Research has showed gravimetric satellite GRACE can detect the coseismic geoid and gravity changes induced by a megathrust earthquake. In this paper, we use a de-correlation filter of DDK3 to smooth the GRACE RL06 data released by Center for Space Research(CSR). By using time series analysis based on least square approach, we successfully extract the coseismic gravity, geoid, vertical deflection and gravity gradients changes caused by the 2004 Sumatra M_W9.3 earthquake, the 2010 Chile M_W8.8 earthquake and the 2011 Tohoku-Oki M_W9.0 events. The results show that the range of coseismic gravity of the three earthquake cases are -15.5 to 6.5 μGal, -9.1 to 2.1 μGal and -11.1 to 4.2 μGal, respectively. The range of coseismic geoid is -5.9 to 0.8 mm, -3.0 to 0.8 mm and -3.2 to 0.5 mm. The variation range of north-south vertical deviation is -1.2 to 2.2 mas, -0.9 to 1.0 mas and -1.1 to 1.4 mas. The east-west variation ranges are -1.8 to 1.0 mas, -0.8 to 0.8 mas and -0.7 to 1.0 mas. Among the components of the gravity gradient, the coseismic variation of rr component is the largest, followed by the variation of the rθ component. The spatial distribution of coseismic signals of the three earthquakes is as follows, the coseismic geoid and gravity change signals are distributed asymmetrically at two poles. The vertical deviation is negative-positive-negative or positive-negative-positive three-pole distribution; the signal of gravity gradient change presents a complex multipole distribution. Our GRACE coseismic signals agree well with some theoretical models better than others. The time series analysis method in this paper is robust and can be used to obtain the co-and post-seismic geodetic deformations and gravity field changes which is often used to constraint the fault slip and Earth’s viscosity structure.

Based on GRACE-FO data of January 2019, using the dynamics method with 24 hours arc length, we analyze results of orbital fitting and time-variable gravity field recovery by adopting different accelerometer parameterization strategies. The results show that: the results of orbital fitting are related to the number of parameters in the accelerometer calibration, and the results would be better with the increasing of accelerometer calibrated parameters. Further, there is not the conspicuous improvement in fitting orbit whatever the strategies are used in directions of the three-axis accelerometer. However, the performance of results of the recovered gravity field is influenced by the number of parameters in the accelerometer calibration, and present some differences if we take different strategies at different directions. Through the cases we investigate, the result would be better if we estimate more accelerometer parameters on the Y direction in SRF frame than other cases, especially, at degrees greater than 40.

Based on basic data, such as the installation log of Malingshan seismic station, combined with field survey results, we make a comparative analysis of the apparent resistivity data of the station in the past 10 years using the three-dimensional finite element method and combining auxiliary observation data such as water level, rainfall and air temperature. Because the station is located in Anqiu-Juxian fault of Tanlu fault zone, the underground structure is relatively complex. There is a height difference between the measuring electrodes at EW oriented, and the apparent resistivity characteristics are clear and complex. The comprehensive study shows that the NS apparent resistivity change of the station is mainly affected by the temperature, the EW apparent resistivity is mainly affected by the water level change, and the water level influence magnitude is about three times that of the temperature. In July and August, the two directions are affected by the short-term rapid rainfall at the same time.

In this paper, we use re-initialization of all satellites, additional ionospheric constraints and additional coordinate constraints to solve the above re-convergence problem. Based on real-time observations of NOTA station network, we test the three methods. The results show that they effectively improve positioning accuracy. In seismic monitoring, appropriate methods can be selected according to the actual situation to obtain high-precision real-time dynamic PPP processing results, thus better capturing seismic signals and improving the capability of seismic monitoring.

In this paper, the GNSS raw Doppler observation and the carrier phase derived Doppler observation are used to verify the performance of single antenna attitude determination in a kinematic environment, and we compare the accuracy of the attitude determination for the kinematic platform in different motion states. The experimental results show that when the kinematic platform is moving at a lower speed, the RMS values of the heading angle calculated by the carrier phase derived Doppler observation, the raw Doppler observation are both within 0.35°, and the RMS value of the pitch angle is within 0.15°. The attitude result calculated by the carrier phase derived Doppler observation is more stable than the raw Doppler observation. The heading accuracy calculated by using the carrier phase derived Doppler observation when going straight and turning are within 0.15° and 0.81° respectively, and the accuracy of the calculated attitude results when the carrier is going straight is higher than when turning.

Because the gross error in the time series of GNSS is a three-dimensional position gross error, this paper proposes an iterative least-squares gross error detection method on 3D coordinates. According to the experimental simulation data, the gross error detection in one direction and in three directions of north, east, and up is compared with the 3σ criterion and the interquartile range (IQR) criterion, respectively. The results show that the effect on three directions is much better than that in one direction. And the detection efficiency of three directions detection is beyond 95%. The 3σ criterion detected more gross errors than the IQR criterion, but the misjudgment rate is higher. Application in real GNSS time-series data shows that gross errors in the raw type of some global IGS stations provided by SOPAC (Scrips Orbit and Permanent Array Center) range from 0.1% to 4.7% of the station data. Time series detected by IQR gross error in three directions is cleaner than its cleaned trend type.

Aiming at the characteristics of complex noise and poor precision of GNSS vertical coordinate time series, we use the locally weighted regression(LOESS) method to denoise the vertical coordinate time series of 289 GNSS stations in the crustal movement observation network of China. Firstly, we use the LOESS method to denoise the preprocessed time series, and obtain the denoised time series and noise series. Then, we use the Durbin-Watson test for autocorrelation of denoised series, and the Wilcoxon rank sum test method for significance test of the standard deviation, noise term and velocity uncertainty of the series before and after denoising. Finally, we use the signal-noise ratio and the above three indicators to quantitatively evaluate the denoising effects. The results show there is no autocorrelation in the noise series of each station after denoising, and each evaluation index has significant correction after noise reduction by using LOESS method. LOESS method can effectively reduce the noise of GNSS vertical coordinate time series, and improve the accuracy of the GNSS vertical coordinate time series.

We propose a new strategy of SLR data short-arc analysis of ETALON satellites in this paper, which also solves satellite orbit parameters, earth rotation parameters and range bias parameters. The new strategy is verified using the data from January to October in 2018. The results show that: the root mean square of residuals for ETALON-1/2 satellites is 1.11cm and 1.08cm respectively; compared with IERS-C04 product, the precision of X_p,Y_p and LOD parameters is 2.21 mas, 2.26 mas and 218.30 μs/d respectively; compared with the final orbit of ILRS, the orbit accuracy of ETALON-1 satellite in radial, track and normal direction is 1.6 cm, 8.5 cm and 6.8 cm respectively, while that of ETALON-2 satellite is 2.1 cm, 8.9 cm and 8.7 cm respectively.

Integer ambiguity is affected by atmospheric delay error and it is difficult to correctly fix integer ambiguity for low elevation angle satellite between BDS reference stations. We propose the algorithm of triple-frequency ambiguity resolution for low elevation angle satellite between BDS reference stations. The advantage of long wavelength for extra-wide-lane integer ambiguity is used to determine double difference extra-wide-lane integer ambiguity. The double difference wide-lane integer ambiguity is searched and solved using the closer characteristic of ionospheric delay error between wide-lane combination observations. The linear relationship of double difference integer ambiguity is selected as a constraint condition, double difference integer ambiguity for high elevation angle satellite is searched and determined using dual-frequency ionosphere-free combination observation. The space linear model of double difference ionospheric delay error between reference stations is established using fixed double difference integer ambiguity for high elevation angle satellites, and the ionospheric delay error for low elevation angle satellites is weakened. Double difference carrier phase observation equation with fixed double difference integer ambiguity for high elevation angle satellite is selected as distance constraint, double difference integer ambiguity for low elevation angle satellite is searched and determined. This algorithm is tested by the measured triple-frequency data of BDS CORS network. The results indicate that triple-frequency integer ambiguity for low elevation angle satellites between BDS reference stations can be fixed accurately with the method.

In order to explore the optimal combinations for BDS long-baseline ambiguity resolution, we select two optimal combinations of extra-wide lanes and an optimal combination of narrow lanes of BDS based on fuzzy clustering analysis to jointly construct the TCAR model, and use three models of GF_TCAR, GB_TCAR and GIF_TCAR to resolve the ambiguity of the measured long baseline for BDS. The experimental results show that all three models equally solve the combination ambiguity results of extra-wide lanes. Using GF_TCAR to calculate the adjacent epoch ambiguity difference of the optimal narrow-lane combination (3,0,2) obtained by fuzzy clustering analysis method is better than that of the commonly used combination (4,2,5). When using GB_TCAR to solve IGSO satellites, the narrow-lane combination (3,0,2) is superior to the combination (4,2,5), but the results of the two narrow lanes are consistent when solving GEO and MEO satellites. The narrow-lane combination (3,0,2) calculated by GIF_TCAR is slightly inferior to the combination (4,2,5).

Aiming at the characteristic that Beidou-3 satellites can broadcast four-frequency signals, we propose a real-time cycle slip detection and repair method. The method combines three geometry-free phase combinations and one code-phase combination. Firstly, the high-quality geometry-free combination under multi-frequency conditions is discussed, and then three geometry-free phase combinations and one code-phase combination are selected to detect the cycle slips, combined with the LAMBDA method to repair the cycle slips in real time. Finally, we use the measured data under the different observation conditions to verify the effectiveness of the proposed method. The results show that the method is less affected by ionospheric delay than three-frequency condition, and can detect and repair the cycle slips in real time.

This paper provides an adaptive waveform re-tracking algorithm based on the primary peak (AMWRT). First, the primary peak is extracted, and second, the primary peaks are retracked by offset center of gravity (OCOG) method and threshold method. Finally, the retracked results are compared to obtain the best observed water level. The AMWRT, OCOG and threshold methods are applied to the middle reaches of the Yangtze river for comparison. The results show that: 1) The long-term water level sequence observed by altimetry satellites is more consistent with the trends measured by the hydrological station; 2) When the AMWRT algorithm is compared with OCOG and threshold, the AMWRT method performs well, and the accuracy is improved by 30% compared to the best-performing classic threshold_0.5 method. The absolute mean error (MAE) and root mean square error (RMSE) are reduced to 0.55 m and 0.72 m, respectively.

Based on the existing tropopause empirical model, we discuss the influence of regional relative elevation on T_m model. Based on this, the h₀T_m regression model of China region and the regional model h₀T_m-Qz of Qinghai-Tibet plateau region are established. The results show that: 1) Taking ERA5 grid data as reference values, the root mean square error (RMS) of h₀T_m model is 2.43 K, which is 1.15 K (32%) and 0.63 K (21%) higher than that of Bevis formula and GPT2w-1 model, respectively; 2) taking radiosonde data as reference values, the RMS of h₀T_m model is 2.48 K, which is 1.19 K (32%) and 2.06 K (45%) higher than that of Bevis formula and GPT2w-1 model, respectively especially in western China, the h₀T_m model shows more significant advantages; 3) the regional model h₀T_m-Qz of Qinghai-Tibet plateau, which takes into account the relative elevation of the region, is better than the single factor (surface) model in this region. The accuracy of the single factor (T_s) regional model T_sT_m-Qz and Bevis formula is improved by 0.54 K (19%) and 2.50 K (51%), respectively.

Using the acceleration waveform records of 214 strong seismic stations in Beijing-Tianjin-Hebei region captured by the M5.1 earthquake in Guye district of Tangshan city on July 12, 2020, we analyze the variation characteristics of the records number with the epicenter distance and draw the contour map of peak acceleration. We compare the linear elastic response spectrum recorded by 6 soil stations with the designed spectrum of China, and analyze the amplification effect of soil layers on ground motion using the traditional spectral ratio method. By comparing the observed data with some commonly used GMPE(ground motion prediction equations), we analyze the attenuation relationship of PGAs with different distance. By calculating the normalized residuals between predicted and observed values, we discuss the influence of deep overburden soil and soft soil on ground motion. We study the effects of local site conditions on the amplitude and main period of ground motion, plot the spatial distribution of the spectral acceleration with different period, and explain why the ground motion at medium distance is significantly magnified combined with the site environment.

The purpose of this paper is to study the engineering and time-frequency characteristics of ground motion in different regions of near-fault. Considering the relationship between the characteristics of near-fault ground motions, such as the rupture directivity effect, fling-step effect, hanging wall effects, and fault types, the near-fault ground motion of Chi-Chi earthquake in Taiwan is selected as the research object from the database PEER. Its near-fault ground motion is divided into the rupture forward region(RFR) ground motion, the hanging wall region(HWR) ground motion, the foot wall region(FWR) ground motion, and the rupture backward region(RBR) ground motion, according to the relative positions relationship between the fault rupture and the ground motion observing station. Further, we compare and analyze the characteristics of three elements of four types of near-fault ground motions, and analyze the difference of energy distribution in time-frequency by the Hilbert-Huang transform. The results show that: 1) The seismic characteristic parameters such as PGA, PGV, PGD and PGV/PGA of RFR, HWR, FWR and RBR are significantly different, leading to great influence on the seismic response of structures. 2) The ground motion in RFR region shows significant long-period effect, the ground motion in RBR region has the characteristics of high-frequency and large-scale value, and the ground motion in HWR and FWR shows the characteristics of high-frequency large-scale value and significant long-period effect. 3) The high values of ground motion energy in RFR, HWR,FWR and RBR are mainly concentrated in 0 to 5.2 Hz middle and low frequency band, among them, the maximum instantaneous energy in HWR is the largest, and the maximum instantaneous energy in RFR is the smallest.

We use ground-based RAR mode and ground-based SAR mode to monitor longitudinal vertical deformation of railway bridges as the train passes by. The experimental results show that: 1) the deformation trends and deflection are basically consistent between the two modes. Moreover, the RAR mode and SAR mode corresponding monitoring point of longitudinal vertical deflection deformation precision maximum deviation is 1.85 mm. Through external verification and analysis, both RAR and SAR modes can realize millimeter level deformation monitoring of the bridge. 2) RAR mode has fast and simple data processing, which has real-time processing and visualization capability. SAR mode has a relatively large amount of data, so it is technically difficult to process the data, but it can reflect the trend of planar deformation at the same time and has the ability to capture deformation of the object. SAR mode can also extract the deformation of feature point, lines and planes.

In this paper, two same type very broadband seismometers are installed in the seismic observation station three kilometers away from the subway. One is equipped with a shielding case, the other is not. By calculating the acceleration power spectrum density of continuous 24 hours observation data, we evaluate the influence of subway electromagnetic field on the observation of very broadband seismometer. The results show that the influence of subway electromagnetic field on the 3 components of the very broadband seismometer is mainly in the frequency band of 10 to 1 000 s, affected mostly in the vertical direction, followed by the horizontal direction.

Considering the limit of the PCA method in detecting ionospheric anomalies, we propose a detection method that combines PCA and sliding time window. We use three strong earthquakes as examples to test its anomaly detection capability. The results show that this method retains the advantages of each of the two methods. Compared with the sliding time window method, its detection results are more intuitive and concise, and have stronger resistance to environmental disturbances as well as relatively higher confidence and reliability of anomaly detection. Compared with the PCA method, it can be better used for long-time anomaly detection and accurately distinguish between positive and negative anomalies. On this basis, this paper analyzes the anomalous perturbation of the pre-earthquake ionosphere for the 2020 Oaxaca earthquake using this method. The study showed that the TEC anomalies increased 13 days before the earthquake, and decreased 12, 11, 9, and 6 to 4 days before the earthquake in the ionospheric TEC, which is located in the southerly region of the epicenter and mainly manifested as negative anomalies. Finally, combining these four earthquakes, it can be found that the ionospheric anomalies mainly appear in the southerly direction of the epicenter, and most of them are conjugate structures.

We calculate the strain parameters of hourly point values, daily average values, 5-day average values and 30-day average values, based on five sets of four-component borehole strain observation data installed in the Xinzhou area of north Shanxi province. We calculate and analyze the tidal factors of borehole strain observation data by harmonic analysis of Venidikov. The results showed that the rate of change of the maximum principal strain ε₁, minimum principal strain  ε₂, surface strain ε_a, and shear strain ε_s of the five measuring points maintain a relatively constant level. The principal direction φ is unchanged roughly, and the rose map of tidal factor is also almost unchanged, indicating that the strain field is relatively stable. The results of the study are basically consistent with the characteristics of current seismic activity, focal mechanism solutions and fault properties in the area.