The long-term gravity changes of the Chile region are obtained by GRACE gravity satellite data. The gravity changes before and after the 2010 MW8.8 Chile earthquake are further analyzed through the time-series of gravity changes of typical points. Finally, the coseismic gravity changes detected by GRACE are compared with the results calculated by the spherical dislocation model. The results show that the strong earthquake caused significant positive and negative gravity changes between the upper wall and foot wall of the seismogenic fault. The time-series of the gravity changes of typical points all clearly reveal the jumping phenomenon caused by the strong earthquake, as well as the abnormal signals caused by the three strong aftershocks. Especially, the gravity changes of a typical point in the epicenter area presents characteristics of "decrease-increase-decrease-increase-earthquake occurrence" from 2007 to 2010 before the strong earthquake. The magnitude and distribution of coseismic gravity changes observed by GRACE are in good agreement with the results calculated by the spherical dislocation model. In the different stages of the earthquake, the gravity changes between the two sides of seismogenic faults are also different. In addition, because of the effect of stratum viscosity after the strong earthquake, the gravity changes around the seismogenic faults will continue to show the trend of positive and negative changes.

In order to improve the calculation efficiency of adjustment models with uncertainty based on F-norm, a directly iterative algorithm is developed. The algorithm does not use singular value decomposition (SVD), is simple in the concept, and is easy to program. Another algorithm of SVD-equations is also given when the iterative algorithm is divergent. The results of the binary linear fitting and AR model in settlement observation illustrate that the two proposed algorithms could be practiced and are equivalent to the algorithm of SVD-iteration.The directly iterative algorithm is more suitable when the iterative algorithm is convergence, which has faster convergence rate and higher calculation efficiency. Moreover, the algorithm of SVD-equations can be used when the iterative algorithm is divergent.

In order to analyze the efficiency of multi-GNSS combined PPP technology in different complex conditions in the Asia-Pacific region, GNSS observations from five stations are used to simulate three kinds of shelter conditions: surrounded shelter(urban environment),  unilateral shelter(canyon environment) and headspace shelter(large bridge subgrade environment). The service advantages and positioning efficiency of multi-system PPP technology are analyzed from the five aspects of available satellite number, PDOP value, available rates, positioning accuracy and convergence time. The numerical results show that, compared to single GPS PPP results, the four-system GNSS PPP results can be improved 300%, 40%, 2%, 20% and 50% in terms of available satellite number, PDOP value, available rate, positioning accuracy and convergence time in the normal environment, and 300%, 57%, 25%, 39% and 52% in the complex environment.

It is difficult to get the optimal result using empirical weight ratio in combination positioning, due to differences in observation noise and orbit accuracy between GNSS systems. Based on this, Helmert variance component estimation is introduced into GPS/GLONASS/BDS combined positioning, determining the reasonable weight ratio of the GPS/GLONASS/BDS combined single point positioning and baseline solution. Based on a set of experimental data, the experimental results show that the method can reasonably determine the weight ratio of the observations of the GPS/GLONASS/BDS combined positioning. The optimal weight ratio of pseudo-range observations is 5∶1∶1, the optimal weight ratio of phase observations is 1∶1∶1, which improves the accuracy and reliability of the GPS/GLONASS/BDS combined positioning.

In this paper, a single observation Kalman filter is proposed to speed up the gain matrix computation. The OpenMP technique is introduced to accelerate the covariance matrix update. Lastly, 55 reference stations are selected to estimate real time GPS satellite clocks from day 79 to day 89 in 2017. The estimated clocks are compared with IGS final 30 s clocks. PPP in static and kinematic solutions for 10 stations are carried out. The results show that the differences are within 0.5 ns, the precision in horizontal can be better than 1 cm in static and 1~2 cm in kinematic. The precision in vertical can reach up to 1.5~3 cm in static and range from 3~4 cm in kinematic. Furthermore, the time cost for each epoch is reduced to 4 s from 6 s when 8 slave threads are involved.

Aiming at the low accuracy of the traditional tropospheric delay model, a high precision fusion model for the northern hemisphere is established based on the Hopfield model, using the neural network model error compensation technique. Taking the zenith tropospheric delay (ZTD) as the approximate “true value” of the meteorological sounding data of more than 120 observing stations in 2010 provided by the University of Wyoming, this paper analyzes and compares the Hopfield model, the traditional BP model, and the computational accuracy of the fusion model. The results show that the root mean square error (RMSE) of the Hopfield model is 35.31 mm, the RMSE of the traditional BP model is 30.34 mm, and the RMSE of the fusion model is 23.31 mm.

By testing the length of time series, the influence of geomagnetic index on the exclusion of geomagnetic activity and geomagnetic activity is analyzed. For the periodic variation of the TEC value of the ionosphere and its inhomogeneous distribution with latitude, we use the seasonal time series model, SARIMA and the Holt-Winters exponential smoothing model. 48 regions of different latitudes of the northern hemisphere are predicted using the TEC grid data provided by the IGS Center, and the daily mean relative accuracy and root mean square error are defined to assess forecast accuracy. The results show that all three models can reflect the cyclical changes of ionospheric TEC values, but the root mean square error decreases with the latitude and shows an overall trend of growth; at 25° and 55° north latitude it shows a maximum value, while at 45° N, it shows a minimum value.

Using GAMIT software and observation data from 58 CORS stations in Sichuan province from 2012 to 2014, precipitable water vapor (PWV) is deduced. The spatial and temporal distributive characteristics and variation trends of PWV in Sichuan province in the past three years are analyzed. The results show that the spatial distribution of PWV is higher in the east than in the west. Using multiple linear regression and principal component analysis, the relation of PWV and terrain can be linearized. Taking the CHDU, YBIN and HONY stations in Sichuan province as examples, the method of empirical mode decomposition combined with spectrum analysis is used to reflect the various periodic characteristics. The results show that the three stations have several PWV cycles.

Based on the shipborne shoreline integration measurement system, the integrated measurement data storage format, multi-thread network communication, time synchronization control, and time matching technology, are researched and designed to realize the real-time storage function of shipborne multi-sensor integrated measurement data. Data is stored in a unified file. The experiment proves that the proposed storage method can effectively improve system integration, overcome the redundancy of multi-sensor storage efficiency, and improve the storage performance of measurement data.

This paper investigates effects of different tide models on ocean tide loading displacements in coastal areas of China. We use GNSS continuous observation data from 23 coastal CMONOC stations and four ocean tide models--FES2004, EOT11a, TPXO7.2 and Chinasea2010. Firstly, tide wave parameters of the four ocean tide models are calculated, showing that discrepancies of tide wave parameters among different tide models can reach the centimeter-level in the up direction. Then we apply ocean tide loading corrections to the coordinate time series of 23 stations using tide wave parameters of these ocean tide models respectively. Results indicate that ocean tide loading correction has a great impact on wRMS of the coordinate time series. For most stations, wRMS is reduced by 10%~70% after correction. Compared with stations in other investigated coastal areas, the reduction of wRMS in the east China sea coastal area, which can be more than 50%, is most significant. Among four ocean tide models, the correction based on FES2004 is more remarkable, with wRMS reduction reaching 67.5%.

A modified EEMD method is applied to the GRACE derived terrestrial water storage time series of the four basins. Through this means, various periods of hydrological cycle are captured for these basins, indicating that GRACE derived TWS changes contains lots of cycles. Consequently, the Yangtze, Yellow, Huaihe and Pearl river basins have periods of 3.05, 2.37, 2.37, 2.13 years, respectively.

According to history and instruments, south Longmenshan fault, which is considered a seismogenic structure, has experienced three strong shocks: the 1327 Tianquan, the 1970 Dayi, and the 2013 Lushan earthquakes. These earthquakes triggered rockfall and landslides. We survey the typical landslide and the historical earthquake landslides along the Dachuan-Shuangshi fault across the earthquake area and find the distribution range and transport distances triggered by historical earthquakes greater than the Lushan earthquake. The density, volume and particle size triggered by the historical earthquake reduces from southwest to northeast along the fault. Results from the rock fall of the historical earthquakes and the Lushan earthquake, based on relative chronological tests of buried object and epigenetic lichen, and the formula of sensible radius and epicenter, show that the ground vibration is more heavily triggered by the Tianquan earthquake in 1327. Vibrations from the 1327 earthquake may have triggered the Tianquan county “Dayanbeng” and Shuangshi town “Shaijingshu”. The historical earthquake epicenter may be located in the southwest section of the Dachuan-Shuangshi fault, near Tianquan county Xiangshuixi town. The intensity from the epicenter is greater than 9, and the magnitude of the paleo-earthquake may be great than or equal to the Lushan earthquake.

Based on the deformation-strain calculation of GPS continuous observation station in Tianjin area, the results show that the deformation field and strain rate field are complex, with east-west zoning as well as south-north zoning. It is considered that the Tianjin area is located in different tectonic intersection area, which is affected by the Tangshan-Cixian seismic-tectonic zone, the Zhangjiakou-Bohai seismic-tectonic zone, and other active faults.

Using precise leveling data of the north of Yilan-Yitong fault zone in 1977, 1991 and 2015, we take the pseudo inverse linear dynamic movement model in the case of reference to carry on the adjustment, and reduce to the selected benchmark. We then obtain the two relative velocity fields in the region (1977-1991 and 1991-2015). The previous velocity field shows that roughly Yilang-Yitong fault zone is bounded, the north of Changbai mountain is a relatively uplifting movement compared to the southern section of Lesser Khingan Mountains, with a maximum relative rate of 7.51 mm/a. The second phase of the rate field indicates a relatively low speed of the northern region of the Yilan-Yitong fault zone, with a maximum relative rate of only 3.4 mm/a. The area tends towards a holistic approach to inheritance; the relative motion of the fault line decreases significantly.

Based on waveform data recorded by 19 broadband digital stations of Gansu, Qinghai and Ningxia regional seismic network, we use the CAP method to calculate the focal mechanism solution and the optimal focal depth of Menyuan MS6.4 earthquake in January 21, 2016. The results show that the double-couple solution is: nodalⅠ: strike 143°, dip angle 40°, slip angle 71°; nodalⅡ: strike 347.2°, dip angle 52.6°, slip angle 105.3°. The moment magnitude is MW5.9 and the optimal depth is 7.7 km, which is consistent with the results of others (such as CENC, IGP-CEA and Harvard). Using the arrival time difference between sPn phase extracted by sliding window correlation method, we obtain the focal depth of 8.5 km, which is coincident with the result of CAP inversion, and we validate the feasibility of the method.

Using the Moya method on the waveform data recorded by the Gansu digital seismic network, we calculate the stress drop of 157 earthquakes with ML≥2.0, occurring in Menyuan area before the January 21, 2016 Menyuan MS6.4 earthquake. The results show that the stress drop ranges from 0.02 MPa to 10.86 MPa with an average value of 1.67 MPa. The earthquakes with high stress drop are mainly distributed between Huangcheng-Shuangta and Lenglongling fault zones, and the area is in harmony with the active area of moderate-strong earthquakes. It is found from analyzing the temporal distribution of stress drop that obvious stress high drop values appear before the MS5.1 earthquake. The stress drop of earthquakes gradually decreases and are generally at a low level after the MS5.1 earthquake. About half a year before the MS6.4 earthquake, twice high-value anomalies of the monthly mean values of stress drop appear, showing some relationship with Menyuan earthquake.

In this paper, we select the focal mechanism solution data of 36 shallow-focus earthquakes that satisfies M≥7.0 from January 1, 1976 to December 24, 2016 on the south America plateau, drawing from the global CMT catalog. We use Coulomb3.3 to calculate the static coulomb stress changes in the subsequent earthquake fault surface, and discuss whether there is a stress-triggered effect from previous earthquakes to subsequent earthquakes. The results show that, as the trigger threshold is 0.1 bar, 11 earthquakes are triggered by the previous earthquakes; the triggered earthquakes are 31.4 % of the total; there are 7 suppressed earthquakes and 17 uncertain earthquakes; 22 of the earthquakes have a positive static coulomb failure stress change, which accounts for 62.9% overall. Based on the research results above, we draw the conclusion that there have been some stress triggering effects between the south America plateau strong earthquakes since 1976.

Using the P-wave travel time of seismic data processed by the double difference tomography method, we obtain hypocenter relocation results and P-wave velocity structure within the range of 0~200 km beneath the Japan-Kuril trench subduction zone. This divides the Japan-Kuril trench subduction zone into the north of the Japan trench subduction zone and the south of the Kuril trench subduction zone. The Pacific plate is continuous within the depth of 200 km. Combining bathymetric with fracture feature data, we identify the fracture classification of the Japan-Kuril trench subduction zone.

In this paper, we calculate receiver functions of body waves under 30 stations in Hubei province, using data from the teleseismic earthquake (2014~2015), obtaining thickness, poisson ratio, and CCP profile (AA′). The results show that: (1) the crust thickness in western Hubei is thicker than in the east; crustal thickness in northwest and southwest Hubei is at about 45~55 km, in the east it is about 30~35 km; (2) the poisson ratio in Hubei province is about 0.23~0.27, which may be related to the moderate rock compositions; (3) the Moho surface looks like “stairs”. In combination with related literature analysis, eclogite phase changes occur below the Earth’s crust in the Wudang mountain area, so that the alternating density phase changes of lower crust delamination causes abrupt changes in western Hubei regional crustal thickness.

After the 2011 Tohoku M_W 9.0 earthquake, 43 vertical pendulum tiltmeters recorded the Earth’s free oscillation signals that were excited by the great earthquake. We find that the tiltmeters have a special sensitivity to detect low-frequency Earth’s free oscillation (1~5 mHz). Even in very low frequencies (<1.5 mHz), the modes can be identified with high SNR. We select 36 records out of 43 stations in this study. Using signal enhancement methods (product spectrum over 36 records), even the gravest modes are detected. All of the spheroidal oscillation modes (₀S₃~₀S₃₈) and almost all of the toroidal oscillation modes(₀T₃~₀T₂₈) and 21 harmonic modes are observed.

In this paper, based on synthetic seismograms, the effects of receiver depth, S wave velocity and attenuation structure on noise level are investigated. We find that the amplitude ratio of converted acoustic wave and P wave decreases exponentially with depth, when the receiver is deep enough and where S wave velocity is larger than acoustic wave velocity. Combined with the borehole seismic data of the 2015 Tianjin explosions, features of noise level are further discussed and examined, showing consistency with the characteristics of synthetic data.