We attempt to quickly detect the fault screep deformation from the large GNSS observation network, and accurately invert the spatio-temporal distribution and evolution characteristics of faultcreep. GNSS network filtering, crustal deformation information, and spatio-temporal distribution inversion of fault creep are combined into one method. According to the whole GNSS network spatio-temporal observation array and the high space correction of fault deformation, and based on the principal component analysis of GNSS displacement spatio-temporal series covering the fault surface, the spatio-temporal distribution and evolution process of creeping fault can be detected and retrieved quickly by using principal component information. Taking the 2005 Sumatra M_W8.6 post-seismicslip and 2006 Guerrero slow slip event for examples, spatio-temporal inversion of fault creep is rapid realized. The results agree well with previous research.

By referring to the idea of weight fitting, inequality constraints are added to the integral rotation of regional crustal movement and the Euler vector part of the linear strain model. The Euler parameters of the regional crustal movement are constrained to the vicinity of the average Euler parameter value determined by the multi-year observation data. An overall rotation and linear strain model with inequality constraints is established. The fitting analysis of the velocity field data of recent GPS observations from crustal movement observation network of China in Bohai rim region shows that the proposed method enhances the smoothness of the overall motion trend of the block, with higher fitting accuracy to the velocity field as well as further enhanced physical interpretation ability of the model.

In this paper, by exploiting Sentinel-1A satellite ascending and descending data, we obtain a three-dimensional coseismic deformation field and source parameters by using synthetic aperture radar differential interferometry(DInSAR) and Okada uniform sliding model inversion. The results show that the fault slip caused by this earthquake is generally along the north-south direction, and the uplift occurs in the middle of the subsidence area on both sides, with the maximum vertical deformation around 18 mm. The seismogenic fault parameters of the Rongxian earthquake accords with the characteristics of the reverse fault, having length of about 9.5 km and width of about 2.1 km. The accumulative deformation in the study area from 2018-11 to 2019-03 is obtained using time series analysis technique. The results suggest that the deformation fluctuation in this area is stable before and after the earthquake, and the earthquakes are the main reason for the large accumulative deformation in the area.

In this paper, we use the equation of rotation in the entire block and the linear strain model(RELSM) of block motion to describe the internal deformation of the block. We use the robust estimation method with high breakdown point to screen abnormal stations. The station screening and fitting analysis of the velocity field data of GPS observations from crustal movement observation network of China in Bohai rim region are discussed. The result shows that the method can better balance the smoothness of the overall motion of the block and differences in local tectonic movements. Moreover, it can also suppress the influence of gross error, so as to obtain a better velocity field fitting result.

In this study, professional seismic interpretation and structural modeling software are used to comprehensively interpret stratum and fault in 2D shallow seismic active fault profiles. We utilize seismic data, borehole data and geological information from the Jiaozuo area. At the same time, the plane combination and spatial distribution law of faults are studied to realize the organic combination of geological and seismic factors, then the distribution model of the bottom interface of Q₃ stratum, Q stratum, N stratum and the Quaternary fault model are constructed. On the basis of these models, the Quaternary fault model is used as the basic framework, and the near-surface stratum distribution model is used as the main stratification, then the structural morphology and fault development of the Q-Q₃ geological body and the N-Q geological body are simulated. These geological bodies reflect the spatial structure characteristics of the bottom interface of Neogene stratum to Q₃ stratum.

Based on the observation data recorded by Capital circle earthquake network, the mainshock epicenter of the 2016-09-10 Tangshan M4.1 earthquake is accurately determined by the Hypo2000 method. The best double-couple solution of the plane Ⅰ of the mainshock is 122°, 60°, and -42° for strike, dip, and rake angle respectively; the plane Ⅱ is 236°, 54°, and -142°; the moment magnitude is M_W4.3, which are analyzed by local waveforms inversion method. Combining with local earthquake depth phase sPL, the focal depth is determined as 6 km. The aftershocks are relocated by hypoDD, the results show that the focal depths are in the range of 4~9 km, distributed within a straight zone, indicating that the seismogenic fault should be a NE-trending fault with high dip, the characteristic of which is consistent with plane Ⅱ. We infer that plane Ⅱ is the seismogenic fault plane, and plane I is the auxiliary plane. Compared the parameters of the seismogenic fault with the properties of local faults, it shows that the seismogenic fault of Tangshan earthquake should be the Tangshan-Fengnan fault.

The Coulomb stress change around the source fault induced by the 2008 Wenchuan M_W7.9 earthquake  is calculated based on the fine coseismic slip distribution model. The reliability of Coulomb stress change and the impact that the strike, dip and rake of receiver fault, the friction coefficient and Skempton coefficient towards the reliability of Coulomb stress change is assessed from the CV (coefficient of variation). The unreliable Coulomb stress change is eliminated according to the threshold value and the relationship between the reliable Coulomb stress change and the distribution of aftershock is analyzed. The results show that the most reliable Coulomb stress change is found inside positive and negative Coulomb stress change area, and the closer to the boundary between positive and negative, the more sharply the reliability decline. Among the parameters analyzed in this paper, the dip angle of receiver fault has the greatest effect on the reliability of Coulomb stress change. 79.3% of aftershocks locate in the distribution area of reliable Coulomb stress change and the aftershock triggering rate decreases by 3.1% after excluding the unreliable Coulomb stress change and aftershocks in unreliable area.

Based on GPS observation data from Gansu continuously operational reference system(CORS) and crustal movement observation network of China(CMONOC),  we calculate the vertical deformation of 48 stations. We perform a comparative analysis between GPS vertical displacement and GRACE vertical deformation calculated by CSR RL05M data. The results show local characteristics of vertical deformation in the study area. There is significant difference between eastern Gansu province and the other areas. Correlation coefficient, RMS reduction, and annual amplitude reduction are larger than the other areas. The mean correlation coefficient of GPS and GRACE vertical deformation time series without trend term is 0.72. The mean annual amplitude of GPS and GRACE are 6.00 mm and 3.70 mm respectively. Mean annual amplitude reduction and RMS reduction are 0.51 and 0.29.The results indicate strong consistency. GRACE vertical deformation can explain more than 50% of the annual GPS vertical displacement signal. An average of about 29% of the crustal deformation contained in GPS vertical displacement time series is induced by environmental loading.

Based on ordinary Kriging extrapolation method, the noise variance is estimated by frequency radially averaged power spectrum, and then the downward continuation of stable gravity field is calculated by the iterative Wiener filter. The simulation example and actual data are separately used to do experiments. The accuracy of simulation example is significantly improved after extrapolation, the mean square error decreases from 6.72 μGal to 1.89 μGal, and the relative error decreases from 9.26% to 4.44%. The continuation result of actual data after extrapolation reflects the shallower  anomaly information with high frequency better and avoids obtaining false values in the boundary.

In order to solve the interference problem of geomagnetic stations, this paper analyzes the characteristics and spectrum of different interference sources. It is found that the interference of multiple noise sources has no dominant frequency band and is highly nonlinear. In view of the fact that the geomagnetic observation signal is vulnerable to the interference of rail transit noise, instrument noise, high-voltage DC noise and driving pulse noise, this paper puts forward a new method based on a nonlinear kernel adaptive filtering algorithm which can adjust the filtering parameters independently under the excitation of external environment. Through the processing of the actual disturbed geomagnetic signal, it is verified that the nuclear adaptive filter has the characteristics of suppressing the common geomagnetic filter noise in the station and provides a new method for the geomagnetic interference suppression technology.

We use more than 5 years continuous data recorded with the observatory superconducting gravimeter(OSG-066) at Lijiang station, Yunnan province, China. High precision tidal parameters are obtained, and the local gravity tide model is established to provide a reference for accurate gravity tide deduction in Lijiang area. The correction effectiveness of gravity influence of ocean tide is computed with 9 new global tide models. In addition, the atmospheric gravity admittance is studied with regression analysis in time domain and wavelet analysis in frequency domain. The parameters of free core nutation are further estimated with the observed tidal parameters. The results show that gravity influence corrections of the 9 different ocean tide models are almost the same at Lijiang station, and correction accuracy of the tide model NAO.99b is slightly higher among them. Results of obtained atmospheric gravity admittance demonstrate that significant seasonal variation characteristics are found in both time domain and frequency domain; significant inverse correlation is found between precipitation data and atmospheric gravity admittance. The eigen period of free core nutation is 431.5(411.2, 454.0) sidereal days, while the quality factor is -2 796.

In the process of empirical mode decomposition(EMD) denoising, there is inaccurate problem when using the correlation coefficient criterion to determine the inaccurate problem of the intrinsic mode function(IMF) K value. This paper adopts the composite evaluation index(T), considering the curve smoothness(r) and root mean square error(RMSE) of two indices, presenting an improved EMD denoising method. We verify the method using 9 simulated data and measured GPS elevation time series data from two reference stations of CMONOC. The experimental results show that the composite evaluation index is more accurate than single correlation coefficient index in determining K value, which can more reliably identify the boundary point between noise and signal, and the noise reduction effect is better.

We select the sample points of the construction polynomial based on the dynamic circle radius selection method, and select three weight functions, including the local distance ratio, the inverse distance weighted, and weighted iteration method. We compare and analyze the accuracy and reliability of the local polynomial interpolation results. It is shown that the results of three methods are similar without gross error. When the gross error is added, the results of the first two methods deviate from the interpolated surface without gross error, and the undulation of weighted iteration method is consistent with previous result.

Aiming at the problem that the current GPS real-time clock bias determination algorithm is based on real-time observation data stream, which limits the reliability and integrity of the product to the network quality, and that the accuracy of broadcast ephemeris and IGU-P is low, we propose and implement a real-time clock bias determination algorithm based on the connection of hourly observation files. The algorithm is realized based on hourly observation files data. By combining clock bias estimation with adaptive ultra-short-term clock bias prediction, the real-time clock bias is determined. 30 days online result shows that, STD of GPS real-time clocks generated by the connection method is 0.25 ns, equal to IGS RTS, much higher than STD of broadcast ephemeris and IGU-P. Furthermore, this method can improve the reliability and integrity of GPS real-time clocks.

In order to improve the data processing efficiency of Bernese GNSS software, the intel math kernel library(MKL) is applied to Bernese GNSS software for precise orbit determination, and the computational efficiency of multiple MKL matrix inversion functions are compared with that of Bernese GNSS software. Experiment is analyzed based on Beidou/GNSS data of 200 stations distributed worldwide in March 2019, the results show that the time used by the step of pre-elimination is significantly more than that of matrix inversion, and the efficiency of data processing is not improved by using MKL. However, when the pre-elimination strategy is not applied, intel MKL can significantly improve the matrix inversion efficiency. The dpotri function has the highest efficiency in matrix inversion calculation, and the average computing time is 133 s, which is 13 times faster than that of Bernese GNSS software.

In this paper, we analyze observation data collected from iGMAS and MGEX reference stations based on the self-written data processing algorithm. The result indicates the outstanding performance of Beidou new signals. In the aspects of data integrity rate and signal-to-noise ratio, BDS-3 signals are weaker than GPS. In terms of the multipath error and pseudorange noise, BDS-3 signal is at the same level as GPS. Regarding ionosphere delay rate, BDS-3 signals are better than GPS. On the whole, the performance of Beidou-3 is equivalent to that of GPS.

Using the BDS differential code bias (DCB) products released by MGEX (multi-GNSS experiment), the variation characteristics of the monthly stability of BDS satellite DCB during high and low solar activity periods are compared and analyzed. The auto-regressive integrated moving average (ARIMA) model is utilized to predict the short-term BDS satellite DCB at different solar activity levels. The results show that during the high solar activity periods, the standard deviations of the daily solutions and the monthly stability of the BDS satellite DCB are significantly worse than those during the low solar activity periods, and the stability of the BDS satellite DCB varies with the BDS satellite constellation types. The time series prediction results of ARIMA are in good agreement with the MGEX-released DCB values, which is better than polynomial fitting method.

The higher-order ionospheric(HOI) delays for GPS, GLONASS, Beidou and Galileo in three MGEX stations at different latitudes in different solar and geomagnetic conditions are calculated. Moreover, effects of HOI on multi-GNSS precise point positioning (PPP) are studied and the impact of different solar and geomagnetic activity on the accuracy of PPP is demonstrated. The results show that higher-order ionospheric delay decreases with increasing latitude. Solar activity is the main factor to affect the higher-order ionospheric delays. The second-order ionospheric delay is less than 2 cm and the third-order ionospheric delay is about 1 mm at low solar activity. The second-order ionospheric delay is more than 2.5 cm and the third-order ionospheric delay reaches 5 mm at high solar activity. When a geomagnetic storm occurs, the delay has a certain degree of growth. After correcting the higher-order ionosphere delay, the improvement effect of positioning accuracy is better at lower solar activity than at higher solar activity.

There is a difference between the different orbit types and BDS-2/BDS-3 satellites observation values in the Beidou hybrid system, so this paper proposes a method, based on Helmert variance component estimation model of Beidou UERE, to optimize the stochastic model of the Beidou hybrid system. The single point positioning experiments based on the equal weight model, the ordinary Helmert variance component estimation model, and the proposed optimization model are carried out. The calculation results of the three models are compared and analyzed with respect to the positioning accuracy and the positioning success rate. The results show that the proposed optimization model has high positioning accuracy on the E, N and U directions compared with the equal weight model and the ordinary Helmert variance estimation model. It can reduce situations in which the positioning accuracy is low due to the poor spatial geometry of the satellite constellation, and it improves the positioning success rate.

This paper combines the theoretical step signal with the polynomial fitting method to propose a method to eliminate the step calibration signal in the seismometer record. Comparing the power spectral density (PSD) of the micro-tremor signals at different times, we find that the signal which is eliminated by the proposed method is close to the PSD of the real signal. At the same time, all phases of the step calibration signal eliminated by this method are clearly restored.

The step change of the atmospheric pressure coefficient of Qingdao station in a short period(less than 128 min) occurred in January 2018. Through on-site verification analysis of the observation system, construction situation near the station and observation environment, the two factors of observation system and construction near the station are excluded. Using discrete wavelet transform and regression analysis, the paper finds that the variation of atmospheric pressure coefficient of borehole water level and volumetric strain is synchronous. To determine the reason for the variation of atmospheric pressure coefficient of the volumetric strain, the paper studies the water level variation of volumetric strain borehole on the day of drilling construction, the trend variation of volumetric strain of Qingdao station, and variation of atmospheric pressure coefficient of water level at other adjacent seismic stations. The results show that drilling construction is the cause of variation of atmospheric pressure coefficient of volumetric strain and the change mechanism is qualitatively analyzed.