We focus on the southeastern margin of Qinghai-Tibet plateau, which presents weak distribution characteristics and spatial characteristics of seismicity, and put forward the crust-mantle coupling model of strong earthquake gestation. Using data collected from the China National Digital Seismic Network in Yunnan and Sichuan Regional Digital Seismic Network, and setting in the southeastern margin of Qinghai-Tibet plateau seismic stations teleseismic records by SKS waveform data, through the upper mantle anisotropy results for the shear wave splitting data, we describe the southeastern margin of Qinghai-Tibet plateau in the upper mantle anisotropy images. Then, we carry out correlation processing and analysis respectively with the velocity field data observed by GPS in this area, the data of regional tectonic stress field,and the medium density data at a depth of 13 km. The proposed model is verified to a certain extent. Analysis results show that through the research on the coupling characteristics of the deformation field and stress field, and the anisotropy of fast wave polarization direction with the direction of GPS crustal movement velocity field changes, the maximum stress direction and level of the overall correlation is larger, the mutual effects among various spheres of crust-mantle and asthenosphere shown on the flow of seismic anisotropy are closely related to the stress transfer mechanism. At the same time, we can also get the key role of weak zone(low speed and high conductor) in the study of coupled crust-mantle motion through correlation analysis.

Based on previous studies, the rupture parameters of the 1970 Tonghai MS7.7 earthquake are obtained. Using the model, we study the changes of Coulomb rupture stress under different lithospheric rheological properties. We discuss the effects of static co-seismic deformation and short-term post-earthquake deformation on potential seismic activity in the Xiaojiang fault and the Red River fault. Results show that on some segments of the two faults, the coseismic change of Coulomb failure stress(ΔCFS) attains 0.12-0.50 bar at 7.5 km depth. Relaxation of the lower crust and the upper mantle adds the postseismic ΔCFS on the two faults up to 0.22-0.90 bar after 48 years. This means that the 1970 event is increasing the potential of seismic activity along the two active faults. In this case, we should pay more attention to these two active faults in future research by increasing field observations to monitor the seismic activity and the crustal motion pattern.

In this paper, we utilize Sentinel-1A SAR data and D-InSAR technology to obtain the coseismic deformation, inverse the fault geometry, and coseismic slip distribution of the 22 May, 2016 Dinggyê M_W5.3 earthquake. The results show that the seismogenic structure is a NS strike normal fault with dip angle 43°,surface trend east and rupture length 9 km. The coseismic slip mainly occurs within 2-5 km, with a maximum slip of 0.24 m; the moment magnitude is MW5.3. The coseismic fault of the 2016 Dinggyê earthquake is a new blind normal fault in the Dinggyê-Shenza normal fault system.

In this paper, using 3-D Google maps, we plot the current global distribution map of major tectonic fault zones, providing a new method to observe and interpret the movement and evolution of oceanic crust. Our technique plays a very important role in further research on the distribution and the mechanism of the evolution of active oceanic faults, the expansion of the seafloor, mantle convection, etc.

Drawing on the known seismogenic structures of future large earthquakes, we propose to use empirical formulas to determine the parameters of length, width, seismic moment and sliding distribution of the large earthquake faults, and use the random finite fault method to predict future large earthquakes. Considering the M_W7.8 earthquake that occurred in New Zealand on November 13, 2016, the ground motion time history and response spectrum of 12 bedrock stations, are simulated by stochastic finite fault method. The simulation error is determined by the average ratio of the simulated spectrum amplitude to the recorded spectrum amplitude. The results show that the period is in the range of 0-10 s and the simulation error is between 0.92-1.08. The standard deviation of the simulated error of different frequencies is not more than 1, and the width of 95% confidence interval has not changed significantly with the frequency. The simulation results reflect the average effect of ground motion records. Although a specific simulation may differ greatly from records, for engineering purposes we are interested in whether the simulation results are equivalent to the average results of seismic records. The quasi-random method is used to retrieve the initial rupture point and dislocation slip distribution of the New Zealand earthquake fault, simulate the ground motion, calculate the variation of simulation error with frequency, and obtain conclusions similar to those of the original model. This further confirms our proposed method of obtaining the seismic source parameters and the use of stochastic finite fault methods for the prediction of future large earthquakes is reliable, especially for the far-field simulation of large earthquakes.

The bubble oscillation equations and pressure wavelet equations for large-capacity air-gun source are established, and several correction factors that have important influence on the simulation results are introduced, including open thermodynamic model, muzzle throttling,evaporation and condensation,  gun body’s influence and air release rate to improve the traditional method of single air-gun pressure wavelet calculation. By comparing the simulation and the measured wavelets of reservoir experiment, influences of each correction factor on the wavelet simulation is analyzed, and the pressure wavelet of the large-capacity air-guns array is further calculated by the improved single-airgun signal. Comparing the wavelet waveform predicted by the improved simulation method with the measured signal, It shows that the improved model can simulate the source wavelet of single gun and gun array well. Finally, the reason why the model produces simulation error is analyzed.

Based on high-rate GNSS data which are collected from the New Zealand earthquake on November 13, 2016 and processed using the precise point positioning(PPP) module of PANDA software, this paper analyzes surface deformation and calculated magnitude. Then the impact of GNSS seismic wave measurement accuracy on magnitude estimation is quantitatively assessed. The results show that when the accuracy of peak ground displacement(PGD) is within 2 cm, the error of magnitude estimation is within 0.1 magnitude units. In addition, because the magnitude estimation is related to the epicenter distance and PGD of the station, when the epicenter is far away, the estimated magnitude is more sensitive to PDG.

Based on short cross-fault level data from Gansu province since the 1980 s, using the research method of “point-line-surface”, the spatial and temporal evolution characteristics of strain background, fault activity and short-term deformation anomalies before 8 earthquakes of magnitude 5.5 or above from 1990 to 2016 are comprehensively analyzed. The results show that the following typical abnormal characteristics appear on different levels before multiple earthquakes:1) Earthquakes usually occur in the high value area or the high gradient zones of vertical deformation accumulation rate, and the orderly distribution of deformation field occurs before the earthquake; 2) There is often significant difference in motion between different faults or between different segments of fault zones before earthquakes, and the medium-term motion rates in seismogenic faults are significantly enhanced, resulting in imbalance of regional motion; 3) With the approach of the strong earthquake, significantly abnormal sites in the spatial distribution obviously migrate to, and concentrate at, the near seismic region, and abnormal intensity gradually increases in some cross-fault sites near the seismic source region. The time, space and intensity evolution characteristics of these typical anomalies before the earthquake indicate that fault activity and stress field near the earthquake zone have some representation before the fault movement instability, fault activity before the earthquake has different instability fluctuations, and the stress appears to transfer to the near source area.

In this paper, the Hongqinghe coal mine in Ordos city is taken as the experimental object, the interference processing and time series analysis of Sentinel-1A satellite image are carried out. By analyzing the similarity degree of the statistical attributes of time dimension samples, the pixels with the same scattering characteristics are selected for homogeneous filtering, and the phase is adjusted according to the phase closure principle in the complex field to realize the DS-InSAR processing in this area. The accumulated subsidence area of the coal mine is about 32.3 km2, and the subsidence results are consistent with the spatial distribution of the coal mine. Compared with the results of PS-InSAR, DS-InSAR has better effect in subsidence monitoring.

We design a robust adaptive Kalman filter model to deal with the dynamic characteristics of rapid subsidence areas and the influence of gross errors in observation vectors on Kalman filter results. The model can identify two states of steady settlement and rapid settlement. Robust estimation is used to reduce the influence of gross errors in the observation vectors. In order to reduce the errors of the state model, an adaptive factor is used to adjust the dynamic model to improve the accuracy of the filtering results. This model is applied to the data processing of subsidence monitoring in a mining area for verification. Compared with the results of robust Kalman filter, the conclusion shows that the filtering model is better.

According to the wide range of surface deformation, seventeen descending Sentinel-1A images are used to obtain subsidence information of Fushun from 2015 to 2016, based on the short baseline subset InSAR technique. The surface uneven deformation is analyzed using the information entropy method, and then the profile results of the typical regions in the east-west and south-north directions are extracted. Finally, the reliability of the experimental results is verified with existing. The results show that: 1) From June 2015 to June 2016, the deformation centers are distributed in the Fushun east open-pit mine, Fushun west open-pit mine, Tayu, Nanhuayuan, Xintun, etc. 2) The deformation information entropy of both Yulin and Zhanqian street is greater than 1.0, and the surface deformation is severely uneven, which should be monitored continually in the future. 3) The maximum subsidence is about 78 mm in the north side of the Fushun west open-pit mine; four subsidence funnels formed from west to east in the Fushun east open-pit mine, and the maximum subsidence is around 250 mm; the maximum value of the north subsidence region is 55 mm in Tayu township.

According to different remote sensing platforms, the progress of soil moisture monitoring using GNSS-R technique with ground-based, tower-based, airborne and space borne platforms are presented in this paper. A detailed review of GNSS-R joint with microwave radiometer is also given in this paper. And we discuss the key and difficult points of GNSS-R/IR in soil moisture inversion.

We put forward a new model parameter method to retrieve the atmospheric refractive based on the traditional model parameter technique. In addition, a new method for estimating the tropospheric height is proposed. Eight IGS stations in China and surrounding areas in July 2016 are selected for trial calculation. Using the radiosonde as the reference, statistical results show that the atmospheric refractive derived from new model parameter method is closer to the truth.

This paper studies the stochastic model of DCB for GNSS systems by post-processing phase smoothed code data. To resolve the problem of estimating the uncertainty of the parameters of the weight of the elevation angle and the equal-accuracy, the paper presents a weighted method, i.e., variation coefficient weighting, to effectively influence the stochastic model. The index weights are determined according to the decision matrix of the index value of the evaluation object. Preliminary results show that: 1) The fitting precision of the model is related to observation data and the station location. The more stable the ionospheric variation of the station position and the adequacy of the observational data, the better the model fitting accuracy. 2) The model fitting accuracy is better than by the weight of elevation angle using variation coefficient weighting. 3) In the accuracy and stability of parameter estimation, the variation coefficient weighting method is superior to the method of the weight of elevation angle, and the calculation accuracy of single-day data can be improved by 0.2 ns.

A series of vibration tests with different frequencies are conducted on a shake table to clarify the characteristics of low-frequency measurement. We analyze the results from two aspects: power spectral density of acceleration and displacement derived from integration of acceleration. We find that accurate displacement can be recovered by double-integration of acceleration(with appropriate high-pass filtering) on the condition that the frequency of vibration signal is above 0.05 Hz. In order to recover accurate displacements of vibration signals, especially in low frequency, we combine GNSS and accelerometer. The results show that this shake table system is an appropriate platform for earthquake monitoring studies; the lower limit of accelerometer low-frequency measurement is around 0.03-0.05 Hz, and the result provides reference for the proper choice of cut-off period for data processing of accelerometers as well. Meanwhile, the data fusion of GNSS and accelerometer data can combine the strength of each instrument and make the displacement derived more accurate and reliable.

This paper introduces the principle of centralized real-time autonomous navigation(AutoNav) based on the parameter decomposition filter, and also presents the simulation method for inter-satellite cross-link observations. Based on the simulated observations for the simulated constellation of BDS, 60 days’ AutoNav is performed. The influence of observation noise and residual systematic errors to the precision of AutoNav is analyzed. The results indicate that observation noise has no distinct effect on the precision of AutoNav. However, the residual systematic error has a great influence and needs to be carefully corrected. Otherwise, there are no differences regarding the influence of the observation noise and residual systematic error on GEO, IGSO and MEO satellites.

The accuracy of solutions for ill-posed equations can be significantly improved by using the reasonable equality constraint between the adjustment parameters. In this paper, an ill-posed total least squares model with equality constraints is proposed by applying equality constraints to an ill-conditioned total least squares model. The regularization criterion for constrained ill-posed total least squares is established, and the iterative solution of parameters and its variance-covariance matrix are obtained according to the Lagrangian multiplier method. Finally, numerical examples and ill-posed trilateration network example are used to verify the correctness of the formula. The results show that new method not only improves the ill-posedness of normal matrix by regularization criteria, but also complies with error of coefficient matrix according to EIV criteria. At the same time, it also considers reasonable prior information between parameters; therefore, the accuracy of solution is significantly improved.

Using CryoSat-2 satellite altimetry data, the sea ice thickness in the Beaufort Sea is retrieved. The accuracy of the results is evaluated using the ULS data in October 2010 to 2013 and IceBridge data in 2011. The results show that the maximum and standard deviations of the difference between the sea ice draught depth and ULS draught depth are 14 cm and 4 cm, respectively. The average and standard deviations of the difference between the sea ice thickness and IceBridge sea ice thickness are 2.7 cm and 65.7 cm, respectively, which are better than Laxon’s (2013) research results. The results are optimized by 2.1 cm and 6.6 cm, respectively. On this basis, the variation of sea ice thickness in summer and winter of the Beaufort Sea from 2011 to 2017 is studied. The sea ice thickness in summer and winter has similar distribution characteristics, and the coverage in March of winter is wider and thicker. The interannual variation of sea ice thickness in March of winter of 2011 to 2017 is further analyzed. The results show that the sea ice thickness in 2011 to 2017 shows an overall downward trend, with the smallest in 2012 and the largest in 2014.