The shadow matching algorithm can improve the GNSS positioning accuracy of mobile terminal in urban canyons, but the signal-to-noise ratio (SNR) threshold used to judge the visibility of satellites in traditional shadow matching algorithm is a fixed single threshold that cannot ensure the correct judgment of the visibility of all satellites, which affects the positioning accuracy. To the above situation, we propose a shadow matching algorithm based on dynamic threshold. The judgment threshold of each satellite changes with the mutual correction of geometric visibility and SNR visibility. Through the calculation and comparative analysis of two shadow matching algorithms for mobile phone GNSS original positioning results and Kalman filtering positioning results, we prove that the average point error of dynamic threshold method can be reduced by 15.1% in standard single point positioning and 22.4% in Kalman filtering compared with single threshold method, and the positioning accuracy has been improved.

The coefficient matrix of the error-in-variables (EIV) model is ill-posed, and the precision of the coefficient matrix and the observation are not equal, so we derive the regularized solution of the ill-posed weighted total least squares model using the Lagrangian multiplier method. We prove the existing regularized solution of the ill-posed total least squares model to be a special case of the new model. On this basis, we propose the regularized robust solution based on the median method for ill-posed weighted total least squares model, and the effectiveness of the new algorithm is verified by two examples of the first kind Fredholm integral equation and ill-posed trilateration network. The results show that the least squares solution and the total least squares solution have poor accuracies and seriously deviate from the true value due to the influence of ill-posedness and the outliers of the model, while the accuracy of the regularized solution has been improved for weakening the ill-posedness of the model taking into account the errors of coefficient matrix and the observations. On the basis of regularized solution, the regularized robust solution reconstructs the weight matrix using the equivalent weight function, which can effectively resist the influence of gross error and has the highest accuracy.

On the basis of linear regression model, considering the measurement errors of independent variables and dependent variables at the same time, this paper first unifies many models to be selected into the linear regression model with constraints, adopts the hypothesis testing theory with multiple alternative hypotheses, and then constructs the hypothesis testing statistics with Lagrange multipliers. We propose the optimum selection algorithm of the linear regression model. The experimental results show that the proposed algorithm can obtain the optimum linear regression model, which is in accordance with actual observations and simpler than the improved linear hypothesis method.

According to the climate type, the mainland of China is divided into five regions, and a number of sites are randomly sampled from each region for analysis. Based on the fast Fourier transform(FFT) method, we use GNSS data from crustal movement observation network of China(CMONOC) in 2016 for characteristics analysis of PWV in short term frequency domain, and extract GNSS PWV cycle characteristics of different seasons. The results show that the characteristics of PWV frequency domain have obvious regional variation and seasonal difference. There are obvious periodic changes in the GNSS stations of plateau mountain climate, tropical monsoon climate and subtropical monsoon climate. The PWV amplitude is higher in the tropical monsoon, subtropical monsoon and coastal regions, and it is lower in plateau mountainous and temperate continental regions.

We use the tropospheric products of 224 IGS sites with high-precision around the world in 2019 to validate and analyze the accuracy of the zenith tropospheric delay and atmospheric horizontal gradients information estimated by the GPT3 model. The results show that differences appear in spatial and temporal distribution, and the mean bias and root mean square error of zenith tropospheric delay, horizontal gradient in south-north and west-east direction are 0.12 mm/37.2 mm, -0.05 mm/0.49 mm and 0.08 mm/0.57 mm, respectively.

We carry out precise orbit determination research of the GRACE-FO satellite, and process the 14-day on-board GPS data by the reduced-dynamic method, and perform accuracy analysis. For the comparison of the reduced-dynamic orbit with ambiguity resolution parameters by UPD method and the float solution orbit, we carry out carrier phase residual analysis, overlapping arc comparison, reference orbit comparison, and KBR check. The results show that the carrier phase residual of the fixed solution orbit of the GRACE-FO satellite is about 1 cm, which is 1 to 2 mm larger than the float solution. The RMS of 3-day fixed solution overlapping arc difference is less than or equal to 9.4 mm in the radial, tangential and normal directions, which is better than float solution of 12.4 mm. Compared with the precision scientific orbit provided by GFZ, the GRACE-C satellite reduced-dynamic fixed solution orbit has a mean RMS of less than 1 cm in all directions, which means that the calculated orbit has a high consistency with PSO. The residual RMS value of the KBR check after ambiguity resolution decreases from 9.6 mm to 6.7 mm, which indicates that the fixed solution can further improve the relative position accuracy between GRACE-FO satellites. The ambiguity resolution can improve the orbit determination accuracy of GRACE-FO satellite and provide more reliable orbit service.

We summarize the typical gravity changes (gradient zone and four quadrants distribution) and their quantitative parameters before seven strong earthquakes in the mainland of China. The results show that the earthquake prediction indicators used currently have a certain degree of uncertainty related to the observation time, the spatial distribution of stations and the spatial difference of variation amplitude. The spatial range of tidal factor anomalies before the Lushan earthquake is nearly 10 times the typical variation range of mobile gravity, which may be related to the fact that the measuring precision of continuous gravity stations are one order higher than mobile stations. Based on the locked shear force model, we propose a method for earthquake numerical prediction using the typical gravity changes. Before strong earthquakes, four-quadrants gravity changes on surface caused by strike-slip couple, or two-quadrants (gradient zone) gravity changes on surface aroused by dip-slip couple will appear under the continuous pushing of the locked double-couple shear force existing at the pregnant earthquake source. Preliminary results of the retrospective prediction cases show that the method is effective in predicting the magnitude of strike-slip and dip-slip earthquakes and can also give out the location and type of the future earthquakes.

The global geomagnetic field model EMM2017 is used to calculate the global distribution of the difference between the total magnetic field intensity anomaly (ΔT) and the projection component (T_ap) of the magnetic anomaly vector (T_a) in the direction of the main magnetic field, and the amplitude, spatial distribution characteristics and influencing factors of the difference are analyzed. This study can provide reference for the construction and application of the global lithospheric magnetic field model.

The time, magnitude and epicenter of the Tangshan M5.1 earthquake on July 12, 2020 are all consistent with the prediction of China electromagnetic mid-year consultation on May 22, 2020. Before the earthquake, according to the “practical handbook of earthquake electromagnetic analysis and prediction methods” and the “manual on technical methods of earthquake electromagnetic analysis and prediction”,  the electromagnetic group analysed the data of electromagnetic stations in the mainland of China. The results show that there are many anomalies such as daily variation of geomagnetism, harmonic amplitude ratio, and geoelectric field at the electromagnetic stations in the capital circle. The comprehensive prediction before the earthquake is basically consistent with the actual earthquake. The prediction index system construction of geo-electromagnetic discipline from 2016 to 2019 has achieved initial results.

There are single point missing and continuous multi-point missing in geomagnetic data recorded by seismic instruments, which is not conducive to seismic data processing and earthquake prediction. In order to quickly process the non-seismic abnormal data, we propose a time series autoregressive moving average (ARMA) prediction model for geomagnetic data interpolation processing. We compare the ARMA model with mean interpolation and linear interpolation. The results show that the mean standard errors of missing single point of mean interpolation, linear interpolation and ARMA model are 0.110 2, 0.006 9 and 0.000 1, and the mean standard errors of missing continuous multi-point are 0.258 23, 0.194 2 and 0.004 86, respectively. The results indicate that ARMA model has a low standard error in single point missing and continuous multi-point missing, which can well maintain the curve shape of the actual observation sequence, and the interpolation effect is better. It is expected to become a new method of geomagnetic data sequence processing.

In this study, the aftershocks of Gaoling M_L4.8 earthquake occurred on November 5, 2009, in Shaanxi province are relocated by the HypoDD, and the focal mechanism solution of the main shock is determined by the CAP method. Based on the results, we analyze the characteristics of seismogenic structures. The results show that the depth of main shock is about 6 km, and the aftershocks are mainly distributed in the depth range of 2~12 km. Seismic sequences are distributed in clusters on the plane and almost vertical on the depth profile. The focal mechanism solution of the main shock are strike 103°, dip 61° and rake -74°, and strike 252.40°, dip 32.78° and rake -116.44°, respectively. The centroid depth is about 7 km, the moment magnitude is about 4.2, and the P-axis direction is northeast. The focal mechanism results show that the main shock is a normal type earthquake with slightly strike-slip movement. According to the results of relocation and focal mechanism solution, it is suggested that the intersection of Weihe fault and Weinan-Jingyang fault is an important potential seismic area.

In this paper, we collect digital waveform data both from mobile stations and permanent stations, and study the shear-wave splitting characteristic of Wulong M5.0 earthquake sequence based on SAM method. We obtain 94 pairs of shear-wave splitting parameters from 5 stations. The results show that the fast polarization directions (FPDs) are generally NW direction, which is obviously controlled by regional principal compressive stress field. The variation of the dominant FPDs from NW to NNW is affected by several NE-trending faults and complex structures in the region; the dominant FPDs of stations in the east and west of the mainshock show a certain spatial distribution change. The FPDs are scattered in the early stage after the mainshock and become stable in the late stage, indicating the regional stress tends to stabilize in the late stage. The delay time basically increased before strong aftershocks. However, the delay time decreased in a short time before some strong aftershocks. It is coincides with the long-term stress accumulation and short-term stress relaxation before earthquakes. There is no obvious relation between the delay time and the focal depth．

Based on the artificial shallow seismic reflection profile, we find that the Baodi fault in Jingzhuang area of Xianghe county is a nearly EW trending normal fault with a spade-shaped profile. It is an active Quaternary fault and controls the deposition of Quaternary. Judging from the comparison of structural observation drilling cores on both sides of the fault and the data of ¹⁴C dating and OSL dating, the up-breakpoint of Baodi fault is about 5-7 m away from the surface, its main activity occurred in Late Pleistocene, and there is no obvious activity in Holocene and late Middle Pleistocene. The comparison of the cores shows that in the last 100 000 years, the Baodi fault has been seismically active twice, with a cumulative displacement of 2.6 m. The recent earthquake occurred at 22 740±70 a BP with a co-seismic displacement of 1.4 m; the other earthquake occurred between 97.1±4.4 ka and 98.0±5.3 ka with a co-seismic displacement of 1.2 m. According to the drilling research, combined with previous research results on fault activity and seismogenic laws in the eastern section of Zhangjiakou-Bohai tectonic belt, we conclude that the Baodi fault has been inactive since Holocene, and the possibility of inducing moderate and large earthquakes is low. In addition, the Jingzhuang ground fissures in Xianghe county are structural ground fissures controlled by the Baodi fault. The shearing stress caused by the uneven ground subsidence on both sides of the Baodi fault also accelerates the development of the Jingzhuang ground fissures.

Based on the Sentinel-1A image data of 35 scenes covering Shanghai from 2018 to 2020, we use the PSInSAR method to obtain the annual average settlement rate and cumulative settlement of three metro lines in Shanghai, and analyze the reasons. The results show that: 1) The uneven settlement occurs in Shanghai, which mainly occurs near Xujing village, Minhang district in the south of the main urban area of Shanghai, with the maximum cumulative settlement of -47.3 mm. 2) The obvious settlement funnel appears in Minhang district of Shanghai, which is positively correlated with the urban infrastructure construction and groundwater level changes in this area. 3) There are some section of Shanghai metro lines 5, 8 and 16, with obvious settlement phenomenon, and the settlement around metro line 5 passing through Minhang district is the most obvious, with a maximum settlement rate of -12.0 mm/a. This study shows the spatio-temporal characteristics of surface settlement and the relationship between surface settlement and its influencing factors in Shanghai and along the main metro lines.

Based on the gravity recovery and climate experiment (GRACE), GPS and absolute gravity data from 2003 to 2013, we discuss and analyze the crustal deformation in the southern margin of the Qinghai-Tibet plateau. The result shows the LHAZ station rising at a rate of 1.5 mm/a from 1999 to 2015, which indicates that the LHAZ station has always been uplifting. We analyze the vertical time series of four stations (LHAZ, TPLJ, CHLM, JMSM) of the remote continuous GPS station package (RCGSP) constructed by southern California integrated GPS network (SCIGN). The GPS results show that the four stations have an average uplifting rate of 2.3±0.11 mm/a from 2003 to 2013. Meanwhile, the vertical displacement time series indicates that the four stations have a slight upward trend from 2003 to 2013, with an average rate of 0.35 mm/a. Combined with absolute gravity data, considering many factors (crustal uplift, erosion, GIA), the gravity variation of LHAZ station in recent years is about -0.97 μGal/a and the rate of crustal bottom thickening at LHAZ station is about 4.4 cm/a, which reveals the characteristics of crustal deformation in the southern margin of the Qinghai-Tibet plateau: crustal uplift and bottom thickening.

According to the principle of centrifugal force generated by the object rotating around the origin and the synthesis and counteraction of centrifugal force, two servo motors are used to drive two eccentric masses to rotate at the same time, and linearly change the rotation frequency of the servo motors through the Beidou timing control system, so as to develop a controlled accurate seismic source (CASS), which launches a linear frequency modulation (LFM) signal to the ground. The energy of LFM signal has good aggregation characteristics. We calculate the initial frequency and frequency modulation rate of LFM signal by Wigner-Hough transform. CASS is a kind of green source with long-term operation and high repeatability. It can be applied to monitor the fault activity, receive the LFM signal reflected to the ground from subsurface interface at different depths, quantitatively analyze the temporal variation law of the wave velocity on the same path, study the relationship between spatio-temporal evolution characteristics of the fault activities and the occurrence of earthquake, and explore the numerical earthquake prediction.

Through the method of keywords co-occurrence analysis, we carry out an analysis of the development and application status of deep seismic sounding (DSS) technology in China. The results show that the data processing of deep seismic sounding is mainly based on the forward modeling method of 2D dynamic ray tracing, while the 3D crustal model is mainly constructed by Kriging interpolation. DSS method can reveal the seismogenic tectonic, but rarely summarizes the relationship between the regional crustal structure and earthquake distribution and activities. In terms of seismic sounding results, including the type of crustal structure and crustal thickness, are closely related to the formation of geothermal, metal, and oil and gas mineral resources. However, there are no available DSS database or data products for mineral resources exploration. Ocean deep exploration has become a focused area in the application of DSS method. The onshore and offshore deep seismic experiment should be strengthened to give full play to the advantages of the two different observation modes.

In order to improve the calculation accuracy and efficiency of the seismic response spectrum, we introduce the linear recursive filtering method(LRFM) to calculate the seismic response spectrum. We derive the general expression of the response spectrum, which is based on the single degree of freedom dynamic system equation. To verify the calculation accuracy and efficiency of this method, we use the synthetic sinusoidal simple harmonic wave as the system input to compare the results of response spectrum which are calculated by the LRFM, the Duhamel step-by-step integration method, the Newmark-β method and the analytical solution method (ASM), and verify the error comparing with the ASM. We prove that the calculation method of the LRFM has good stability and high efficiency. In order to further prove the effectiveness of this method in calculating the actual data, we select the acceleration data of strong earthquakes with different prominent periods and spectrum from the ESM strong earthquake database, and the response spectrum of the LRFM and ASM is compared and analyzed under different damping conditions. The results show that the LRFM is consistent with the ASM in calculating the velocity response spectrum. The calculation result of the acceleration response spectrum has some errors with the increase of the damping, but the overall calculation accuracy could meet the requirements. The LRFM proposed in this paper could quickly and efficiently calculate the response spectrum information, so it has great significance for the rapid assessment of seismic intensity characteristics of the site and the stress condition of engineering structure.

We study the moon phase modulation effect of small earthquakes above M1 in Shandong inland area since 2008, combining with the variation of seismic b value in this area. The results show that there are anomalies of low b value and high modulation ratio in the area before the 2020-02-18 Changqing M4.1 earthquake, and we calculate the deflection phenomenon of P-axis of the composite nodal plane solution, which is very consistent with the variation period of low b value of small earthquakes. In addition, before the 6 earthquakes above M4 in this area, the abnormality of modulation ratio ≥0.33 occurs 4 times. In 2011 and 2016, the two Fanxian earthquakes did not show the above changes, which may be related to the activity mechanism of oil recovery in Fanxian small earthquake concentration area.