Earthquake models are restricted to the academic and scientific level| further, they evolve over time.  We researched the history of earthquake model evolution: models that consider only the hypocenter region (fault zone)→models that consider hypocenter region together with surrounding tectonics→models that consider upper crust brittle layer and deep crust ductile layers coupling with upper mantle→models that consider the earthquake as an activity of present continental deformation dynamic system’s self-organization evolution process.  The evolutionary trend is from hypocenter to dynamic system over a hundred years.  Its characteristics are: time-space domain expands, dimension rises of the models| from reductionism mechanism to holism| from classical mechanics to complex system theory and nonlinear dynamics| new models contain and surpass old models, and are better approaching nature.  Benefitting from the unprecedented achievements of seismic geodesy, strong correlativity and interdisciplinary methods, new earthquake models are developing.

This paper presents the formula of the inverse Poisson integral, which is used for continuation of airborne gravity anomaly and to assess the accuracy of airborne gravity data.  The downward continuation of airborne gravity data is performed by FFT and iterative methods.  The accuracy is evaluated based on the comparison and analysis of downward continued results with terrestrial real gravity data.  The choice of low-pass filters for downward continuation is also discussed. The experiment shows that the FFT method obtains better results compared with those derived from iterative methods.

The influence of Newton central differentiators upon acceleration in airborne gravimetry is discussed.  It is concluded theoretically that there are almost no differences between the results of Newton central differentiators after low-pass filtering, when the effective band of airborne gravimetry is located in the pass bands of Newton central differentiators.  The experimental results show that when the sampling interval is 1 s, the average velocity is 155 m/s, the resolution of gravity data is 10 km, and the standard deviations of the differences between the results of Newton central differentiators are less than 2.1 mGal.

This paper introduces the continental tectonics environmental monitoring network of China in 2013-2014, and analyzes the absolute gravity measurement results in the western region.  The study shows that the results are characterized by 89.5% reliability, with precision better than 5 μGal.  These results satisfy the technology requirements of the Continental Tectonics Environmental Monitoring Network of China for absolute gravity measurement.

There are large differences of Moho local features in the southern area  of the north-south earthquake belt.  This article applies the Parker-Oldenburgiterative inversionmethod with constant density and variable density models on the southern area of the north-south earthquake belt.  The results indicate that the variable density model’s reverse results are more reliable, and that the two reverse results’ general morphological characteristics are consistent.  Between the differences of theinversion results, the maximum difference is less than 2 km, and the results of the variable density model are much deeper.  The mantle uplift in Panzhihua and the mantle concave in Dongchuan are extremely significant, indicating that deep geologic processes have important influence on regional tectonic earthquake activities.

The paper firstly analyzes the relationship between the area of Bouguer gravity anomaly and faults.  It mainly utilizes measured gravity section (Suhaitu-Pingluo-Etuokeqi), and gets a more exquisite section of the Bouguer gravity anomaly, and a figure of normalized full gradient of the section of the Bouguer gravity anomaly (G^H). The results of the section show that: 1) The main active faults are corresponding with G^H contour strip or cut local region, which generally show strong deformation characteristics.  For example, the Bayanwulashanshanqian fault (F1) or the southeast boundary of Alxa block is in accord with the western change belt of G^H| the belt is about 10 km wide and extends to about 30 km.  2) The seismogenic structure of the Alashanzuoqi Ms5.8 earthquake is not Dengkou-Benjing fault (F3), but is controlled by F3.  3) The Yinchuan-Pingluo fault (F6) is the seismogenic structure of Pingluo Ms8.0 earthquake, and its focal depth is about 15 km.

In this paper we use the EGM2008 gravity field model to calculate the height anomaly, combining this with SRTM and DTM2006.0,to futher calculate the remaining terrain model. We then calculate the RTM height anomaly, in order to compensate for the lack EGM2008 gravity field model shortwave signals. We subtract the RTM height anomaly form the measured height anomaly in the gravity field model to get the residual abormal anomaly height. The surface fitting method using residual height anomaly model allows us to solve problems of elevation benchmark system deviation. The calculations are proven by actual GPS/leveling data and  comparison of the results without RTM.The method can significantly improve the abnormal elevation accuracy of GPS points,
refining for regional geoid purposes

Based on the fitting of GPS height measurement with two engineering examples, this paper analyzes how to achieve the third, fourth class leveling’s requirements. We compare the evations obtained from the GPS height fitting and conventional level, statistically analyzed the precisions of GPS/level observation elevation fitting, analyzed the influence of GPS observation grade to the precisions of GPS/level observation height fitting,assess the level’s grade of GPS elevation fitting by using the conventional level tolerance,and anazlyzed the practicability of GPS elevation fitting. The results show that: the GPS height fitting measurement can replace the leveling measurement under certain conditions.

The study of several gravity field interpolation algorithms indicates that there is a large interpolation error in the high frequency regions of gravity field| furthermore, we can achieve increased interpolation accuracy for a small price by supplementing measuring points in the high frequency region.  Wavelet analysis can perceive the spatial frequency information of a gravity field.  This paper proposes a strategy to effectively improve the interpolation accuracy of a gravity map.  Our strategy, which takes the wavelet detail information of gravity reference map with a certain resolution as the transition probability, uses a random sampling scheme to add measuring points in the high frequency region of a gravity field.  The contrast with the scheme of increasing measuring points distributed uniformly shows that, in the case of achieving roughly the same degree of improvement of interpolation precision, the strategy uses fewer measuring points, and has a smaller extreme value of interpolation error.  It therefore has higher efficiency.

With data obtained from a high precision strapdown inertial measure unit, in this research, static experiments are carried out for the land based inertial gravimetry, initially examining the influence factors for static inertial gravimetry.  Results from our static experiments show that with a moving average filter, most high frequency noise from the quartz flexible accelerometers is mitigated, and the corresponding standard deviation of gravity observation that derived from the test IMU is achieved at about 1.5 mGal with 30 s averaging.  However, due to instrument instability and observation environment, an upward temporal trend in the observed gravity value can still be noticed.  By modeling the temporal trend with regression analysis and detecting the initial unknown IMU observation bias from the beginning known gravity point, the error of observed gravity value can be calculated and compensated.  Using this method, in our vehicle-based experiment the observation precision at most ZUPT points is successfully improved from 12-20 mGal to 2 mGal. 

The ice sheet 3D model, which includes ice surface, subglacial topography and internal structure, efficiently combines the internal and external ice sheet structure.  It plays an important role in research on the mass balance of the Antarctic ice sheet, ice core drilling and ice-sheet models.  With the Bedmap2 data and ice radar data collected during the 29th Chinese National Antarctic Research Expedition (CHINARE 29, 2012/11), we build an Antarctic local ice sheet 3D model of an 11.3 km × 11.5 km region located in the Gamburtsev mountains on the cross section from Zhongshan station to Dome A.  This article highlights the data processing during the building of the 3D model.  The ice radar collection uses the ice radar system developed by China.  In the paper, the processing of ice radar data is described in detail, including data pre-processing, conventional image corrected technology and the gain of subglacial topography.  The depth of isochronous layers and bedrock layers can be extracted from the ice radar data.  We then generate the isochronous layers DEM (elevation 1 729 m-2 718 m) and bedrock layers DEM (elevation 2 601 m-2 950 m) with a suitable interpolation method. Combining with the ice surface DEM (elevation 3 679 m-3 745 m) extracted from the Bedmap2 data, we finally build a 3D model that includes a topographic of the ice surface, isochronous layers internal ice sheet and the subglacial bedrock.  We then examine the model and make a preliminary analysis.

The Gaoqiao fault zone, which extends along the strike direction for more than 40 km, is a large-scale fault in the Three Gorges dam area.  There were a series of seismic activities along the Gaoqiao fault zone before and after impoundment of the Three Gorges Reservoir.  In this paper, the authors investigate the fractal dimension and mineralogy of the clay-rich fault gouge samples collected from the Gaoqiao fault zone based on grain size distribution and X-ray diffraction analysis.  In this study, the main clay minerals of fault gouges are illite and chlorite, and the content of clay minerals is generally less than 35%, which is conducive to the stick-slip.  The results demonstrate that the fractal dimension of fault gouges in the Gaoqiao fault zone range from 2.20 to 2.67, which indicates that the fault is essentially characterized by stick-slip.  The mechanical properties of the fault and the fractal dimension of fault gouges show that it is divided into two segments.  On the basis of the above, coupled with the analysis of seismicity, it is clear that we should pay attention to the moderate-sized earthquakes occurring on it in the future.

By using the least-square method constrained by the co-seismic deformations of the Wenchuan earthquake derived from GPS, we inverse the slip models, with and without considering ramp-decollement separately. Based on the slip models obtained, we simulate the co-seismic gravity changes. The results show that the co-seismic horizontal deformation of the GPS stations and the gravity changes of the gravity stations modeled by the slip model with ramp-decollement are both closer to the actual measurements.  This verifies that a large-scale nearly horizontal ramp-decollement,which existed at the southern part of the mid-fracture of the Longmenshan, has participated in the co-seismic slip of Wenchuan earthquake.

By using the least-square method constrained by the co-seismic deformations of the Wenchuan earthquake derived from GPS, we inverse the slip models, with and without considering ramp-decollement separately. Based on the slip models obtained, we simulate the co-seismic gravity changes. The results show that the co-seismic horizontal deformation of the GPS stations and the gravity changes of the gravity stations modeled by the slip model with ramp-decollement are both closer to the actual measurements.  This verifies that a large-scale nearly horizontal ramp-decollement,which existed at the southern part of the mid-fracture of the Longmenshan, has participated in the co-seismic slip of Wenchuan earthquake.

In this paper, on the basis of cross-fault flow deformation layout monitoring and geological structure research results collected on north China, the layout monitoring time and monitoring cycle, the cross-fracture characteristics of monitoring sites, monitoring sites distributing area, type of monitoring sites and measuring pile type are analyzed.  The results show that the layout monitoring situation of different monitoring units varies considerably, that layout measuring time and monitoring cycle are not uniform, that there is a monitoring blind area, that cross-fracture activities corresponding with regulatory requirements are lower, and that most monitoring sites require only standard monitoring.

A series of earthquakes (the Xianyou ML5.0  Earthquakes series,totally 73 events), recorded by digital seismic network of Fujian province, are selected to calculate the apparent stress in this region.  The paper gives the time distribution and the area of the apparent stress level.  The features of apparent stress change before and after the main shock are discussed.  Finally, the relation between apparent stress, magnitude and hypocenter depth is obtained.

Along with the growth of GNSS, the number of GNSS satellites isalso increasing. The combination of different GNSS satellites could guarantee improved satellite availability. In this paper, unifying multi-GNSS time system is analyzed.  Further, a coordinate system and a multi-GNSS single point (SPP) positioning model and a method of determining the weight matrix for multi-GNSS are implemented. The performance of the combined GPS/GLONASS/BDS/Galileo SPP is assessed using the observations from 8 GNSS stations. The position performance is compared between standalone GPS and multi-GNSS with 15°,30°and 45° cut-off elevation angles.The results show that 15° cut-off elevation angle satellite availability is best, and that multi-GNSS advances positioning accuracy only a little. When cut-off elevation angle is 30°or 45°, the satellite availability is worse| standalone GPS is often unable to ensure continuous and accurate positioning due to lack of available satellites, but multi-GNSS still could guarantee the continuity and integrity of real time positioning. In the BDS coverage area, the rate of epochs that can be used for locationis over 95%,and the horizontal direction error is less than 30 m. This paper has reference value for navigation in complex environments, such as an urban environments and thick forests

The algorithm of real-time GLONASS satellite clock corrections is investigated and the real-time satellite clock corrections are used for RT-PPP.  Based on software developed by ourselves, undifferenced ionosphere-free phase and range observations collected from global GNSS stations are used to estimate real-time GLONASS clock corrections.  The results show that the real-time GLONASS clock corrections are comparable with the ESA final precise clock corrections.  The RMS of real-time clock is better than 0.5 ns with respect to the ESA final satellite clock.  The accuracy of GLONASS RT-PPP is a centimeter in static mode, and better than 30 cm in simulated kinematic mode, when real-time satellite clock corrections are used.

Since orbit determination accuracy can be affected by the distribution of stations, the relevant formulas are derived.  In this paper, one station selection strategy, taking into account the number of stations, the quality of observation data and the geographic distribution of stations, is proposed.  Using the global observation network data of IGS and iGMAS, three different station distribution contrast schemes are designed.  Then, the GPS precise orbit determination experiment is carried out.  The results show that the orbital accuracies of three schemes with well-distributed stations are obviously better than those which do not consider station distribution.  The 3D orbit accuracy can reach about 6.0 cm using 30 well-distributed global stations, and the level of 3.5 cm using 40 stations.  The accuracy level reaches 2.8 cm, if 50 stations are used.  However, extra stations beyond 50 have almost no effect in improving orbit accuracy.

We use the singular spectrum analysis method to spatial filter with GAMIT/GLOBK processing of high precision CZCORS coordinate time series.  To exclude common mode errors of coordinate time series, a maximum likelihood estimation method is used for noise characteristic analysis.  The optimal white noise model is “WH”, the optimal colored noise model of the N, U components is “WH+FN”, and the optimal colored noise model of the E component is “WH+FN+RWN”.  Employing the optimal colored noise model to estimate the regional CORS sites velocity field, shows a southeast overall movement trend, and the overall displacement value and movement direction is similar to the surrounding IGS stations.

Using the observed data of IGS stations, a Chinese regional ionospheric delay model is established using a least squares polynomial model with restrictive conditions.  Additional restrictive conditions are imposed in the spatiotemporal domain on the model, in order to improve the model’s continuity.  As a result, the polynomial model with restrictive conditions can somewhat more accurately reflect the time-varying characteristics of the ionosphere.  We can use the model research and analysis to determine diurnal variation and seasonal variation characteristics of region ionospheric delay in China.

This paper provides a detailed analysis of the ionosphere using GEO satellites of China’s Beidou navigation satellite system.  Taking Curtin University Beidou continuous observation station’s data as an example, the paper extensively analyzes the ionospheric TEC variations in a sample day in equinox, summer and winter, as well as seasonal variations in 2013 and 2014.  In addition, based on the temporal variations in TEC, the paper improves the Klobuchar ionospheric model broadcast by GPS satellites.   Finally, the effect of the improved model is validated using a year’s data.  The paper shows that the accuracy of the new Klobuchar model is obviously superior to the former model.  The precision of the new model increases more than 10% in most months, and it reaches over 64% in computing the ionosphere vertical TEC.

This paper mainly uses the interval quartering algorithm of singular spectrum analysis iterative to do interpolation about GPS coordinate time series. This method extracts main characteristic components from an incomplete time series of coordinates by an adaptive filter.  Based on the improved singular spectrum analysis iterative interpolation, it is then compared with the Lagrange interpolation method.  Our research shows that the interval quartering algorithm of singular spectrum analysis iterative not only has higher interpolation precision, but also solves the interpolation efficiency problem. The computational efficiency is greatly improved.

Based on the Hurst index characteristics of noise series, a new method determining the smoothing parameter for semi-parametric models is proposed in this paper: a binary search method based on the Hurst index.  We apply it to analyze simulated and real GPS coordinate series, and the numerical results show the proposed method can effectively determine the smoothing parameter, and the seasonal signal of GPS coordinates time series can be completely separated by using the determined semi-parametric model.

This paper is concerned with the problem of lack of CORS stations when using GPS data to correct the DInSAR atmospheric delay error.  The paper puts forward a generalized curve surface fitting interpolation model and compares the precision of modeling regional tropospheric delay by using the inverse distance weighted interpolation model (IDW) and the curve surface fitting interpolation model (CSF), based on GPS observation data and SAR imaging data from Beijing.  Further, it tries to conduct an atmospheric delay correction for SAR differential interferograms.  The results show that compared with IDW, CSF can better reflect the delay features of the regional single difference tropospheric delay.  In addition, the interpolation precisions of the two experiments are 7.1 mm and 3.5 mm respectively.  By using CORS observation data and the CSF model, the atmospheric delay errors in DInSAR are effectively reduced and the precision of DInSAR measurement results is improved.

Aiming at the high precision requirements of position and orientation, a post-processing algorithm based on tightly coupled carrier phase DGNSS/INS and smoothing is proposed.  Ambiguities in resolution, navigation status errors and IMU errors are estimated by two Kalman filters| RTS smoothing is used in data post-processing.  Test results indicate that high precision positioning and orientation can be achieved, and that the algorithm is an effective data processing method.

Based on the characteristic and applicable scope of plane fitting, quadratic surface fitting and GA-BP neural networkmodels, nonlinear and linear combined methods are proposed to integrateadvantages of each model and improve the accuracy and reliability of height fitting.  We consider an RBF neural networkcombination, a weighted least squares support vector machine (WLSSVM) portfolio, an optimal weighted combination and an optimal non-negative variable weight combination.  The consequences of these different combined methodson GPS height fitting precisionare compared and analyzed.  The results showthat different combination methods generate different accuracy of GPS Height Fitting.  The WLSSVM and optimal non-negative variable weight combinations are superior to the others: they have stronger reliability, can better control the residual extremes, effectively shorten the error range, and havehigher conversion accuracy.

With respect to 3D deformation monitoring of a sea-crossing bridge, the schemes of measuring robot and 3D laser scanner are implemented.  For both, the maximum range of 3D deformation monitoring data is less than 0.5mm.  The monitoring data of the measuring robot is high precision, and the 3D laser scanner technology applied to bridge deformation monitoring is feasible.  Based on the wavelets and Kalman filtering theory, the Kalman filtering model of bridge deformation monitoring after wavelet denoising is established.  The experiment shows that applying Kalman filtering through the monitoring data after wavelet de-noising processing improves the reliability of the deformation analysis.  Comparison of the results obtained by the methods of standard Kalman filter data and adaptive Kalman filtering, and the analysis of sum of squares error, mean square error and average relative error precision, shows that the adaptive Kalman filtering method is superior to the standard Kalman filtering method.

A unified approach for cartesian coordinate transformations is presented.  Firstly, the unified expression of transformation equations in two- and three-dimensional space is given, and then the orthogonal Procrustes analysis is used for estimating the transformation parameters.  The orthogonal procrustes analysis provides a non-iterative solution for transformation parameters without any prerequisites on the magnitude of the parameters.  Consequently, neither good initial values nor iterations are required during the parameter estimation process.  Through numerical examples, it is illustrated that the proposed approach is capable of giving accurate, efficient, stable and reliable parameter estimates and that it is applicable for cartesian coordinate transformations in various situations

Nearly all discussions about attaching linear inequality constraints are based on parameter adjustment.  This paper proposes an approach to solve condition adjustment models with linear inequality constraints.  The main idea includes taking constraints on the Lagrange multiplier according to the K-T theory, calculating the multiplier through an iterative algorithm, and then obtaining the optimum solution of the unknown parameters.  The feasibility of this method is illustrated in an example.

Currently, using the point cloud simplification, it is hard to achieve high precision, superior simplification rate and high speed.  This paper proposes an adaptive point cloud simplification algorithm.  Firstly, we use the PCA to estimate the normal of each point and to compute the angle between the normal vector and the reference plane.  The characteristics of the surface can be determined by the local entropy of normal vector angles.  The superior results of simplification can be derived according to the different simplification rate, counter to the characteristics of the surface.  The results show that the proposed approach can reach a balance in the simplification precision, simplification rate, and simplification speed. 

Using the National Institute of Metrology(NIM) designed iodine-stabilized He-Ne laser as a substitute for the FG5’s Winters Model 100 iodine-stabilized He-Ne laser, we do some experiments of comparison.  The test results show the NIM’s iodine-stabilized He-Ne laser can properly work with FG5 absolute gravimeter, and the NIM’s iodine-stabilized He-Ne laser could totally take the place of the original Winters Model 100 iodine-stabilized He-Ne laser.

A broadband seismo-electromagnetic radiation observation system is developed for better seismo-electromagnetic signal identification under complex global electromagnetic environment. The system is composed of a space electric field sensor with frequency band of 0.1 Hz-100 kHz and a high speed data acquisition device with sample rate of 512 kHz and clock synchronization precision better than 100 ns. The experiment at 6 sites in Yunnan province shows that the system is capable of recording clear waves of electromagnetic pulse and the signal from the same source at different sites in a large area.  This is ideal for identifying seismo-electromagnetic signals through detailed signal information and position.

This paper briefly introduces the role of the seismic monitoring system used in nuclear power plants, and carries on the design of its hardware and software based on the standards of seismic monitoring systems. Wuhan Institute of Seismologic Instrument produced a series of products, which are successfully applied in practice.  Finally, combined with the current development trends of nuclear power plants, the paper presents the developmental requirements of the seismic monitoring system.

The duration time of a mainshock’s wave increases with epicentral distance.  Aftershocks that occur within the duration time are always merged in the coda wave and missing in existing catalog earthquake swarm.  By comparing duration times of Ludian 6.5 seismic waves, we get at different epicentral distance.  It is argued that strong acceleration recording near the source region is good for merged earthquakes detection because the duration time of the mainshock is shorter. Then twice aftershocks merged in coda wave of Ludian 6.5 are detected using strong acceleration recording near the source region| the magnitudes are M4.6 and M3.6 individually.

Based on big data research methods, earthquake precursory observation data of many items prior to the Ludian earthquake is used to detect abnormalities.  The combination of the anomaly detecting results fit well with this earthquake.   Furthermore, calculation of this region shows that the results obtained from the method introduced in this paper relate well to earthquakes greater than 5.0.  The method used in the paper is a trial to introduce the big data idea to the earthquake observation data application.

We collected the continuous thermal infrared data of MODIS/Terra satellite remote sensing in the Pingguo 4.5 magnitude earthquake zone from December 2012 to March 2013.  We processed it through the clouds, selected the thermal infrared data between 5:00-7:00 a.m. in Beijing time (the best period for observation to conduct land surface temperature data retrieval), analyzed the time evolution process of land surface temperature anomalies before and after the earthquake, as well as the relationship between abnormal spatial distribution and active fault.  Further, we discuss the impact of non-structural factors such as earthquake zone topography, landform, and seasonal weather on land surface temperature anomalies.  The results show that: 1) There was a phenomenon of thermal infrared abnormal warming near the epicenter two months before the earthquake in Pingguo, and there was a correspondence between the abnormal warming and earthquake occurrence time.  Significant warming occurred in the first half of the month, while 7 days before the earthquake the abnormal warming reached its peak, and dropped rapidly after the earthquake.  2) Through the relative analysis of non-structural factors such as earthquake zone topography and landforms, seasonal weather, the structural “warming” information, which was anti-seasonal and due to the abnormal warming of the earthquake zone.  3) Abnormal warming developed from the epicenter along the faults in the NW direction, and it was relatively consistent with the seismogenic tectonic of NW direction.  After full consideration of the influence of non-structural factors such as earthquake zone topography, landforms and seasonal weather on the abnormal warming, it is inferred that this thermal infrared warming could be a short imminent anomaly phenomenon before the earthquake.

Examining the degree of influence of various factors affecting the change of water level of Taiyuan well, the paper presents the discussion and research on the rapidly rising water level of Taiyuan well.  The results show that the rapidly rising water level is not only closely related to increased rainfall and reduced mining underground water of Jinci Spring, but also has a certain relationship with the size and direction of the stress field of Taiyuan basin.  Therefore, it is the common result of non-structural factors and tectonic activities.