We analyze the displacement time series of the GEONET stations in Japan and the three-dimensional deformation characteristics in northeast China before and after the 3·11 Tohoku-Oki earthquake. Combined with seismic and magnetotelluric results, we discuss the present-day crustal movement and geodynamic environment of the Changbaishan volcano, northeast China. The results show that the Ou mountain is the crustal movement boundary in northeast Japan. After the 3·11 Tohoku-Oki earthquake, the stations located in the two sides of the boundary move to E direction. The effect of earthquake on the deformation in northeast China is not over yet. Affected by the plate subductions and crust-mantle magma movement, the present-day crustal deformation and earthquakes in Changbaishan volcano is related to the regional stress changes and magmatic activities.
We collect the mobile leveling data of 6 sites along the Tan-Lu fault zone and “Huoshan seismic window” in Anhui area in recent 30 years, and process the data using the “deformation rate accumulative analysis” method. We study the relationship between anomaly period and surrounding earthquakes, and analyze the overall characteristics of deformation rate accumulation of all sites in the same period. The results show that: 1) The deformation rate accumulative analysis can effectively extract the data of more than 2 times the standard deviation of each survey line, and the trend changes with smaller rates can be suppressed. 2) The deformation rate accumulative analysis not only extracts the deformation rate, but also retains the transition change of original data to a certain extent, and the frequency of deformation rate accumulation in the curve transition period increases obviously. 3) The effect of reflecting earthquakes in the period of continuous deformation rate accumulation is better, but lacks one-to-one correspondence, so it needs to be analyzed in combination with the survey line. 4) The frequency of deformation rate accumulation of each survey line in the study area differs in each period, indicating that the fault activity has certain time-sharing characteristics, and the time-sharing activity has group characteristics on a whole. 5) When the deformation rate accumulation of cross-fault leveling data in this area changes intensively, the possibility of MS5.0 earthquake in east China will increase.
In order to study the feasibility of measuring earth rotation parameters by digital zenith photographing and positioning method, we combine the principle of digital zenith photographing and positioning with the difference of space-time datum conversion, and analyze the earth rotation parameters by multi-stations and UT1-UTC measurement principle by single station. The relevant mathematical models are derived. The simulation and actual measurement verification results show that the current high-precision zenith photographing and positioning system (accuracy 0.05″) can achieve the measurement accuracy of ±5 ms for UT1-UTC parameters and ±0.06″ for polar motion parameters through reasonable station layout nationwide. In addition, the measurement accuracy of UT1-UTC by single station is 5-8 ms, which can meet the needs of specific military engineering, and provide data support for the joint calculation of ERP parameters by multi-stations. The research results can provide theoretical reference for the development of earth rotation parameters measurement in China.
Aiming at the problem that GNSS elevation time series inevitably contains gross errors, and the difficulty of detecting gross errors in nonlinear and unstable elevation time series, we construct a WT-MAD gross error detection method based on the traditional MAD method, which introduces wavelet analysis. We use experimental simulation data and the measured elevation data of three IGS stations of LHAZ, BJFS, and TWTF to conduct experiments. The WT-MAD method is compared with the 3σ method, the IQR method, and the MAD method based on least squares. The results show that the WT-MAD method established in this paper can more effectively detect gross errors in the GNSS elevation time series and provide more “clean” data for the subsequent analysis and processing of the GNSS elevation time series.
We propose a PPP partial ambiguity resolution method based on the UPD(uncalibrated phase delays) of receiver. We correct the float ambiguity using the satellite UPD product, and obtain the UPD of receiver by the trigonometric function independent of integer. We estimate the UPD of receiver based on the undifferenced method, and carry out partial ambiguity resolution according to the undifferenced narrow-lane ambiguity covariance. In order to verify the method, we select a global distributed network of 225 server stations and 14 receiver stations. The results show that, compared with PPP float solution, the positioning accuracy of the traditional method and the new method in static situation is improved by 68.3% and 76.4% respectively, and the convergence time is reduced by 56.3% and 69.9% respectively. The positioning accuracy of the traditional method and the new method in dynamic situation is improved by 65.7% and 84.7% respectively, and the convergence time is reduced by 61.8% and 71.1% respectively. The results indicate that the new method can effectively improve the positioning accuracy and convergence speed of PPP.
We select the data of 52 MGEX stations in the world for 7 consecutive days to statistically analyze the performance of two pseudorange inter-frequency bias (IFB) models in multi-system combined precise point positioning. The first IFB model estimates an IFB parameter for each GLONASS satellite, the second model uses a linear function of the frequency number to estimate the IFB parameter. The results show that the positioning accuracy of the two pseudorange IFB models in the three directions of E,N and U under static condition are equivalent, the accuracy of two models in E and U under dynamic condition are roughly the same. The positioning accuracy of the second model is 21% higher than that of the first model in the N direction. Under the same condition, the convergence speed of the two pseudorange IFB models are not much different and both can reach the specified accuracy. Since the first pseudorange IFB model needs to estimate more parameters and comprehensively consider the positioning accuracy and convergence time, it is recommended to use the second model for pseudorange IFB estimation when performing multi-system combined precise point positioning.
Based on the wavelet analysis method, using the observation data of GPS, GLONASS, Galileo and Beidou, we select the better quality data to extract the instantaneous tide level through the quality analysis of different types of signal-to-noise ratio data. The experimental results show that the sea level retrieval using multi-GNSS multipath reflectometry based on wavelet analysis greatly improves the time resolution of tide level inversion results, and the accuracy is equivalent to LSP method, which can reach decimeter level.
In order to analyze the effect of day-boundary discontinuities on GNSS carrier phase frequency transfer, PPP station clock offsets are estimated based on 10 days of GPS and BDS observations from co-location sites of multi-GNSS experiment(MGEX) and BIPM. On the basis of analyzing the statistical characteristics of day-boundary discontinuities of clock offset estimation results, we study the influence of day-boundary discontinuities on the estimation of relative frequency deviation between two clocks from the aspects of theoretical analysis and experimental verification. The results show that the day-boundary discontinuities are in accordance with Gaussian distribution, causing the jumps of relative frequency deviation at the boundary epochs of two clocks, making it difficult to accurately estimate the relative frequency deviation in the total period, seriously affect the clock frequency comparison results.
We focus on the problems of low accuracy and easy to fall into local optimum in the short-term prediction of ionospheric TEC based on neural network. We use the TEC data and geomagnetic activity index provided by the CODE center to establish an improved Elman neural network model based on the sparrow search algorithm(SSA). The BP model, Elman model and SSA-Elman combined model are used to predict 5 days continuous TEC in the middle and low latitudes during the ionospheric quiet period and disturbance period. The experimental results show that when the optimized Elman neural network model is used to predict 5 days continuous TEC, the root mean square error of single day can reach 1.443 TECu, and the correlation coefficient can reach 0.976, which is better than BP model and Elman neural network model.
Based on the GPT3 model, GNSS observation data and meteorological data of multiple radiosonde stations in Yangtze River Delta, we evaluate the precision, reliability and real-time performance of PWV calculated by GPT3 model, 2 surface meteorological data methods and GNSS, compared with the radiosonde PWV as the true value. The results show that the GPT3 model can be used to obtain real-time PWV with low precision. GNSS-PWV has the highest precision, but it needs measured meteorological parameters, which limits its widespread application. The 2 surface meteorological data methods have the higher precision, but they need surface dew point temperature data.
According to the GNSS precipitable water vapor(PWV), we carry out accuracy evaluation and correlation analysis of MODIS PWV in southwest China, construct the regional correction model and single-site correction model of MODIS PWV, and carry out a reliability test. After sub-regional MODIS PWV correction and image superposition, we obtain the corrected MODIS PWV distribution in southwest China. The results show that the regional model can be used to replace the single-site model, and the seasonal correction model of MODIS PWV in southwest China has a significant effect. The root mean square error(RMSE) in spring, summer, autumn and winter are better than 8 mm, 9 mm, 11 mm and 4 mm, respectively.
This paper uses sounding data from 2015 to 2017 to establish a China regional Tm model(BET model) that considers surface temperature, surface water vapor pressure, elevation and latitude. Using the Tm data of sounding station in 2018 as a reference value, we analyze the accuracy of BET model, and compare it with the Bevis and GPT3 models. The results show that the average annual RMSE and bias of BET model are 3.15 K and 0.04 K, respectively. Compared with the Bevis model, the GPT3 models with resolutions of 1°×1° and 5°×5°, the average annual RMSE decreases by 29.2%, 32.8%, and 39.1%; the average annual bias decreases by 96.4%, 96.7%, and 97.4%, respectively. The accuracy and stability of the model at different elevations and latitudes in China are better than the Bevis and GPT3 models.
Based on 2 074 known stations with actual measured velocities in Chinese mainland, we carry out the application research of velocity field in Chinese mainland in CGCS2000 coordinate conversion. We use different interpolation combination methods to fully consider the distance factor between unknown points and known velocity stations, so that the data of known velocity stations can be fully utilized. Compared with the calculations using velocity field grid model, the velocity accuracy of unknown points obtained by this method is improved substantially, and the accuracy of epoch conversion improves by up to 24.00 cm, 10.89 cm and 17.82 cm in X, Y and Z directions of the points, which is a guideline to improve the accuracy of epoch conversion using known velocity stations.
Based on the three-dimensional point-mass modeling approach with spatial constraint, we introduce a hydrological model to calculate the correlation coefficient between geographical points. The L-curve method is used to determine the optimal regularization parameters. The results show that the three-dimensional point-mass modeling approach after the introduction of hydrological model has higher proportion of signal-to-noise ratio greater than 0 than the method with zero order Tikhonov constraint. We use this method to monitor the drought in southwest China in 2010. We also use principal component analysis method to analyze the water storage anomaly in southwest China after removing seasonal signals. The results show that the water storage in southwest China has an obvious negative anomaly from autumn 2009 to spring 2010.
Based on the characteristics analysis of active oceanic faults, we find oceanic crust growth, symmetrical structure, fold structure and volcanic structure in active oceanic crust fault. Oceanic crust faults not only release heat, but also create pushing forces that grow new oceanic crust and leave traces of evolution on the oceanic crust topography. The growth of oceanic crust inside the fault promotes the relative movement of oceanic crust on both sides of the fault, and the comprehensive influence of the activities of many oceanic crust faults causes the relative movement of oceanic crust blocks. The mid-oceanic ridge is the product of oceanic crust growth and the transform fault region is the boundary zone of current relative movement of oceanic crust blocks.
We use the cross-fault deformation observation data of Jinzhou seismic station to analyze the current activity characteristics of Jinzhou fault. The long-term trend analysis results show that the Jinzhou fault is dominated by right-lateral and normal fault activities, and the overall activity level is relatively weak. The results of segmented analysis of fault activity show that the activity rate of northern section is higher than that of southern section. The GPS data shows that 2011-03-11 MW9.1 earthquake in Japan had a mitigative effect on stress accumulation in northeast China. Combined with the analysis of cross-fault data of Jinzhou fault, we conclude that the annual fault activity rate has been significantly weakened in 2011-2014, and the fault activity has been consistent with its background data since 2014, but the activity level is relatively weak.
Using the waveform cross-correlation method, we identify ML≥2.5 repeating earthquakes in Qingshiling, Gaizhou, Liaoning province, which are recorded by Yingkou and Xiuyan stations from 2012 to 2017. We apply the Sato single-scattering model to calculate the frequency dependence of coda Qc value of repeating earthquake sequence and variation characteristics with time. Combined with the seismic activity characteristics of Qingshiling earthquake swarm area in Gaizhou, the results show that the stress reflected by Qc change has a certain correlation with moderate-strong earthquakes. The Qc value fluctuated obviously after ML≥4.0 earthquakes in Gaizhou area, and has the characteristics of local adjustment and recovery. We infer that local stress accumulation is not enough to lead to larger earthquakes, and local accumulated energy is released by medium earthquakes for many times.
Aiming at the classification of natural earthquake events and blasting events, we use 80 natural earthquake events and 20 blasting events in Gansu and its surrounding areas to establish datasets, and apply deep learning convolutional neural network(CNN) method to build two models with different structures for training, and use 500 waveforms of natural earthquakes events and blasting events out of the training sets as test datasets. The accuracy of training and testing is more than 90%. The results show that two training models designed in this paper have a certain generalization ability; especially the Inception V1 model has good effect in the classification and recognition of natural earthquake events and blasting events.
To meet the needs of time accuracy testing of seismic digitizer, we use a specific circuit to extend the PPM(pulse per minute) output by the common reference clock source to a duty cycle of 50%. With the extended PPM, the time tag accuracy of seismic digitizer is tested at room temperature, constant low temperature and constant high temperature environments. We study the influence of ambient temperature on time tag accuracy. The results show that time tag accuracy and drift trend of seismic digitizer are affected by the change of ambient temperature. A temperature change of about 1 ℃ in half an hour can cause the time tag accuracy change of microsecond stage and the drift trend change of time tag delay. The cumulative effect of temperature change reflected by temperature to time numerical integration has good correspondence with the drift trend change of seismic digitizer. The time tag delay at constant temperature environment shows monotonous characteristics. The drift rate is determined by the crystal resonators used by individual digitizer, and it is associated with the ambient temperature.
The earth resistivity observation data of Longnan Hanwang deep well are obviously interfered by the surface and have poor seismic reflection ability. In order to suppress the interference, improve the quality of observation data, and reflect the efficiency of deep wells, we make full use of the spatial distribution of electrodes in the original well holes to optimize the observation electrode distribution, abandon the use of ground electrodes, apply intermediate electrodes instead, and build a new observation device system. The results show that the signal-to-noise ratio of observation data of the optimized device is significantly improved, the amplitude of seasonal change becomes smaller, and the discrete current generated by natural environment interference and human activities is suppressed, making the data change identification more intuitive.
