On the basis of introducing the empirical mapping function NMF and the dynamic mapping functions VMF1 and GMF, using one year’s grid data provided by ECMWF, the time varying characteristics of these three mapping functions and their variations with height, latitude and elevation angle are concretely researched and analyzed are compared. Finally,we compare  the influence on slant path delay with the variations of elevation angle in the three kinds of mapping functions are compared.

Based on analysis of GPS observation data decoding and data quality indices, we compile a new program for decoding the original observation data and calculating and analyzing the quality indices, such as multipath effect, cycle slips and so on. The quality indices can be displayed intuitively by graphics in real-time. Moreover, the experimental verification is carried out through engineering practice, and the results prove the correctness and reliability of the designed software, providing guidance and help for practical work.

We introduce the basic accuracy assessment theory and method of broadcast ephemeris. We then analyze and compare the short and long-term trends of the broadcast ephemeris accuracy, clock bias accuracy and SISRE accuracy of different types of BDS satellites, using 28-month-long broadcast ephemeris data from January 2013 to April 2015. The results show that the orbital accuracy of GEO satellites is about 1.8 m, and the clock bias accuracy is better than 6 ns. As for the IGSO and MEO satellites, orbital accuracy is better than 1.2 m and clock bias accuracy is better than 4 ns. These results are better than the accuracy of GEO satellites overall. The long-term trend of BDS broadcast ephemeris accuracy indicates that its accuracy has a tendency to improve gradually.

This paper first analyzes the accuracy and stabilities of multi-GNSS experiment (MGEX) BDS  DCB. It then performs a comprehensive analysis of the difference of daily and weekly DCB, as well as computer noise components and parameters of the BDS DCB. The results indicate that the IGSO DCB is the best, followed by MEO, while the last is GEO. The accuracy and stability of DCBB1B2 is better than the others, DCBB2B3 is the worst. A large amount of experimental data shows that the noise component of BDS is mainly Gaussian white noise (WN) as well as first order Gauss-Markov (GM) process noise. The DCB WN is 0.16 ns, 0.15 ns, 0.09 ns; the GM process noise is 0.093 ns/d, 0.049 ns/d, 0.058 ns/d, and the correlation time is 42.63 d, 85.79 d, 42.63 d.

According to the characteristics of GM (1,1), BP neural network and Kalman filter, we propose an optimal non-negative variable weight combination forecasting model based on grey correlation. This new model can control the effects of the errors. The calculation results are compared with the variable weight combination model based on the error square of minimum and the optimal weighted combination model or the three single method. The results show that the new model has the highest accuracy.

The Gauss-function model, based on least square collocation, is used in anomaly assisted identification of regional cross-fault deformation characteristic curves. The precursory anomaly judgment examples of regional deformation of cross-fault in Lushan Ms7.0, Kangding Ms6.3, Ludian Ms6.5 and other earthquakes, show that least square collocation can be fitted in regional cross-fault deformation and helps in identifying anomaly characteristics. The method has certain significance for the deformation of cross-fault identification．

The influence of observation precision to inversion results is studied from singular value, the distribution of data and a priori information of the inversion model. To take the reciprocal of the shortest distance between observation and the subdivision fault for weight, the model considering observation precision is established. On the basis of the dilocation model, the velocity of Longmenshan fault is calculated by GA. The results show that the data out of singular values, with the model considering observation precision, on the premise of reasonable a priori information, are stable and reliable.

This article analyzed the strain field and sectional strain accumulation characteristics about Qinghai-Tibet block with the 2004-2013 GPS velocity field data. The results show that: 1)The stable Alashan and Ordos blocks stop the migration for northeast on the Qilian and Helan faults for east; 2)The high value areas of principal compressive strain rate and the maximum shear strain rate distribute the eastern of Qilianshan faults and from the eastern of Haiyuan fault to the western of Liupanshan fault, the high value areas of compressibility surface expansion distributes at the eastern of Qilianshan faults; 3)The relative movement features are mainly compression and both left-lateral strike-slip of the eastern of Qilianshan faults, mainly left-lateral strike-slip and both compreesion from the eastern of Haiyuan fault to the western of Liupanshan fault.

Using Bohai Bay region GPS rate data from 1999 to 2013, Ying-Wei fault zone is divided into two blocks, and we calculate the principal strain rate and the strain rate on the fault zone. The regional strain-stress field calculation results are studied. Results show that the 1999-2011 Bohai Bay regional crust as a whole presents NEE-SWW compressive stress and NNW-SSE tensile stress effect. For 2011-2013, the regional crust presents a N-S direction, tensile stress is nearly in an E-W direction. The main cause of the change is the Japan 3.11 earthquake in 2011. Bohai Bay area is different in localized changes obviously. The Ying-Wei fault zone presents normal strain and the shear strain rates are also different at different times; in 1999-2011 it presents dextral shear strain-shear stress, whereas in 2011-2013 it shows sinistral shear strain-shear stress.

Based on seismic phase date recorded by Sichuan digital seismic network, this study analyzes the spatial and temporal variations. In this study, we employ the idea of repeat earthquakes, select seismic active area as generalized repeat earthquakes, and analyze the velocity ratio of the cluster zone to different stations around. The results show that the velocity ratio obviously decreased before the Wenchuan earthquake and increased after the Lushan earthquake. These changes can be observed both north and south of Longmenshan fault. The temporal changes of velocity ratio from 2000 to 2013 show the integrity and difference between Wenchuan MS8.0 and Lushan MS7.0. It seems that it was the same dynamic process, the two large earthquakes occurred in this process.

Tunnel excavation encounters a variety of geological challenges, such as faults, caves, etc. Currently tunnel seismic prediction adopts TSP, which is used widely, but research about TSP is currently focused on its engineering application examples. In this paper, we use the finite element method to simulate the tunnel seismic wave field, employing wave field snapshots and combined time recording methods to study the impact of caves on the characteristics of the propagation of seismic wave field tunnel. We also employ inversion processing of the time record of the model containing the cave. We compare our method with the velocity scattered image and reflection interface position, velocity processing with the default values set by using TSPwin. The results are good for tunnel seismic velocity cave layered model with single prediction, and have a higher prediction accuracy of the P-wave. Finally, we process and analyze engineering examples to verify the numerical simulation the conclusions.

Based on the Helmert condensation and the KTH methods, this paper studies the effect on regional geoid modeling caused by the lateral density perturbation. The simulated research shows that the effect on geoid modelling caused by lateral density variation generally does not exceed one centimeter in plains and less undulating areas; these variations can be ignored within the precision limitation. However, for mountainous areas, this effect reaches to the centimeter or even decimeter level, which should be considered carefully in centimeter-accuracy geoid modelling. These effects show significant variation in mountainous areas; the errors are supposed to exist in the results derived from the Helmert condensation method due to neglecting the outer-zone topographical effects. In total, for regional geoid modeling in mountainous areas, various geoids based on variable crust density should be computed by different approaches, and GPS/leveling data can be used to choose the optimal methodology for geoid modeling in these regions.

Based on information on evenly distributed IGS stations, the paper selects 150 healthy IGS stations distributed in coastal areas and 29 inland IGS stations.  Displacements of IGS stations caused by non-tidal ocean loading are calculated using an ocean bottom pressure grid provided ECCO, results of which are used to analyze differences of effects of non-tidal ocean loading on IGS stations between coastal and inland areas. Finally, to investigate its effects on IGS station coordinate time series, the calculated displacements are used to correct coastal IGS station time series provided by Scripps orbit and permanent array center(SOPAC). Results indicate that U components of stations in coastal areas (less than 50 km) are affected on average by 5 mm, compared to only 1 mm for inland stations. After non-tidal ocean loading correction is applied, the RMS of height time series of 71% coastal area stations are reduced, with a maximum reduction of 13%. Furthermore, non-tidal ocean loading effects along the west coasts of the Atlantic and the Pacific ocean are more obvious than along the east coasts. Considering the obvious effects in these areas, it is strongly recommended that when using station coordinate time series in west coast areas of the Atlantic and the Pacific oceans, that non-tidal ocean loading corrections must be applied.

By comparing minute value of the geoelectric fields Ex and Ey at different times at Siping and Yushu stations with the geomagnetic and strain components of Changchun station, this article analyzes the relations of all physical quantities by further calculating their dominant period with FFT. The results demonstrate that: firstly, daily variation of the geoelectric fields at the two stations present double peaks and one valley, and the dominant period of NS and EW trending is 12 h, 8 h. Compared with geomagnetic field, By and Ex have high suitability of fit and Bx and Ey have suitability of fit. The dominant period 12 h is the common period. Secondly, the geoelectric storm recorded has obvious synchronous change. Thirdly, daily variation amplitude between Ey and Bx has a relative synchronous change, as does that between Ex and By. Moreover, it varies obviously with the season. Finally, some short-impending anomalies may be clearly recorded at two stations, but the form and occurrence time of anomaly before earthquakes differ: the nearer the epicentral distance, the earlier occurrence time and the bigger amplitude of anomaly.

The mainframe of RZB borehole strainmeters has been upgraded, incorporating the core functions of original RZB borehole strainmeters and the merits of other premonition instruments, while retaining its excellent advantageous functions. The upgrades include: capabilities of responding to wide-band signals in RZB multi-component (sampling rate in second level); stabilization and continuities provided and guaranteed by an embedded low-power dissipation, high-stability and high-speed ARM920T kernel, together with customized WinCE; a touch screen on network board facilitating on-site configuration and maintenance; reinforced lightning-proof power supply and charging auto-adapt with DC/AC switching. The software has also been enhanced, with capabilities of remote upgrades and logging (detailed logs are provided for data tracking, exploring and future analysis). Using the new instrument, data analysis finds that the continuous rate has increased substantially, and the minute details of the sample is less than the recorded seismic response process.

We analyzes the Danjiangkou well network 3 wells water level and water temperature observation data from 2013 to 2015. The normal dynamic individual test items are established. We study the different coseismic changes of water level caused by the Nepal M8.1 earthquake on April 25 in 2015 through the Danjiangkou fluid underground well network. The mechanism is explained. Corresponding relationships between magnitude earthquake response and surface wave magnitude and distance are deduced. We analyze the cause of the earthquake coseismic response waveform. The result indicate that well dynamic affect by rainfall water level has certain rules: monthly normal dynamics generally exhibit tidal and pressure effects; furthermore temperature dynamics are smooth, varying in the range of about 0.1 ℃.; different pulse changes of the Tangpa well are caused by different mechanisms of change between water temperature and well water. Different earthquake response characteristics on water level between Tangpa and Leizhuang wells may be associated with their different fault strikes. The water temperature of Shengwan well decreases much less than Tangpa well, caused by different lithology and groundwater type.

In this paper, according to the different ways of release, we perform a detailed analysis of earthquake early warning information release time-delay. The results show that the release time-delay by warning receiving terminal is less than 1 second. It is the same with radio or television release, but efficiency is greatly influenced by time period. Using the internet (new media, mobile phone APP, third-party platform to push) release time-delay is greatly influenced by user concurrency: ten million users of release time delay is 30 to 40 seconds, so we argue that we should use fractal area classification release strategy to effectively improve release time. Analysis of the Wenchuan earthquake for example, shows that within 20 to 50 km, more than 95% of Internet users have 1-7 seconds of warning time; more than 95% of the Internet users who are in the biggest rupture area (Beichuan) will gain more than 30 seconds of warning time.