Since the KBR phase center and the installation matrix of the star sensor may change after the satellites enter the orbit, there will be a systematic bias of the inter-satellite pointing of GRACE-FO satellite. In order to correct the bias, it is necessary to uplink the inter-satellite pointing calibration parameters into the satellite attitude and orbit control system. In this paper, we derive the correlative algorithms of inter-satellite pointing calibration and accuracy evaluation, and propose the process and implementation strategy of inter-satellite pointing calibration based on in-flight path. Based on the post-processing path, we evaluate the inter-satellite pointing performance of GRACE-FO since its launch. The STD and RMS of the inter-satellite pointing angle indicate that the GRACE-FO inter-satellite pointing meets the requirements of pitch and yaw direction 1 mrad and roll direction 10 mrad for the normal operation of KBR. The magnetic torque of the satellite attitude and orbit control system and the attitude control thruster can meet the requirements of attitude control.

To establish the optimal strategy for the Beidou precise orbit determination receiver PCC(phase center correction) model, we first compare the IGb R3 and IGb 14 calibration values of GPS, and the small difference between the two sets of model values demonstrates the feasibility of using the IGb R3 instead of the IGb 14 calibration values of the Beidou receiver antennae when they are not provided. Further, we design three sets of receiver PCC models with PCC assignment 0, GPS L1/L2 frequency correction value substitution, and IGb R3 calibration value substitution to analyze their effects on the Beidou precise orbit determination. The results show that the receiver phas center offset（PCO） has a greater influence on the precise orbit determination, and the IGb R3 model has the best result for the Beidou-3 precise orbit, with an average orbit fitting accuracy of 3.4 cm, followed by the L1/L2 substitution value of GPS. It is recommended that using the IGb R3 value for Beidou-3 precise orbit determination before the precise PCC model of the Beidou receiver is published in the latest framework.

Based on GAMIT10.71, we study the influence of high-order ionosphere (HOI) delay on the estimation of tropospheric parameters of B2I, B2a and B3I frequency signals of Beidou navigation satellite system(BDS). The experimental results show that during low levels of solar activity, the maximum influence of HOI delay on the zenith total delay(ZTD), north south gradient(NS_grad) and east west gradient(EW_grad) of B2I is 0.80 mm, 0.60 mm and 0.99 mm respectively. The maximum influence of ZTD, NS_grad and EW_grad of B2a is 3.60 mm, 10.77 mm and 10.74 mm respectively. The maximum impact of ZTD, NS_grad and EW_grad of B3I is 1.60 mm, 3.28 mm and 5.90 mm respectively. During the period of high solar activity, HOI has a greater impact on the estimation of tropospheric parameters. The experimental results further show that the influence of HOI on tropospheric parameter estimation is higher in the daytime than at night, and the influence of HOI in low latitude is higher than that in high latitude.

In order to improve the accuracy of sea level change prediction, we propose combining CEEMDAN with LSTM to get the CEEMDAN-LSTM method adopting the idea of decomposition-prediction-recombination. The results show that the combined CEEMDAN-LSTM method outperformed the direct LSTM prediction(MAE=16.87 mm,RMSE=21.51 mm) and the existing combined EEMD-BP neural network method(MAE=10.4 mm,RMSE=15.44 mm) with lowest prediction error (MAE=8.89 mm,RMSE=11.34 mm).

Geocentric motion sequence contains complex noise and the real signal is difficult to extract effectively, so we propose a noise reduction method LOESS-EMD combining the locally weighted regression(LOESS) and the empirical mode decomposition(EMD) methods. Firstly, we fit the geocentric motion sequence with the LOESS method to obtain the fitted time series and the residual sequence. Then, we perform the EMD method on the residual series and extract the low frequency signal from it. Finally, we reconstruct the fitted time series and the low frequency signal in the residuals to obtain the denoised time series. In the simulation data experiment, compared with the LOESS method, the LOESS-EMD method reduces the root mean square error by 31% and improves the signal-to-noise ratio and percentage of remaining energy by 16% and 0.16 percent point. This method is used to denoise the geocentric motion sequence provided by the International GNSS Service(IGS) 3rd reprocessing campaign(Repro3), and the experimental results show that the LOESS-EMD method can effectively reduce the noise of the geocentric motion sequence and thus improve its accuracy.

Taking the nearly three slow slip events(SSE) in Boso Peninsula in Japan in 2011, 2013-2014, and 2018 as examples, we use the independently developed GPS coordinate time series analysis(GTSA) software to batch process the coordinate time series data of multiple GPS stations in the peninsula for a total of 10 a from 2009 to 2019, and build a coordinate time series model containing periodic SSE. The results show that: 1) The mathematical characteristics of the hyperbolic tangent function are very consistent with the motion law of SSE. At the initial stage, the hyperbolic tangent function accelerates slowly, then slides rapidly, and finally decelerates slowly to return to the steady state; 2) The precise modeling of SSE by the function can effectively determine the center of SSE occurrence time, duration, and surface displacement; 3) The duration of SSE in 2011 can reach 50 d at most, and in 2013-2014 and 2018 can reach 25 d at most; 4) The horizontal displacement of the three SSE is toward the southeast, and the elevation displacement is shown as subsidence. The maximum horizontal displacement of the SSE in 2018 was 5.2 cm, and the maximum elevation displacement in 2013-2014 was 3.8 cm.

Using GNSS horizontal velocity field data from 2009-2011, 2011-2013, 2013-2015 and 2015-2018, based on block and negative dislocation model, this paper inverts the dynamic change process of locking degree and slip deficit of the Anninghe-Zemuhe fault zone in corresponding four periods. We study the dynamic characteristics of the Anninghe-Zemuhe fault zone based on the GNSS velocity field across fault profiles. The results show that the locking distribution of the Anninghe-Zemuhe fault zone changes periodically, but the Shimian-Mianning section is always in a highly locked state. From 2013 to 2018, the movement rate of Sichuan-Yunnan block first decreased and then increased compared with the South China block, which corresponds to the locking and then unlocking process of the Zemuhe fault zone. According to the cross-fault GNSS velocity profile, the motion characteristics of the Anninghe-Zemuhe fault zone show sectional changes.

Using the Snoke approach, we solve the focal mechanism of the July 2, 2018 Fuliang M_L4.1 earthquake in Jiangxi province. The results show the best focal mechanism is: nodal plane I, strike is 140°, dip angle is 87°, slip angle is 4°; nodal plane Ⅱ, strike is 50°, dip angle is 86°, slip angle is 177°. The P axis azimuth is 275°, and it is strike-slip fracture mode. The results of nodal plane Ⅱ are in good agreement with those calculated by CAP method, and the P axis azimuth is consistent with the regional stress field. Outcrop geological profile features show that the NE-trending Yifeng-Jingdezhen fault is an early-middle Pleistocene fault with moderate and strong earthquake generating ability. The attitude of the fault is consistent with the strike and dip angle of nodal plane Ⅱ. We conclude that the Yifeng-Jingdezhen fault is the seismogenic structure of the earthquake.

Based on the data of Ningxia regional seismic network and earthquake warning network, we calculate the focal mechanism of the July 12, 2022 Yanchi M_L4.0 earthquake by the Hash method. We obtain the focal mechanism, moment tensor solution and centroid depth of the earthquake by the gCAP method. We determine the focal mechanism central solution, and calculate the relative shear stress and normal stress on the two nodal planes of the focal mechanism solution. Finally, we carry out the polarization of fast shear wave and time delay of slow shear wave on the station that conforms to the shear wave splitting window. The results show that the Yanchi earthquake is a natural strike-slip earthquake event. The focal mechanism central solution of different methods is strike 99°, dip 71°, slide 8° for nodal plane I and strike 6°, dip 83°, slide 161° for nodal plane II ,and the P-axis azimuth is 54°. The relative shear stress generated by the stress tensor on nodal plane I and nodal plane II of the focal mechanism central solution is 0.982 and 0.942, respectively. The near-station shear wave splitting shows that the fast shear wave polarization direction is 55° and the slow shear wave time delay is 1.7 ms/km. We find that the P-axis azimuth of the focal mechanism and the polarization direction of the fast shear wave of the Yanchi M_L4.0 earthquake are mainly controlled by the regional tectonic stress field, and the slides of the two nodal planes from the focal mechanism are basically consistent with the slides of the shear stress. Combined with existing research, we consider that the seismogenic fault may be a parallel associated structure near the EW-strike basement fault in the Ordos block. Under the principal compressive stress of regional tectonic stress field in the NE-SW direction, after a long period of stress accumulation, the rupture occurred on the optimal release nodal plane of shear stress, triggering the M_L4.0 Yanchi earthquake.

Through uniaxial compression test, we obtain the elastic modulus and Poisson’s ratio of the drill core of the borehole strainmeters in Jiangsu province. We calculate the theoretical coupling coefficient and the measured coupling coefficient of the borehole strainmeters using surface strain theory, and discuss their correlation. On this basis, we introduce the observation accuracy, study the relationship between the observation accuracy of the instrument, the coupling coefficient and the data quality, and determine the threshold of the difference of the coupling coefficient. The results show that: 1) The observation accuracy of the borehole strainmeter is positively correlated with the theoretical coupling coefficient and the measured coupling coefficient. The better the consistency between the theoretical coupling coefficient and the measured coupling coefficient, the higher the observation accuracy of the borehole strainmeters and the better the data quality. 2) The threshold of difference between the theoretical coupling coefficient and the measured coupling coefficient of the surface strain at each station in Jiangsu province is 0.1. When the difference is less than 0.1, the observation data quality is higher. When the difference is between 0.1 and 0.2, the downhole coupling has a certain, but not significant, effect on the observation accuracy. When the difference exceeds 0.2, downhole coupling problems occur, resulting in low instrument observation accuracy and poor data observation quality.

On the basis of the Bouguer gravity anomaly data of EIGEN-6C4 model, V19.1 terrain data and CRUST1.0 model crustal data, we calculate the effective elastic thickness (T_e) of lithosphere in north China by the joint inversion method of admittance and correlation function methods. The results show that the T_e value in north China has obvious lateral inhomogeneity. The Ordos block is a high value area, and the T_e is about 30-60 km. The north China plain is a low value area, and the T_e is about 5-25 km. The Ludong-Huanghai block is a middle and high value area, and the T_e is about 20-50 km. The high T_e value area of the north China Craton corresponds to the stable Ordos block, and the middle and low T_e values correspond to the relatively active Taihang mountain tectonic belt and Tanlu fault zone. From the relationship between T_e value and surface heat flow, the area with high T_e value generally corresponds to the area with low heat flow value, while the area with low T_e value generally has high heat flow value, which is consistent with the general distribution law of lithospheric thickness and surface heat flow. According to T_e and the spatial distribution of earthquakes, the seismic activity in north China is mainly concentrated in the middle and low value areas of T_e, indicating that the area with smaller T_e has poor stability and may be more conducive to the preparation and occurrence of earthquakes.

Using the repeated measurement data from Hainan island mobile gravity survey network of 13 periods from 2016 to 2022, we analyze the dynamic changes of differential gravity field and cumulative gravity field in the study area, and emphatically analyze the spatial and temporal changes of gravity field in the southwest of Hainan island before and after the Sanya M_S4.2 earthquake on August 20, 2019. The results show that before 2 and 3 years of Sanya M_S4.2 earthquake, in the area near the epicenter of the Sanya M_S4.2 earthquake a gradient zone of cumulative change of gravity field appears, the anomaly range of gravity variation gradient zone before the earthquake is 60 km, and the maximum variation of gravity field is about 60 μGal. After the Sanya M_S4.2 earthquake, the gravity field in the area near the epicenter continued to rise and began to decline in coordination until the second half of 2020. During the decline, a swarm event occurred with the Ledong 3.2 earthquake on August 8, 2021 as the largest earthquake. The wavelet multi-scale decomposition of the dynamic changes of the cumulative gravity field in Hainan island from September 2016 to August 2019 shows that the gravity field changes in southwest Hainan island are abnormal in the third, fourth and fifth order wavelet detail maps.

We use the convolution filtering method to eliminate rainfall correction processing for the resistivity of Shandan ground, and then carry out seismic response efficiency statistics on the original data and the data after rainfall removal respectively through the combination of different processing methods before the normalized monthly rate analysis, so as to select the optimal data processing method before the normalized rate analysis. The results show that the NS channel is suitable for decanting and deperiodic processing. Both EW and NW channels are suitable for deperiodic processing, and the seismic performance of each channel after the removal of rainfall is better than that before, indicating that rainfall has a great interference effect on the extraction of resistivity anomalies of Shandan station. Since the Qilian mountain seismic zone showed an unprecedented level of activity in 2022 and gradually recovered to a calm level in August, we select the rate overlimit anomaly of Shandan resistivity before moderate strong seismic activity in the Qilian mountain seismic zone and its vicinity from 2014 to 2022 to summarize the seismic response index. We predict the seismic risk of the gradually quiet Qilian mountain seismic belt by using the rate overlimit anomaly of the Shandan resistivity from August to December, and determine whether the seismic activity level near the Shandan observation station is really reduced. The results show that the rate overlimit anomaly still exists in the Shandan resistivity, which indicates that the seismic risk still exists in the Qilian mountain seismic zone and the seismic activity level has not weakened.

Based on the generalized three-cornered hat method, we evaluate the uncertainties of global basin terrestrial water storage (TWS) changes calculated by five latest released GRACE/GRACE-FO time-varying gravity field models. Then, we investigate the effects of geographic location, climate type, and basin area on the uncertainties of basin TWS. The results show that: 1) The average uncertainties of basin-scale TWS changes for COST-G, CSR, JPL, ITSG, and GFZ models are 0.41 cm, 0.63 cm, 0.66 cm, 0.81 cm, and 0.97 cm, respectively. 2) Uncertainty in basin TWS changes is strongly correlated with geographic location and basin area, while climate type has little impact on the uncertainty. 3) There are large differences in the basin TWS changes from different solutions when observed data is poor.

For the traditional calculation method for Gaussian filtering coefficient, when the filtering radius is large, the reference spherical radius is small, and the spherical harmonic order is high, there is often computational instability. In order to solve this problem, the extended exponential method is applied to the backward recursive calculation of the spherical harmonic coefficient of Gaussian filtering. The numerical calculation and practical applications show that, the method proposed in this paper realizes the high-precision computation of ultra-high-degree Gaussian filtering with arbitrary radius on any radius sphere.

We analyze the spatial and temporal variation characteristics of the scale factor coefficients of 12 CG gravimeters. The results show that the scale factor coefficients calibrated in different time periods within the same reading range vary, and the scale factor coefficients calibrated in different reading ranges within the same time period also vary. The application example proves that the accuracy of scale factor coefficient calibration of the CG gravimeter using the data between gravity datum points in the survey area is better than that of it calibrated at the national baseline field.

In accordance with the relevant theories and specifications of the Standard for Earthquake Gravity Survey,  we discuss several problems urgently need to be solved in practical work, and existing in the currently used data processing software of the gravity survey electronic recording system, such as instrument height correction, section difference calculation, ring closure difference calculation, etc.

From the perspective of observation environment, through the regression analysis of meteorological factors, we find that although there is a linear effect between temperature change, pressure change and strain change, the fitting results are quite different from the actual observed strain change when the strain pressure drops. There is a strong correlation between the start time of the decline and the time when the Meishan reservoir starts to discharge and the time when the water level of the downstream Shihe river rises sharply. The numerical simulation results of the three-dimensional irregular load model show that the early rapid decline of the anomalous strain compressibility of the cave body has little correlation with the load change of the Meishan reservoir in the same period, and has a large correlation with the water level change of the Shihe river. The tectonic stress field in the study area did not change significantly before and after the occurrence of cave-strain and compressive anomaly. Therefore, we preliminarily infer that the abnormal drop of strain and pressure in Jinzhai station cave may be caused by the environmental load changes in the surrounding sites.

In order to compare different time periods and different techniques for estimating earth rotation parameters, we design experiments to estimate ERP using CONT08, CONT11, CONT14, and CONT17 campaigns and IVS-R1/R4 regular observations in 2008, 2011, 2014 and 2017. At the same time, we use GPS data in the same period with CONT observations to participate in the solution. We use the EOP14 C04 sequence provided by IERS as a reference value for accuracy assessment. The results show that CONT17 has a higher polar motion solution accuracy than the other three sets of CONT observations, with an accuracy improvement of about 40%, but the accuracy of UT1-UTC is basically at the same level. CONT observations have more outstanding advantages than other types of VLBI observations, for example, the improvement in polar motion accuracy is about 30% and improvement in UT1-UTC accuracy is about 45%. As for the comparison between VLBI and GPS observations, the former is more advantageous in solving UT1-UTC, while the latter is more advantageous in estimating polar motion.