We evaluate the performance of Tm derived from ERA5 and MERRA-2 reanalysis data over China using 89 radiosonde stations data in 2017, and analyze the temporal and spatial variation characteristics of bias and RMSE of Tm calculated from ERA5 and MERRA-2 reanalysis data. The results show that: 1) The Tm calculated by ERA5 and MERRA-2 reanalysis data have the annual bias of 0.41 K and 0.10 K, and annual RMSE of 1.26 K and 1.34 K when compared with Tm derived from radiosonde stations, respectively. 2) The bias and RMSE of Tm calculated from ERA5 and MERRA-2 reanalysis data show similar temporal and spatial variation characteristics. For the characteristics of temporal variation, the bias and RMSE of Tm calculated from two reanalysis data generally show high precision in summer but low precision in winter. Moreover, Tm calculated from ERA5 reanalysis data shows positive bias in the whole year, while the MERRA-2 reanalysis data shows negative bias in summer and positive bias in other time. For the characteristics of spatial variation, there is no obvious variation characteristics in altitude. However, the RMSE shows an increasing trend with the increase of latitude, and generally remain in 2.5 K.

We propose a new water distribution detection method based on spaceborne GNSS-R using reflected GNSS signals collected by CYGNSS (cyclone global navigation satellite system). First, we process the reflected GNSS signals collected by CYGNSS constellation to obtain the surface reflectance data. Then, the study area is divided into grids according to the spatial resolution of 0.01°×0.01°. After eliminating outliers and invalid values, the reflectivity of each grid is recalculated using the region average algorithm. We use the grid sample variance and standard deviation as measurement indexes. The random walker segmentation algorithm is used to detect water distribution. Finally, the distribution of water in Congo basin is analyzed using the proposed method, and the results of this paper are evaluated with MODIS data. The results show the great potential of spaceborne GNSS-R for detecting water distribution, further expanding the applications of GNSS-R.

According to nonlinear and nonstationary characteristics of ionospheric TEC, we establish a residual correction ionospheric prediction model based on the combination of Prophet and Elman neural network. The model is used to model and forecast the ionospheric TEC time series provided by IGS during different degrees of solar activity. The results show that the modified model can reflect the variation characteristics of ionospheric TEC. The average relative accuracy of the model in low solar activity years and high solar activity years is 92.9% and 92.2%, and the root mean square error is 0.94 TECu and 1.77 TECu, respectively. The accuracy of the modified model is significantly higher than that of the Prophet-Elman model and the single Elman model.

To evaluate the performance of BDS global positioning service and zenith tropospheric delay estimation, based on the observations of 9 MGEX tracking stations evenly distributed around the world in a one-week period, we use GAMP software to perform BDS coordinate-static PPP with the self-calculated precise orbit and clock products of GPS/BDS. Experimental results show that the horizontal accuracy for BDS coordinate-static PPP mode is better than 1 cm, the vertical accuracy is around 1 cm, and the positioning accuracy is comparable with GPS. The estimated accuracy of zenith tropospheric delay is better than 1 cm, and the difference of RMS of errors is within 1 mm compared with the result calculated by GPS PPP. In general, the BDS full constellation already has global positioning service capabilities and ZTD inversion performance equivalent to GPS.

The influence of GEO satellites in BDS high precision relative positioning is studied and analyzed. First, the BDS un-differential observation model is realized, the high precision relative positioning solution model of BDS is derived on this basis. Then, the whole constellation satellites and IGSO+MEO satellites combination are used to carry out in high precision relative positioning processing. Finally, the influence of GEO satellites on BDS relative positioning convergence time, positioning accuracy and PDOP values are further analyzed. The results show that the BDS GEO satellites are beneficial to increase the number of observation satellites, and significantly improve the geometry of the observation satellites and the positioning convergence speed and accuracy. Both the static and dynamic results of the whole constellation satellites of high precision relative positioning are at or better than centimeter level.

Based on the analysis of the orbit characteristics of more than ten low-orbit satellites, such as GRACE and SWARM, we analyze the sampling interval and interpolation methods of the precise ephemeris for the low-orbit satellites. We design two precise ephemeris schemes: efficiency first and precision first, according to the user’s requirement. The experimental results show that the orbit height of low-orbit satellites has a significant influence on the sampling interval and interpolation methods. In view of the accuracy requirements of 3 cm and 10 cm orbits, the storage space occupied by ephemeris files can be compressed by more than 99% through reasonable orbit design.

Aiming at the problem of low accuracy and divergence of positioning algorithm for ultra-wideband (UWB) sensors in time-different-of-arrival (TDOA) positioning mode, we propose a simple iterative least squares algorithm. This method has the advantages of a simple formula, convergent coordinates and high positioning accuracy. The main idea is to transform the TDOA localization equation into a nonstandard least squares form by the Pythagorean  theorem. Then, the equation constraint relation between unknown variables is combined to transform the form with two unknown variables into a form with only one unknown variable. Finally, the convergence coordinate of UWB tag is calculated by iteration. Monte Carlo simulation results show that the proposed algorithm is superior to the iterative constrained weighted least squares algorithm in positioning accuracy in high-noise environment.

We use sandbox experiment for structural physical modeling, and combine with seismic survey results and cross-fault observation data, to analyze whether the buried fault activity monitored by Ninghe cross-fault site is effective. The results reveal that when the observation accuracy is higher than the error, cross-fault site monitoring can effectively monitor the activity of buried faults. The distribution and development of the simulated fault profile are consistent with the shallow seismic profile, and surface deformation can be caused by buried fault activity. Due to the different fault tendency of each secondary fault, cross-fault monitoring data may give rise to the illusion of reverse fault. The model diagram of horizontal point position distribution of Ninghe cross-fault monitoring site is presented.

Based on the viscoelastic finite element model, combined with the active fault detection in Handan area and historical earthquake data and distribution of the lithosphere medium in the studied area, we construct a three dimensional viscoelastic finite element fault model. We substitute the kinematics characteristic boundary conditions into the model, and simulate the changes of stress and strain field caused by the earthquake on the active faults. Then, we caculate the Coulomb failure stress changes of the rupture surface caused by the 1830 Cixian earthquake and the sliding direction. At last, the risk of strong earthquake in the target area is analyzed.According to the simulation model based on actual geology and current kinematic conditions, the Coulomb stress near the epicenter of the original strong earthquake of the Cixian fault is reduced which indicates that the activity of the Cixian fault reduces seismic risk near the epicenter of the original strong earthquake. But Coulomb stress is strengthened near the junction of the middle and west sections of the Cixian fault and the two northeast trending faults (Zishanxi fault and Taihang mountain piedmont fault), and the seismic risk in the future should be monitored.

In order to discuss the relationship between temporal and spatial distribution characteristics of fluid solid tidal parameters and earthquakes, we take the 2020 Tangshan Guye MS5.1 earthquake as an example. We collect and analyze the water level data of 16 wells around Guye from January 2017 to July 2020. The tidal parameters of O1 wave and M2 wave are analyzed by Baytap-G program. This paper mainly studies the trend and spatial distribution of M2 wave tidal factors. The results show that:1) The tidal factors of wells around the epicenter before the Guye earthquake in Tangshan show an upward trend, indicating stress accumulation. 2) Combined with the focal mechanism type of this earthquake, we conclude that the variation pattern of tidal factors in 16 wells around the epicenter is mainly affected by the direction of the main stress of the earthquake, and the tidal factor decreases with the compression in the east-west direction; similarly, the tidal factor increases with the extension in the north-south direction.

Based on the finite element method, we use the postseismic GPS observation data from 5 years of the 2015 MW7.8 Nepal earthquake to constrain the rheological structure of the southern Tibet. We use two different boundary structural models to probe the boundary structure of the Indian plate and the Tibetan plateau, which are the vertical boundary structure constrained by the distance between the main front thrust(MFT) and the Indian elastic subduction slab structure, respectively. The results show that both structures can produce south-southwest horizontal motion pattern consistent with the observed data, and can explain the far-field surface displacements well. The vertical deformation obtained by the vertical boundary structure is larger. In contrast, the subduction slab structure can well explain the uplift of the Sino-Nepal border and the north. The optimal results of steady state viscosity ηm and transient state viscosity ηk of the lower crust of the Tibetan plateau are 1×1018 Pa·s and 1×1017 Pa·s, respectively.

By collecting and processing the data of GPS continuous observation stations in Nepal and GPS reference stations in southern Tibet of China, we obtain the GPS horizontal deformation field three years after the 2015 MW7.8 earthquake in Nepal, which shows that the post-earthquake deformation is mainly distributed in the northern part of Nepal and the border area between China and Nepal, and that deformation in the east-west direction is small, while deformation in the north-south direction is large. The whole earthquake continues to move south, and the maximum post-earthquake displacement is about 10.93 cm. The theoretical surface displacement calculated by the pore elastic rebound model is far less than the GPS observation value, which cannot explain the post-earthquake deformation observed by GPS. The post-earthquake residual slip distribution inversed by the post-earthquake residual slip model shows that the post-earthquake residual slip is mainly concentrated in the downdip extension part of the fault, and the spatial distribution is wide. The seismic moment released by the residual slip is 1.09×1020 Nm. The theoretical surface deformation caused by viscoelasticity in the forward modeling of PSGRN/PSCMP shows that the viscoelasticity relaxation model cannot explain near-field GPS observations, but its movement direction is consistent with the GPS observations in the far-field region. After inversion using the combined mechanism model of viscoelastic relaxation and post-earthquake residual slip, the seismic moment released by residual slip is reduced to 1.08×1020 Nm, and the spatial distribution is more concentrated. The results show that the inversion results of the combined mechanism model are closer to those of the stress driven model on the basis of ensuring the fitting accuracy of the model.

 Based on the information separation of dam deformation data, we use the rescaled range method to judge the dam deformation trend, and then use the optimized extreme learning machine and chaos theory to realize the deformation prediction. The rescaled range analysis shows that the dam deformation always has positive persistence, but its degree has a weakening trend. The progressive optimization of model parameters can not only improve the prediction accuracy, but also effectively improves its stability in the process of deformation prediction, and the average relative error of the prediction model is less than 2%, which verifies the effectiveness of the prediction ideas in this paper. Compared with the dam deformation trend judgment and prediction results, it is concluded that the dam deformation will further increase, but the increase range is relatively small, and tends to be stable.

 Using the gCAP method, we determine several full moment tensor solutions of 87 natural earthquakes, 18 collapse events and 9 blasting events that occurred from 2010 to 2019 in Shanxi. The results show that the gCAP method has a good effect in inverting the focal mechanism solution in Shanxi. The fault movement of Yanhu ML4.8 earthquake in Shanxi on March 12, 2016 was strike-slip and normal fault, which is an inward closed fracture with reduced volume, having an obvious non-double-couple component. Whether the proportion of dual-couple components in the full moment tensor solution of earthquake events in Shanxi is greater than 80% can be used as a basis for distinguishing between natural earthquakes and non-natural earthquake events. For non-natural earthquake events, when the parameter ζ>0, it is an outward explosion source, on the contrary, it is an inward closed source.

The accurate arrival times of Pn phase are first obtained by waveform cross-correlation technique, and then we use the sliding window correlation method to get the time differences of the sPn and Pn phases. The time differences results of sPn and Pn phases are used to obtain the epicenter depth of the M6.2 Taiwan strait earthquake, which is 14.7 km. We present a comprehensive interpretation and analysis. Finally, the results of this paper are compared with inversion results of CAP method which shows our method is more convenient, precise and faster.

Seismic phase observation reports and seismic waveform data are selected from the California Institute of Technology and US Geological Survey and the Southern California Earthquake Data Center. Relocated hypocenters of MW6.4 foreshock and MW7.1 mainshock of Ridgecrest earthquake and its aftershocks with MW≥2.5 within one month after the earthquake are independently and jointly inversed using double-difference location algorithm combined with the waveform cross-correlation technique. The relocation results show that four event clusters are formed from 2 098 events, and seismic source parameters with a relocation rate of up to 97% are achieved. The travel time residual and cross-correlation travel time residual of earthquake sequences decreases to 65 ms and 57 ms, respectively. Compared with the original result, the linear features of the epicenter are more significant after relocation, with aftershocks forming an L-shaped and T-shaped pattern, respectively. The length of aftershocks is about 90 km and the width is about 25 km. The relationship between seismic activity and fault structure is quantitatively analyzed and the results state that the Ridgecrest earthquake sequence starts deep and propagates to the shallow surface gradually. The predominant range of depth distribution is from 4 km to 8 km, showing the characteristics of segmented bands, revealing that the Ridgecrest sequence destroyed the complex fault system composed of multi fault structures with different scales, and the Little lake fault zone is the seismogenic fault of the Ridgecrest earthquake sequence.

This paper proposes to use sliding singular spectrum analysis(SSA) method for long-term gravity tide correction extraction. The experiment uses 61 days static relative gravity measurement data, and sliding SSA method to divide the data into 52 groups, 47 groups, 42 groups, 37 groups, 32 groups, 22 groups, and 12 groups according to continuous time periods of 10 days, 15 days, 20 days, 25 days, 30 days, 40 days, and 50 days. These data are used for gravity tide extraction experiments. We refer to the theoretical values to select the calculation results of the CG-5 gravimeter software (CG5_Data) and the measured gravity data using Eterna harmonic analysis to obtain the tidal wave parameters, and then we use TSoft software to calculate the results (ET_Data). The results show that when ET_Data is selected as the reference theoretical value, the RMS and STD of the residuals are smaller than those calculated by the CG-5 gravimeter as the reference theoretical value. The STD of gravity tide correction residuals extracted by different methods are all less than 11 μGal. Comparing with SSA method, when the appropriate continuous time period is selected, the extraction of gravity tide by sliding SSA method can reduce the dispersion of the result residuals. The residual RMS of 30-day and 40-day sliding SSA extraction results is 0.219 μGal and 0.602 μGal (CG5_Data) and 0.430 μGal and 0.665 μGa (ET_Data) smaller than the SSA extraction results.

The observation data of Delhi YRY-4 have been affected by environmental interference for a long time. Based on various methods, this study judges and analyzes the interference source and mechanism of borehole strain by analyzing the relativity between observation data and meteorological factors, the load influence of environmental water storage and release, and the geochemical characteristics of different water bodies. The results show that there are stable and significant annual periodic air temperature and pressure disturbances, unstable and subtle groundwater leakage disturbances on a long time scale. In a short time scale, there exists load interference of water storage and release in artificial lakes and reservoirs, which last from April to November every year.

We use the stack spectral ration method to explore thirty vertical components of earthquake data recorded by Zhongxiang seismic station, gaining the corresponding  Q0 values of coda wave of Lg-wave for each radiation path. Results show that the values of Q0 have to do with the tectonic activity. In the active regions among Huji-Shayang fault, Tongchenghe fault, Tianyangping fault and Qingfeng fault, the Q0 manifest low values change from 130 to 200, but away from the activity regions of fault, the tectonic activity is stable, and Q0 is as high as 600.