radiation,two crucial problems should be taken into consideration: 1) Aerosphere exists between solid earth and the satellite based sensor,and thus influences the infrared energy of land radiation when it passes through the aerosphere to arrive at the sensors. Moreover,the atmosphere itself is involved in thermal radiation,too. Therefore,to extract the geophysical information from thermal radiation,we have to begin with the estimation of atmospheric influence; 2) for identifying the anomalies induced by fault activity in thermal radiation,we should be able to distinguish the normal radiation from anomalous radiation,as anomaly can only be identified in reference with normality. This paper discusses these problems from the following aspects: 1) define the concepts of land surface brightness temperature (LSBT) and its annual variation field. LSBT would be an important physical concept in extracting the geophysical information from thermal radiation. LSBT should be assigned to the category of radiation energy of land surface,rather than the traditional "temperature" concept. Once the emissivity is known,the temperature can easily be calculated according to Planck law; 2) introduce the split window method for calculating the LSBT and the influence of atmosphere. The calculation shows that the influence of atmosphere on brightness temperature received by satellite based sensor is about ?10K; 3) the annual variation field of LSBT in China is extracted from the LSBT data by using the wavelet and split-window methods; the data used in this study come from the observations of NOAA/AVHRR from 1981 to 2001. The results of this study may provide a basis for further analyzing the anomaly field of thermal radiation.

After a series of complicated modifications the surface temperature of the Earth can be measured. Usually this surface temperature is referred to as land surface temperature. The heat effect that engenders this temperature variation may come from the climate and subsurface heat sources. Among them,the subsurface heat sources are of particular importance to earthquake and active fault studies. Unfortunately,the heat effect of climate is much larger than that of the subsurface heat sources so that it may conceal the information of subsurface thermal activity. Heat Penetrability Index (HPI) method is proposed in this paper for measuring subsurface thermal activity. The first assumption for HPI method is that the sun is heating the land surface synchronously. The second is the horizontal variation of heat exchange rates of rocks on top of subsurface heat sources. Provided that the rocks are heated,then the variation of surface temperature can be observed. According to the afore mentioned concepts and thermodynamics,the HPI can be deduced through correlation analysis as expressed by: where u and v represent surface temperature in two different areas,n is the sample number in the scanning window and D is the Heat Penetrability Index. When subsurface heating event takes place,D will go up,and vice versa. Some examples from experiment and satellite infrared image analysis are presented to test the effectiveness of HPI method.

The calculation of land surface temperature from thermal radiation requires the emissivity value of rock. However,there are lots of rocks in the Earth,and even the rocks of the same type may have obviously different emissivity. Recent methods for measuring the emissivity of rock are relatively complex,and most of them depend excessively on environmental condition. Therefore,when plenty of emissivity data are needed,it is necessary to develop a simple method of measurement. In fact,emissivity is a constant at room temperature,and the radiation of instrument itself and environment can enjoin the inversion as an unknown quantity. Then emissivity can be obtained by least square method through measuring the radiation of rock in a series of temperature and radiation conditions. On the basis of this method,the emissivities of 16 rock samples were measured. The square error of the results keeps in about 0.01,mostly less than 0.01,and the correlation coefficient of all linear fits is larger than 0.99.

The correlation between deformation and temperature is tested in detail on five rock types commonly observed on the Earth's surface. It is proven by both experimental and theoretical analysis that during the elastic deformation stage,the temperature of the rock is directly proportional to the loading force. According to the first and second laws of thermodynamics and Maxwells equation,the following relation can be deduced: σ-σ₀=-(cσρ)/α ln [T/T₀] or T/T₀= exp [-α/cσρ(σ-σ₀)] where σ is stress,T is temperature in Kelvin scale,α is modulus of heat strain,c σ is heat capacity and ρ is density. The results of the experiment are consistent with that of classic thermodynamics. Under Earth's surface condition,i.e. ordinary temperature and pressure,the maximum temperature increment of the sample elastically deformed in an open system is around 0.2K,and the normal rate of temperature rise is approximately 3mK/MPa. At the moment of fracturing,the temperature increment of the samples is much greater than that during elastic deformation,ranging from a couple degrees to more than ten degrees. The temperature increments caused by the elastic deformation of rocks are so minute that not only will they have any effect on climate,but cannot be identified by satellite infrared detectors. Therefore,new methods have to be explored for distilling faint information from infrared images,while the other thermo anomalies aside from elastic deformation should be carefully reviewed.

The infrared signals produced during rock deformation are very faint and tiny,so that the measurements of them require excellent capabilities of detection system and strict working environment. The reliability of the results of infrared experiment,therefore,depends greatly on the reasonable technical specifications of infrared photodetector,high sensitivity recorder for temperature field and effectively controlled thermal background. All of the key parameters are discussed in detail in this paper based on laboratory infrared observation. The features for an infrared camera include spectral band,digitizing resolution,view field etc. For capturing the optimal infrared images,it is suggested that the spectral range should be 7～13μm,the observation range should be from 0.5m to infinity,and high-density focal plane array should be employed as far as possible. In order to confirm the temperature from infrared image,a surface temperature recorder with multi thermometers should be employed as an eyewitness working alongside the camera. A realistic experiment system has been developed based on the afore mentioned discussion. Aside from the traditional functions of infrared camera,the system is fitted with high-speed data transmission component. The infrared focal plane array has 320×240 pixels,and 25 frames of temperature field data are saved into the hard disk per second. Some steel beam bending and steel cylinder compressing experiments were made for testing the system. The tests have proved that the system is reliable and the data quality is improved greatly.

The propagation of acoustic waves through periodically layered media (PLM) is experimentally investigated for the purpose of determining how the velocities and waveforms depend on periodical structures and detection frequency. In our experiments the PLM were made of stacks of plastic and steel plates with different periodical spacing. The acoustic wave frequency used in our experiments was 240MHz. Velocities and waveforms were measured by Digitization Wave Meter with 12-bit resolution and 40MHz sampling rate. The results show that the velocity transition from the time averaged relation to the long-wave approximation (LWA) is related to the critical ratio of λ_i/d_i(i=1,2) (λ_i-the wavelength of wave going through medium i; d_i-the thickness of medium i in PLM) rather than to λ/d (the ratio of wavelength to the thickness of rhythmic layer). The LWA occurs when λ_i/d_i>5(i=1,2),and the velocities (V) approach to the Reuss bound (V^R),whereas the time averaged velocities hold when λ_i/d_i<4(i=1,2). We attribute the behavior of V→V^R to "softer shape" distribution of different rigid components in PLM. The waveforms are also frequency dependent. With larger thickness of rhythmic layer,the attenuation of the first arrival is low and the multiple waves reflected from interfaces can be observed. In the case of LWA,however,the transmitted waves tend to have relatively simple waveforms. This behavior implies that simple waveform does not always reflect simple medium structure through which the wave propagates. As λ_i/d_i～6 (i=1,2),the transmitted wave is obviously attenuated,and consequently it is equivalent to a shield layer for the wave. Therefore,the influences of medium structures on velocities and waveforms should be taken into account when interpreting deep structures and compositions based on seismic data. Tectonic nature of intracrustal low velocity zones is one of the hotly debated topics. An important implication of our experimental results is that PLM are potentially low velocity zones. If extremely reflective deep crust is caused by alternating layers,the relatively lower velocities observed in deep crust should probably be attributed to the rock properties that we wish to image,as well as to its structures,scales and detection frequencies.

The Gutenburg-Richter law of magnitude-frequency relation has been challenged in recent years. Many researchers have found nonlinear sections in the magnitude-frequency relation in some cases. The magnitude-frequency relations shown in acoustic emission experiment of the rocks under confining pressures are studied in this paper.A new AE experiment system related to the confining pressure has been developed. The system can record the energy of AE events ranging from 1 to 10⁶ in relative energy unit.Six experiments of b-value simulation related to the confining pressure were conducted. The specimens were made of granite,marble,and granodiorite with different configurations. The main results of experiments are:1.The characters of acoustic emission of rocks are very similar with those of earthquake,providing the basis of applying AE experiment for the seismological research.2.The time series of AE events is fractally distributed on the time axis and its fractal dimension is a measure of AE activity.3.The AE series has not any characteristic occurrence period. The series of large AE events accords with neither the time-predictable model nor the slip-predictable model.4.There are two types of nonlinearity of b-value in the magnitude-frequency relations of either earthquakes or AE events.5.The magnitudes of AE events are fractally distributed on the magnitude axis. In addition,some problems aboul the theoretical study on b-value and applyingthe new method to seismological research have been discussed.

By the spatial scanning of the frequency-magnitude distribution of earthquakes in Beijing-Tianjin-Tangshan-Zhangjiakou area,it is found that there are three types of the distribution of earthquakes in this area. Type Ⅰ is consistant with a linear b-value model and types Ⅱ and Ⅲ are consistant with a characteristic earthquake model. The area,in which large earthquakes have never occurred in last 22 years,often has a distribution of type Ⅰ,and the area,in which large earthquakes have occurred,often has a distribution of type Ⅱ and Ⅲ. Result of the rock mechanical experiments shows that the frequency-magnitude distribution of acoustic emission events in rocks has a high b-value in the process of fracturing and fracture joining,and has a lower b-value in the process of frictional slip accompanied with stick-slip. But the frequency-magnitude distribution in a whole process of rupturing and slipping of intact rocks is consistant with the characteristic earthquake model. Therefore,it is evident that the frequency-magnitude relationship may represent a stage of structural deformation for an area. The nonlinearity of frequency-magnitude distribution might give us more information about the geometrical characteristics of the rigional structure. The emergence of nonlinearity of F-M distribution in area with moderate magnitudes but without large magnitudes in the considered period may indicate that the area is in a state much closer to the critical point for the unstable sliding.