Continuous gravitational observation data before the Cangwu MS5.4 earthquake of Wuzhou station are analyzed and studied, as well as data before the Hutubi MS6.2 earthquake of Korla station. Ten months’ non-tidal persistent gravity increases at about 5.0 μGal per month are observed at Wuzhou station before Cangwu earthquake and 7 months non-tidal persistent gravity increases at about 5.1 μGal per month are observed at Korla station before the Hububi earthquake. The cumulative variation amplitudes are about 50 μGal and 35 μGal respectively. Environmental interference is eliminated through site verification and theoretical analysis. It is believed that the increase in gravitational non-tidal persistence over several months before the two earthquakes may be related to the seismogenic process. The results of this study can provide an important reference for the construction of continuous gravitational observation earthquake prediction index system.

Variations of the gravity and the ground density of Nepal and its surrounding area are calculated from the GRACE satellite gravity observations; furthermore, the time series of gravity changes at typical points around seismic sources in the region are also obtained. The results show that the obvious negative gravity change rate is concentrated in the Himalaya and Burma arc regions, while the internal Tibetan plateau, northern India and the eastern Southern China block mainly present the significantly positive gravity change rate. The Nepal earthquake occurred in the transfer zone between the positive and the negative gravity changes. The gravity change rate has obvious time segmentation features during 2015: gravity changes decreased from January to April, increasing from May to September, but the gravity changes gradually recovered to increased tendency after December, 2015. Time series of typical points reveals that the gravity around the seismic sources region of the Nepal earthquake overall show long-time decreased feature, the gravity change rates mainly present the characteristics of “increase-decrease-increase-earthquake occurred-decrease” from 2013 to 2016. The changes of the ground density can well reflect the relationship between the ground mass migration and the gravity change rate, which also provides an important reference for understanding the dynamic mechanism of the Nepal earthquake.

New activities since Late Quaternary are discovered by the trenches of the Tancheng-Lujiang fault zone,Xujing of Sixian county, Anhui province, showing reverse fault type colluvial wedge,tensile fracture wedge, fragmentation flow and seismic fault. Based on comprehensive analysis of the phenomena on site, we make preliminary inferences that after the Middle Pleistocene deposits,the trench was pressed first, the Late Cretaceous brick red sandstone thrust collapsed on the Middle Pleistocene deposits from the west to the east, forming a reverse fault type colluvial wedge.Then the tensile fracture activity occurred, splitting the wedge open. High speed extrusion thrust faults occur, along the east edge of the wedge, cross section extrusion,the fragmentation flow was generated and formed a reverse fault type colluvial wedge. The latest activities of the fracture occurred after black soil overlying sediments, showing the small amplitude thrust activity,caused faults on both sides of the black soil thickness difference. It is indicated that after the Middle Pleistocene deposits, the fault underwent successively four activities in the form of “extrusion-tension-extrusion-extrusion”. The latest activity occurred into the Late Quaternary.Although known to be confirmed by the age of the samples tested and results of the analysis of the microscopic observation,these attempts to understand the new activity in the Early Holocene. According to the formation of reverse fault type colluvial wedge,tensile fracture wedge, seismic fault, these activities should be high speed stick slip activity and corresponding to several strong events of prehistoric earthquakes. The above understanding is partly confirmed by the results of the original microscopic observation.

In order to investigate crustal shallow structures and fault shallow tectonic features crossing the Changyuan fault depression, we completed a 49 km multiple shallow seismic reflection profile and a 7 km shallow seismic reflection profile. According to the seismic reflection profiles and combined with the regional geological data, we analyze the crustal shallow structures and the fault shallow tectonic features in the study area. The results show that Changyuan fault depression is a deep east and shallow west dustpan-like subsidence, and the Cenozoic basement of the fault depression plunges from west to east; Huanghe fault is the west dipping spade shaped normal fault, and this fault disrupts 190 m Quaternary strata in the shallow part, disrupting multi-group Cenozoic lays and the crystalline basement.Therefore, it is a buried active fault in the Quaternary period; Changyuan fault disrupts the bottom of Neogene strata in shallow part, and it extends below the crystalline basement, therefore it is a Pre-Quaternary fault. The result provides seismological evidence for active tectonic research and seismic risk evaluation in Xinxiang and its neighboring districts.

 Using the seismic data from north China since 2007, we calculate the earthquake clustering rate based on the nearest-neighbor distance algorithm. We study the characteristics of earthquake clustering of the seven seismic belts in north China，including Tanlu, Yanshan-Bohai, Hebei plain, Yinchuan-Hetao, Fenwei, Huai river, the lower reaches of the Yangtze river-Yellow sea, analyzing the characteristics of seismic activity in recent years. By analyzing the characteristics of earthquake clustering of Haicheng earthquake area, we consider that the earthquake sequence is weakening, and the occurrence of small earthquakes can be regarded as the background earthquake reflecting the regional stress state. Taking the Alxa M5.8 earthquake that happened on April 15, 2015 as an example, we study the characteristics of earthquake clustering before significant moderate-strong earthquakes in north China, and find that earthquake clustering rate will appear high value firstly, followed by a low value in the main earthquake zone, then the earthquake will erupt. This corresponds to the phenomenon of seismic activity intensity-quietude course before moderate-strong earthquakes, and this phenomenon has certain significance to determine the potential seismic hazard areas in the future.

This paper explores different prediction models in bridge deformation prediction and attempts to improve their accuracy. Combining bridge deformation monitoring data and the combination forecasting method, a combined forecasting model of the MC error correction and optimization of the bridge is constructed. Through examples, it is concluded that the combination prediction, as compared with single prediction, has higher predictive accuracy and stability, the highest of the RBF neural network combination.Furthermore, optimizing the error correction model further reduces the prediction error; the relative error of prediction results after optimization of the expected value is 0.86% and the variance value is 0.097 3 mm2. The accurate prediction of the deformation of the bridge verifies the effectiveness of this method.

In order to solve the problem of multi output slope deformation prediction,and improve the accuracy and computational efficiency of the slope deformation prediction model, a prediction model of slope deformation based on multi-output relevance vector machine(MRVM) is proposed. MRVM is established through the improvement of the output function of RVM. In order to improve the prediction accuracy of MRVM, the parameters of MRVM are optimized based on PSO.Taking the slope deformation of a large-scale dry-dock as an example,the horizontal deformation and settlement deformation of slope is predicted based on MRVM. The accuracy and computational efficiency of the model are analyzed. The experimental results show that the accuracy of MRVM is higher than that of BP neural network, support vector machine (SVM), RVM. It is proven that the expanding method of RVM is feasible, and MRVM has excellent prediction accuracy. The calculation time of MRVM is less than that of BP neural network, support vector machine (SVM), RVM. By using the above method, the repeated modeling can be avoided to improve the computational efficiency and simplify the modeling difficulty. Finally, multiple deformation of slope is predicted based on MRVM at the same time.

In order to solve the problem of fixed weighting of the entropy weighting method, two models are proposed: a weight assignment method based on IOWGA operator and a variable weight combinatorial model based on entropy weighting and IOWGA operator. In this method, the prediction precision of each model is used as the induced value, and the single model is orderly weighted according to the magnitude of each period. We use an example to illustrate this method. Experimental results show that the method has the advantages of reasonable weighting process, better model higher prediction accuracy, and is more suitable for deformation analysis and prediction.

The noise of GPS site time series is difficult to model and separate from signal. In this paper, the noise of GPS time series for 9 IGS stations is reduced by empirical mode decomposition (EMD), an adaptive signal analysis method. First, we decompose the GPS time series by EMD and get a series of intrinsic mode components and its trends. Then we use the correlation coefficient to distinguish the intrinsic mode function (IMF) between noise with signal and restructure the intrinsic mode function. Finally, we use the correlation coefficient, signal-to-noise ratio, and percentage of energy, to evaluate the effectiveness of the EMD method for noise reduction. The results show that the EMD method could separate signal and noise in sequence reasonably and it effectively weakens the impact of noise in GPS time series. Furthermore, it could further improve the accuracy of the GPS site time series.

Error equations are established using un-differenced observables and the equivalent double-differenced observation equations of multiple different stations are derived from the equivalent transformation theory. Meanwhile, the satellite and receiver’s clock biases are both eliminated. According to more numbers and stronger correlation of the equivalent ambiguity in multi-baseline solution, we propose a partial ambiguity resolution (PAR) method, in which all equivalent ambiguity is first sorted in ascending order. Subsequently, the maximum variance ambiguity is eliminated in an iterative process, while the ratio test is used to find a subset of equivalent ambiguity, which can be fixed. We validate this method using data from 5 IGS stations. The results show that, PAR presented in this paper not only can improve the fixed rate and success rate of ambiguity resolution, but also can greatly improve the availability of the multi-baseline solution, and the coordinate accuracy of PAR has considerable consistency in that of FAR.

In this paper, the mathematical model of GPS and GLONASS based on the single difference between stations is derived, and the error factors affecting the positioning accuracy are analyzed. Considering that the ambiguity is fixed without cycle slip and gross error, single difference ambiguity can be more smooth using filtering of selecting equivalent weight matrix, and then can be fixed with double difference ambiguity after choosing a reference ambiguity. The dynamic and static short baseline data are measured with different lengths, and the results show that the double difference ambiguity based on adaptive filtering by selection of the parameter weights of single difference model can be fixed well, and success rate is higher than the single epoch solution.

In this paper, we determine precise orbit and clock of Beidou-3 experimental satellites based on data from several tracking networks. The results indicate that the orbit overlap accuracy in radial direction of Beidou-3 experimental satellites is better 7.0 cm for IGSO satellites and 5.3 cm for MEO satellites, respectively, which is equivalent to that of the Beidou-2 of non-GEO satellites. In addition, solutions of static precise point positioning (PPP) show that better than 1 cm in horizontal and 2.6 cm in vertical direction can be achieved when orbit and clock products of the Beidou-3 experiential satellites are included. Compared to the Beidou-2 only PPP solutions, the results are improved by 0.5 cm and 1.2 cm respectively, and speed up the convergence rate, reducing convergence time by about 2 h 35 min.

A real-time high precision positioning calculation method of the BDS real-time orbit correction parameters is presented. The precision of the real-time orbit correction parameters and its attenuation condition are analyzed. The results show that the real-time orbit correction parameters obtained by this method could basically meet the demand of the BDS real-time high precision positioning. The average precision of the GEO satellites’ orbit real-time correction parameters is 2.3 m. The average precision of the IGSO and MEO satellites’ orbit real-time correction parameters is better than 8 cm.

In conducting dynamic monitoring for high-rise buildings using GPS technology, due to large height difference between base and rove stations, the tropospheric wet delay cannot adequately be eliminated by double difference mode. In this paper, a parameter estimation method based on random walk is used to further weaken the residual tropospheric delay.The corresponding data processing method is introduced in this paper. The experimental results show that the residual error of tropospheric delay cannot be ignored with dynamic monitoring for super high-rise building using GPS technology. Furthermore, the residual error of tropospheric delay is strongly dependent on satellite elevation, with the estimation and deduction for residual tropospheric delay relative to the traditional RTK positioning.The method introduced in this paper improves the real time dynamic monitoring precision for super high-rise buildings, especially if the elevation difference is more obvious.

The traditional three-dimensional space coordinate transformation model is restricted to solving the coordinate conversion parameters of small angles. This paper proposes a model that is suitable for the coordinate conversion of large angles by using dual quaternion that can simultaneously describe the rotation matrix and the translation vector. The model solves the problem of the traditional algorithm, and can directly calculate the coordinate transformation parameters without initial iteration value. Finally, a simulation experiment is carried out by the simulation data. The numerical example shows that the method does not need linearization and its calculation is simple and convenient. The correctness and effectiveness of it is also verified by the simulation results.

The new model of multivariate weighted total least squares by transposition processing, which is similar to the weighted total least squares model, is proposed in this paper. The formula for constructing the new coefficient matrix and its variance-covariance matrix is deduced and the solution flow multivariate weighted total least squares is studied. Applying this method, the solution process by the Jazaeri algorithm is deduced. The proposed method is proven to be effective and feasible through example analysis and comparison with other algorithms.

The post-seismic deformation characteristics of the MW9.0 Japan earthquake are studied based on GPS observation. The post-seismic horizontal displacements observed by GPS accumulated to 60~165 cm for the east component, and 20~65 cm for the south component, in nearly 5 years following the main shock, until December 2015. For GPS sites further away from the epicenter, the post-seismic displacements have exceeded their co-seismic changes. Moreover, the post-seismic deformation is still continuing. According to the theories of afterslip and viscoelastic dislocation, the post-seismic deformation is simulated and the GPS observations are explained by these two mechanisms. The simulation results indicate that afterslip plays a major role in the initial stage of post-seismic deformation, but it shows an exponential decay with time. On the contrary, the contribution of viscoelastic relaxation increases with time. The results also suggest that post-seismic deformation can be reasonably explained by combining these two mechanisms. In addition, the mantle viscosity in the seismic area is inferred to be of the order of 1.5×1019 Pa·s from post-seismic deformation observed by GPS. 

 Arctic sea ice freeboards during autumns and winters between October 2003 and March 2008 are determined using ICESat laser altimetry data; the results are consistent with foreign research. Trends derived from freeboard results show a decline of about 2.3 cm each year, which is faster than suggested by previous research. Obvious systematic error is shown in the comparison of ICESat derived Arctic sea ice freeboards obtained by different author and methods. This systematic error is analyzed. The results show that the window size of height filtering, the range of sea surface height determination, and the method of sea surface observation filtering, all lead to biases of several centimeters in freeboard calculation.

This paper uses data recorded by the BB channel of JCZ-1T ultra broadband seismometer at Korla seismic station. We use power spectral density estimation methods, acquiring the fundamental frequency spherical free oscillation 0S7～0S60  of the Indonesian, sea of Sumatra M7.8 earthquake of March 2, 2016 and the Ecuadorian M7.5 earthquake of April 17, 2016, without any correction pretreatment. We conclude that the measured values coincide with theoretical values. Except some modal, the error ratio is less than 0.1% in the rest, by comparing with PREM theoretical values. The conclusions suggest the significance of further research on the internal structure of earth and focal mechanism.

This paper uses sequence analysis and wavelet spectrum methods to discuss co-seismic deformation waves (CSDW) records characters and its conduct mechanism in Fujian-Guangdong area.The deformation data, based on 10 stations, includes 40 precursory anomalies in total. The result show that: (1) Arrival-time of CSDW is related to epicentral distance, the records include d-P and d-S waves, the records from different items are not consistent, and the same measured items are similar to the recorded waveform. (2) To a certain extent, the amplitude ratio shows the direction of the source, and further reflects the direction anisotropy of the amplitude ratio. (3) By calculating the cross(self)-correlation and filter processing, we can identify the CSDW bands in 0.02~0.04 Hz; the peak value is about 0.025Hz. (4) The wavelet spectrum band of Chile earthquake CSDW is mainly 64~254 s, the main energy is concentrated in 128~254 s, which is different thanthe Nepal earthquake CSDW. Different measuring points are different, the band of Chaozhou and Xiamen station is 16~32 s and 16~64 s, respectively. Results suggest the CSDW records include P wave and S wave elements except surface waves; in addition, the conduction mechanism of CSDW, which is caused by different focal mechanisms, may be anisotropic.This study has reference significance for the study of earthquake induced strain.

In this paper, typical methods of earthquake early warning and intensity reporting are analyzed, and an asymmetric sensor based seismic intensity instrument is proposed for earthquake early warning and seismic intensity rapid reporting. This instrument is composed by an asymmetric three-component accelerometer and a special data collector built by the ARM embedded system. Through tests, the special data collector has a dynamic range greater than 107 dB. In the asymmetric three-component accelerometer, the selected class B accelerometer holds a measuring error less than 0.8% and a dynamic range greater than 105 dB. The selected class C accelerometer holds a measuring error less than 5% and a dynamic range greater than 61 dB. Therefore, the instrument meets the technical requirements for the comprehensive application of earthquake early warning and seismic intensity rapid reporting. Also, the instrument has the characteristics of low cost and low power consumption, and is suitable for density applications.

 According to the working principle of the SS-Y tensometer,this article analyzes disturbance characteristics from air pressure change of Huaibei seismic station.The results show that air pressure is the main interference on the extensometer, and the relation between the SS-Y tensometer and air pressure is direct ratio. Through quantitative analysis, mathematical formulas are established to catch tensometer and air pressure change disturbance.