Taking the Tianshan tectonic belt area as an example, we give the crustal deformation vertical movement rate vector maps, using precision leveling and GNSS, respectively. We analyze and discuss in depth the differences and their causes. Based on the principle of complementary advantages, we give the deformation image of the fusion of the two. Based on many years’ research experience using geodetic deformation data to predict strong earthquakes, we consider that the vertical deformation map obtained by this method has more important application value in the study of strong earthquake prediction.

We propose a reconstruction method of three-dimensional (3D) multi-level deformation in time series. Firstly, OT DInSAR are applied to retrieve the discontinuous multi-level deformation in LOS direction. Then, the deformation prior fusion model and probability integration method are integrated to build the reconstruction model of three-dimensional multi-level deformation in mining areas. Finally, the 52303 workface in Daliuta coal area is selected as the research object to verify the feasibility and precision of the proposed method. After processing six high-resolution TerraSAR-X image pairs, we acquire the results of time series three-dimensional multi-level deformation for this workface. The monitoring results by our proposed method are consistent with the leveling results. The root-mean-square errors (RMSE) in the vertical, east-west and north-south directions are less than 0.188 0 m, 0.224 3 m and 0.207 4 m, respectively. Therefore, the proposed method can not only accurately measure the three-dimensional mining deformation, but also accurately retrieve the three-dimensional multi-level deformation characteristics with more details from the edge to the center of the subsidence basin.

We use the Sentinel-1 radar satellite image to discuss the feasibility of InSAR technology as an auxiliary means for monitoring the vertical displacement of the sluice. The results show that compared with InSAR technology and leveling measurement, the trend of sluice vertical displacement obtained by the two methods is consistent. The InSAR measurement results are credible and worthy of further promotion and application.

We establish a stable reference frame for the northeast China region(NEChina20) using continuous GNSS observations from 2012 to 2019. NEChina20 is aligned in scale with IGS14. The two coordinate systems are tied at epoch 2020.0. The frame stability(accuracy) of NEChina20 is about 0.5 mm/a in the horizontal direction and 0.6 mm/a in the vertical direction. The stability of the reference frame degrades with the extension of the time window and the expansion of the coverage area. We suggest that the usage of NEChina20 be limited to the Northeast and adjacent areas within the time window from 2005 to 2025. We have also developed a seasonal vertical ground-motion model using 20 years(1999-2019) of continuous observations from four GNSS stations mounted on bedrock sites. NEChina20, working together with the seasonal model, enables accurate and reliable ground and structural stability assessment in the Northeast.

In order to effectively utilize the time information of the ionospheric total electron content (TEC) series, we propose a prediction model combining the empirical orthogonal function decomposition and the neural network of long and short memory. We use the TEC grid data of Yunnan region provided by IGS to model and forecast the ionosphere at different locations and different periods. The experimental results of the prediction model in this paper show that the optimal root mean square error of TEC values for 5 days in the same period of time is 1.83 TECu, which is reduced by 16% compared with the single model, and the optimal average relative accuracy is 91.56%, which is increased by 7% compared with the single model. The optimal root mean square error of TEC values at the same location for 5 days is 1.86 TECu, which is 25% less than that of the single model, and the optimal average relative accuracy is 90.74%, which is 7% more than that of the single model.

In this paper, we test the algorithm of BDSBAS grid ionospheric, analyze the accuracy of the grid ionosphere correction model of BDSBAS with the high-precision grid ionospheric data as the reference value, and further study the effect of the grid ionospheric on the single-frequency positioning. The preliminary test shows that the ionospheric correction number and trend of BDSBAS broadcasted at this stage are consistent with the IGS ionospheric product, and the error is basically around 5 TECu; compared with the GPS Klobuchar model, the positioning accuracy after the correction of BDSBAS grid ionospheric model shows great improvement that reaches 30%-50%.

The GAMIT10.7 software is redeveloped based on the Fortran language. We use the Hopfield, Saastamoinen, Black, UNB3, EGNOS, GPT2w_1+Saastamoinen, GPT2w_5+Saastamoinen tropospheric zenith delay models to calculate the tropospheric delay in northwest China, and analyze the adaptability of different tropospheric delay models in the region. The experimental results show that in the measured meteorological data model: the Saastamoinen model the zenith tropospheric delay has the highest accuracy in northwest China. The average bias value and RMS value of each station are -1.67 cm and 3.83 cm, respectively. The Hopfield and Black models have the same accuracy. In the non-measured meteorological data models, the GPT2w_1+Saastamoinen model has the highest accuracy in obtaining the zenith tropospheric delay in northwest China, followed by the GPT2w_5+Saastamoinen model; the EGNOS model has the lowest accuracy. We find that the accuracy of different tropospheric delay models is affected by seasonality. For the seven tropospheric delay models, the absolute and RMS values of bias in summer are the greatest. The results are comparable in spring and autumn, and the absolute value of bias in winter and the RMS value is the smallest.

Taking into account the assumptions of GNSS water vapor tomography algorithm and the characteristics of water vapor density changing with elevation, we propose an automatic vertical non-uniform stratification method for GNSS water vapor tomography by setting a reasonable threshold value of water vapor density changing with height. To carry out the regional GNSS water vapor tomography test, we use GNSS observation data in Hong Kong for 6 consecutive days, using the automatic vertical non-uniform stratification method and the traditional vertical uniform stratification method. Taking the ECMWF data as a reference, the accuracy of GNSS water vapor tomography is verified. The results show that the mean deviation and the mean root mean square error (RMSE) of the wet refractive index obtained by the automatic vertical non-uniform stratification method is 3.9 mm/km and 4.6 mm/km, which decreases by about 20.4% and 23.3%, respectively, compared with the vertical uniform stratification method.

This paper uses GPS water vapor retrieval and GPS-interferometric reflectometry (GPS-IR) technology for precipitation analysis and judgment. First, considering that before and after precipitation occurs, and that atmospheric water vapor content, ground reflection characteristics, GPS tropospheric delay, and signal-to-noise ratio (SNR) amplitude (A) will change, we analyze the correlation between GPS-PWV/SNR-A and precipitation, and combine the two to determine precipitation. The results show that precipitation is positively correlated with GPS-PWV and negatively correlated with SNR-A. The addition of SNR-A data can improve the forecast rate and accuracy for precipitation judgment. Based on the troughs of SNR-A sequence and the peaks of GPS-PWV sequence, in this work the forecast rate of precipitation judgment is about 70%-82% and the accuracy rate of precipitation judgment is about 50%-60%.

We provide a validation method of multipath model of BDS-3 new frequencies. We use antenna group delay test and measured data to verify the multipath error characteristics of BDS meet the requirements of DFMC multipath model. The results show that: 1) The receving antenna used in the experiment meet the requirements of DO-373. 2) In the case of B1c/L1, B2a/L5 and dual-frequency combination, the trend of multipath error curve of BDS and GPS are basically the same; that is, the characteristics of multipath error of BDS and GPS are basically the same. 3) The multipath error characteristics of BDS in the case of B1c, B2a and dual-frequency combination can meet the requirements of the DFMC multipath model.

We calculate the cumulated Coulomb stress induced by the historical strong earthquakes based on the layered viscoelastic rheological model and the regional stress field obtained from the focal mechanism solution of the regional minor to moderate earthquakes in north China. The results show that the current moderate seismicity in north China should be related to the Coulomb stress loading induced by historical strong earthquakes. More than 85% of the moderate earthquakes occurred in the stress loading area, indicating that the main active area is controlled by the stress loading.

We use the sequence stratigraphy paleogeomorphology restoration method to restore the original paleogeomorphology of the three key interfaces of the paleogene Liushagang formation in Fushan sag of Beibuwan basin by calculating the basin subsidence. The simulation results show that the geomorphic gullies are developed vertically and horizontally, and the depressions are highly segmented. In the vertical direction, from Els3 to Els1 period, the paleogeomorphic undulation is obviously episodic, and the eastern and western depressions presents distinctive subsidence characteristics and structural patterns.

We derive the formula of spherical harmonic expansion for surface gravity changes. We analyze the dynamic evolution characteristics of gravity field around the focal area before the 2017 Jiuzhaigou M_S7.0 earthquake based on spherical harmonic method. Analysis of gravity data leads to two main conclusions: 1) The changes of gravity field around the Jiuzhaigou earthquake focal area are obvious from April 2015 to April 2017. There are transition zones of positive-negative gravity changes in the study area during three time periods, which are from October 2015 to April 2016, April 2016 to October 2016 and October 2016 to April 2017. Jiuzhaigou earthquake focal area is located in the intersection area of the transition zone in three periods. Compared with the results of the previous two periods, the spatial trend of the transition zone of positive-negative gravity changes in the last period near the earthquake source rotates about 90°. The Jiuzhaigou earthquake occurs after the direction of the transition zone rotates. The mobile gravity observation data well reflects the precursory phenomenon of Jiuzhaigou earthquake. 2) The spherical harmonic analysis of gravity changes can effectively highlight the dynamic evolution law of large-scale gravity field in the research area and restrains local detail features at the same time. This is really useful to identify the main features of gravity field changes.

In the process of gravity recovery using satellite-satellite tracking data, there are some problems caused by processing massive observation data and solving large equations, such as low computational efficiency and high requirements for hardware level of computing platform. In view of the above problems, we propose a fast heterogeneous parallel algorithm for gravity recovery using energy conservation method. CUDA is used to realize parallel computation of design matrix on GPU, MKL library is used in partition adjustment method and preconditioned conjugate gradient method for constructing and solving normal equation on CPU fast, thus realizing heterogeneous parallel computation of gravity recovery using satellite data. A 120×120 gravity field model, named GM-GraceFO2020h, is obtained by processing the observation data of GRACE-FO satellite from January 1, 2020 to June 30, 2020 using this algorithm. Compared with the existing models and algorithms, the result shows that the accuracy of the model derived by the proposed algorithm is equivalent to that of existing GRACE gravity field models, and compared with the traditional serial algorithm, the inversion time is reduced by 98.479%, and the memory consumption is an order of magnitude smaller.

This paper first discusses the accuracy and feasibility of using different activation functions of BP and RBF neural network methods to fill the gap data of GRACE and GRACE-FO satellites, and fills in the missing data based on the optimal scheme. We use the ITSG-Grace2018 and ITSG-Grace operational time-varying gravity field models to derive the changes of TWS in the Yangtze river basin (YRB) from 2002 to 2020, and finally, combining with the GLDAS model, precipitation, temperature, the Yangtze River Water Resources Bulletin and other data, we comprehensively analyze the changes of TWS. The research results show that: 1) The BP neural network algorithm with the hidden layer activation function as the rectified linear unit (ReLU) is effective in filling the data gap between GRACE and GRACE-FO satellite missions; 2) TWS changes in the YRB have certain regional differences, which are mainly manifested in TWS increase in eastern part of the upstream and most part of the midstream at a rate of about 5 mm/a, and decline in the upstream mid-west, while the downstream is basically unchanged. The GRACE/GRACE-FO long-term series time-varying model can reflect the drought in 2019 and the floods in 2017 and 2019 in the YRB.

Aiming at the difficulty of accurate prediction of seismic observation data, we propose a regression model based on nuclear mixed effects. In order to verify the feasibility of the algorithm model, we perform a simulation experiment with the output data of the geophysical instrument of the Hubei Seismic Station and compare it with the traditional neural network algorithm. The results show that the model can accurately predict the seismic and geophysical observation data and the performance is better than other neural network algorithms. The relative error of the water temperature and water level data prediction is less than 0.05% and 0.48%. The proposed model provides a new research idea for earthquake monitoring and forecasting personnel to accumulate and analyze basic earthquake data. At the same time, it also provides practical foundation and research possibilities for more complex deep learning algorithm framework models.

Through the deep processing of the original station displacement time series data of 43 GNSS continuous stations in Yunnan from January 2011 to December 2018, based on the calculation of the strain parameter grid time series, the surface strain/maximum shear strain short-term anomaly indicators suitable for the short-term and imminent prediction of M≥5.0 earthquakes in Yunnan area are extracted respectively. Based on 27 earthquakes with M≥5.0 that occurred during the period, we use the R value scoring method to evaluate the predictive efficiency of the anomaly indicators. The results show that: 1) Two anomaly indicators have a good indication of the occurrence time of M≥5.0 earthquakes in Yunnan area. 2) When the forecast window length is set to 30 or 60 days, the prediction accuracy of the short-term and impending anomaly index of the maximum shear strain is higher than that of the surface strain under the same prediction window length. 3) Through the comparison of different prediction window lengths on the same anomaly index, we find that the best prediction window length of the surface strain short-term and imminent anomaly index is 90 days (for 27 earthquakes, 23 were reported correctly and 4 were missed. The predictive accuracy rate is 85.19%), and the best prediction window length of the short-term impending abnormality index of the maximum shear strain is 60 days (23 correct reports, 4 missing reports, and the predictive accuracy rate is 85.19%).

In this paper, based on the tidal wave of the telluric field, we establish a model of the seepage(movement) of the rock crack water(charge) seepage, and calculate the dominant azimuth angle α  of the telluric field in Liaoning area. We make a comparative analysis on the degree of change of the dominant azimuth angle α of the telluric field that we combined with typical disturbances such as rainfall, temperature, and magnetic storms. The results show that the dominant azimuth α of the telluric field is limited by typical interference. Combined with the dynamic change of the dominant azimuth α before and after the 2013-01-23 Liaoning M_S5.1 earthquake, we find that the long-term sudden change of the charge movement direction of the rock mass fissures of the same or adjacent geological structure geoelectric field stations can be considered as the precursory anomaly of a moderately strong earthquake.

From middle July to late August 2020, quarantine measures are taken in Xinjiang due to the impact of COVID-19. We select the waveform data of 28 broadband digital seismometers from Xinjiang network from June to September 2020, obtain the corresponding PSD (noise power spectral density) and PDF (probability density function), and count the distribution of PSD values in different frequency bands according to the situation. We analyze the characteristics of noise level changes in Xinjiang during the epidemic period. The results show that due to the impact of COVID-19, the noise level in the high frequency band (2－20 Hz) in Urumqi has dropped by 4.5%, and other areas have experienced different degrees of reduction. The main influencing factor is the continuous reduction of human activities. At the same time, this noise blanking phenomenon is also transmitted from the surface to the underground.