The characteristics of crustal deformation in the area around Longmenshan fault zone before Lushan M7.0 earthquake and Wenchuan Ms8.0 earthquake were analyzed with the data from GPS stations located in the Longmenshan fault zone.The initial result indicates that：(1）Both of the two seimic events occured the fault segment with low strain rate where was compressed and shortened;（2）Due to adjustment of crustal deformation after Wenchuan Ms8.0 earthquake，the horizontal strain rate in the southern segment of Longmenshan fault zone increased obviously.However，the location where Lushan M7.0 earthquake occured still was a low value area with a strain rate of 30 nanostrain/a，while the strain rate in the surrounding area was more than 60 nanostrain/a.The reason formed this phenomenon could be the area where large earthquake will occur was in critical state of breaking，and the accumulation of plastic deformation reaches limit state.Therefor，the increment of crustal deformation or strain rare in the area would keep low value.

Regional tectonic stress field was analyzed with focal mechanism of Wenchuan Earthquake and the fine rupture model which based on nearfield displacement field，and probability of earthquake occurrence of the faults around Wenchuan was estimated.The result shows that the Coulomb stress in DaofuKangding segment of Xianshuihe Fault increased obviously with the maximum value of 0.67 bar.The stress in east Kunlun fault，Jialichayu fault，southern segment of Minjiang fault and northeast segment of ChengduDeyang fault increased a little.The Earthquake also reduced the stress in Maerkang fault，northern segment of Minjiang fault，southwest segment of ChengduDeyang fault and Huya fault.Analysis of earthquake occurrence probability based on changes of Coulomb stress indicates that the probability of earthquake occurrence in Xianshuihe fault is increased close to 2 times.

Three methods of bathymetry prediction：Space Domain Method，Frequency Domain Method and Least Square Collocation Method were compared by actual bathymetry calculating for three experimental areas.The results show that the accuracy of Space Domain Method is slightly better than the others，the calculation efficiency is much lower and not suitable for bathymetry predicting over a large area，however.Frequency Domain Method can be used to calculate bathymetry model quickly by FFT algorithm，and accuracy is acceptable.Detailed statistics information of the bathymetry is needed for bathymetry inversion with Least Square Collocation Method.This is contradicted with the purpose of bathymetry predicting.The bathymetry models for the experimental areas were calculated with Frequency Domain Method combined with the vertical gravity gradient anomaly and the data from ship soundings.The result indicates that it is possible to predict bathymetry wiyh vertical gravity gradient anomalies.

In order to understand the feature of seismic activity before Wenchuan  earthquake in the source initial rupture area，the regional seismic source location of two larger earthquakes in Dujiangyan prior to Wenchuan earthquake was determined again，and  the focal mechanism solution of  the earthquakes was calculated.The result shows:1）Before M8.0 earthquake，seismic activity in the source initial rupture area increased and formed  a belt.The southeast end of the belt locates Dujiangyan Juyuan Town，then spreads in about 42 angle direction  toword  westnorth direction and ends at Dujiangyan Longxi Town，with a length of 20 km.2）The focal depth of the earthquakes in the southeast region is relatively shallow with a depth of 10 km.，In the time series,，the seismic activity extended from southeast to northwest，passed Zipingpu reservoir area and ended in the initial rupture area of Wenchuang earthquake where  focal depth is about 7 km；deeper than that in the southeast region.3）The sliding angle of focal mechanism solution of the ML3.7 earthquake on February 14 is 93 degree，the NE wall moved vertically downward which is consistant with focal mechanism solution of the earthquakes near the initial rupture point of the Wenchuan earthquake.That means that the initial rupture point of the Wenchuan earthquake had been torn before the earthquake，and the necessary condition of Wenchuang earthquake had been ready.4）The ML3.8 earthquake on February 28 located in the Longmenshan  central faults，and the focal mechanism solution is consistent with  Wenchuan earthquake.It，belongs to the direct foreshock.

A rigorous formula for computing altitude anomaly with global geopotential model was proposed，and accuracy of two geopotential models，EIGEN6c2 and EGM08 were checked using GPS/leveling data in different areas.Results show that the accuracy of altitude anomaly of EIGEN6c2 model is 2~3 cm higher than that of EGM08 model，especially，it is up to 18 cm in Tibetan plateau.Morerover，the system difference between 1985 national height datum and global height datum（W0）is 26 cm determined by the formula.

The evolution of strain energy density in the earthquake gestation was analyzed，and the strain energy density rate in SichuanYunnan region was calculated using the GPS data from CMONOC and XIMALAYA，combined with seismic wave structure.The results show that：in 20012007，the strain energy level in Longmenshan
fault zone was very high，the strain energy in Honghe fault，Nujiang fault，Nanninghe fault and Lancang fault was not accumulated.In 20072009，the strain energy in the study area was influenced largely by the Wenchuan earthquake.In 20092011，the strain energy accumulated were similar to that in 20072009.

According to the relationship between the spherical cap function and the point mass model，a model was proposed based on multilayer point mass model structure coefficient of spherical cap harmonic method.Taking a region (32°N34°N，102°E104°E）as the calculating area，the Gravity anomaly was calculated using the three layers spherical cap harmonic coefficient which constructed by EGM2008 model and the Gravity measured observations.Analysis of fitting precision of gravitational field under different spherical cap harmonic order indicates that the average error between the spherical cap harmonic model and the measured observations is less than 0.5×10^-5 ms^-2，and internal consistency accuracy of the model is 4.56×10^-5 ms^-2 when Spherical cap half angle is 0.71°.

The data from continuous GPS stations and survey mode campaign GPS stations were processed using uptodate geophysical correction models and 364 vertical velocities were obtained.The results show that there are uplift areas and subsidence areas in north China.Almost whole north China plain(NCP）surfers from serious ground subsidence，and the largest subsidence rate reaches 144.0 mm/a，average rate reaches 40.0 mm/a.In addition，Shanxi rift also sink due to withdrawing groundwater，the rate is smaller than that in NCP.In the most areas of Shanxi plateau，ground uplifts lightly，with an average rate of 1.8 mm/a，which is larger than that in Sulu and Yanshan orogenic belt.The results reveal that the present vertical motion pattern of north China is consistent with morphologic pattern and neotectonic movement.

The current crustal deformation of the Tangshan area was analyzed with FEA numerical method under constraint of GPS observation，using physical contact model to obtain the discontinuous motion of faults.The results of simulation show that the movement of the faults in Tangshan area is different from each others.The faults in Tangshan area are right lateral slipping for Tangshan Block is pushed by Shanhaiguan plate.Jiyunhe fault zone has accumulated higher stain，therefore seismic risk there is higher.

Based on the dynamic method，observation error of gravity satellite and its influence on gravity fieldsuch were fully analyzed through simulation experiments.The results show that the contribution of orbital determination precision on gravity field is in loworder，while contribution of intersatellite ranging precision on the gravity field is in middletohigh order.When orbital precision keeps at 2-3 cm level，resolution accuracy of gravity field depends on intersatellite ranging data，that is，the higher precision of intersatellite ranging data is，the higher inverse precision of gravity field is.However，orbital determination precision impacts low order part of gravity field when intersatellite rangingrate reaches nanometer level with invariable orbital error.Therefore，improving orbit determination precision might be one of the key issues in new generation gravity field satellite mission.

There are two distinct tectonic units on both sides of central Tanlu fault zone.The west sides of Tanlu fault belongs to Luxi uplift of Huabei plate，while the east sides of it belongs to Sulu block of Yangzi plate.There are obvious differences between Luxi uplift and Sulu block，which was derived from two blocks’ collision and strong reformation of crust at the later geologic period.The1D velocity models of the two blocks were obtained，using reflectivity methods and simulated calculation of deep seimic sounding wide angle reflection and refraction profiles.By The contrast analysis，show the Moho depth of Luxi uplift is deeper than that of Sulu block by 7 km，while the average velocity of crust in Sulu block is only 5.94 km/s，which is much lower than the average velocity of Luxi uplift（6.34 km/s).This is related to late strong reformation of the lithosphere in Sulu block.The collision of Yangzi and Huabei plate，subduction of the plate，strongly compression，exhumation and magma underplating.lead to this strong reformation of the crust.

Earthquake premonitory signals in crustal deformation recording data always contain various disturbances，the premonitory signals must be distinguished from disturbances so as to be used in shortterm earthquake forecasting.So it is important to trace an abnormal signal and distinguish its sort.Because premonitory signals and disturbances come from different sources，they can be sorted and recognized step by step by comparing records of different instruments at a same site or at different sites.The method is demonstrated by tracing an abnormal signal emerging in crustal deformation data before Ninglang M5.7 earthquake.This signal is sorted as a premonitory signal finally.

GPS data of Crustal Movement Observation Network of China 1999-2013，data of Continental Tectonic Environment Monitoring Network of China and other published GPS results were used to calculate a more complete horizontal velocity field covering the Tianshan mountain.Present crustal shortening rates and its spatial variation were analyzed using 5 GPS profiles across the NorthernPamir Thrust Zone，Kepingtage FoldandThrust Belt，Kuche FoldandThrust Belt，Manas FoldandThrust Belt and Turpan FoldandThrust Belt，respectively.Preliminary results demonstrate that the current shortening rates across the entire Tianshan mountain decrease from west to east with nonlinear features，and the current shortening rates in these profiles are about 26 mm/a，20 mm/a，12 mm/a，9 mm/a and 4 mm/a，respectively.

Focusing on the coordinate time series of 5 IGS reference stations in China mainland（WUHN、BJFS、URUM、KUNM、SHAO） during 2012 under ITRF 2008, coordinate time series was calculated by software TriP, gross error was removed by an improved grossremoving method, and noise as well as periodical analysis had been implemented using different noise model combination.Meanwhile，the experimental results were compared with the results calculated by GAMIT/GLOBK. Conclusion is drawn that noise model of IGS station coordinate time series in China mainland contain a trend and periodical movement, and optimal noise model is different from each others. The optimal noise model of URUM, KUNM, BJFS, SHAO stations is BPPL+WN, the optimal noise model of WUHN is BPPL+WN, FN+WN, PL+WN in N, E, U direction. White noise, which is still the main part during optimal noise models is BPPL+WN, is generally greater than 2 mm. BPPL, which is only about one tenth WN component, is between 0.28 mm to 0.89 mm. Besides, noise component of coordinate time series calculated by GAMIT/GLOBK is less than the results calculated by TriP, stations’ optimal noise models calculated by two different softwares are not exactly the same.

The establishment and data processing of ionospheric model based on the integration of GPS/GNSS systems were introduced in the paper.The result of analysis for influence of adding GLONASS data on global ionospheric model taking the result of 001~031 days in 2010 as an example indicates that combination of GPS and GLONASS data does not improve accuracy，on the contrary，the RMS at grids points is increased due to the influence of weaker quality during GLONASS observing.

The interannual variation，seasonal variation，diurnal variation and spatial distribution characteristics of TEC in China are investigated with time series analysis，spectrum analysis and wavelet decomposition using the global ionospheric grid data provided by IGS from 1999 to 2013.The main features are as the followings：TEC varies greatly in the daytime while slightly in the night with the spatial distribution；overall，the TEC value in the low latitude area is higher than that in the mid latitude.In the winter，TEC value in the low latitude is lower than that in the midlatitude.Within a day，peak value appears during  LT 12:0016:00 and valley appears during  LT4:006:00.In the periods，the peak time moves forward with the increase of latitude.This rule may provide certain reference to modify K8 model which set peak time as fixed.The spring is quiet period,in which TEC reaches its maximum value of a year，and the autumn is active period.TEC has obvious semiannual period of about 183 days.In addition，the analysis of  long time sequence shows that winter anomaly does not exist in China，except in 2001.

In the classical ionosphere vertical density profiling，the influence of geomagnetic field is often ignored.To improve the quality of inversion results，the geomagnetic fields should definitely be taken into account when using the bending angle to retrieve the vertical density profile with GNSS ionosphere radio occultation.Therefore，an improved approach for retrieving vertical electron density profile is proposed，which considered the geomagnetic field from International Geomagnetic Reference Fields（IGRF11）in computing TEC by excess phase for each ray path，where the TEC is changed with tangent height and its change ratio can be used to calculate the bending angles.The result of a practice experiment with COSMIC observation data shows that the approach is feasible.

In the paper，the author summerized the progress of the research on models and algorithms of ionospheric CT ，and discuaaed the future development trend and its application prospect of ionospheric CT technique.

〗In the traditional precise point positioning（PPP)，the integer ambiguity can only be estimated as float value，which restricts the further enhance of PPP accuracy for the existence of fractional cycle biases（FCB）in the carrier phase observations from the receiver and satellite ends.In the study，the ionospherefree ambiguity is decomposed into widelane（WL）and narrowlane（NL）ambiguities.The betweensatellitedifference（BSD）operation can remove the FCB included in WL and NL ambiguities at the receiver end.The satellitebased FCB contained in the WL ambiguity provided by the international GNSS service（IGS）analysis center CNES can be used to recover the integer nature of the WL ambiguity.Due to the precise satellite clock corrections provided by CNES assimilate the satellitebased FCB of the NL ambiguity，the integer nature of NL ambiguity can also be recovered after applying the precise clock corrections.The analysis results based on the IGS observation data indicate that the positioning accuracy of BSD PPP fixed resolution is significantly better than the zerodifference（ZD）PPP float resolution.The improvement rate of 3dimensional positioning accuracy reaches 63.8%.

Continuous GPS coordinate series contains modulated amplitude seasonal signals，and using the leastsquares fitting the traditional model will introduce model error.According to the timevarying characteristics，a method to extract the modulated seasonal signals using the varyingcoefficient regression model was proposed.Two groups of data simulated and the real series of GPS coordinate were analyzed with the method，the results show that the varyingcoefficient regression model in extracting the seasonal signals of GPS coordinate series is more effective than traditional model，and the  GPS time series processed by the method can get a more reasonable estimation of speed and noise.

On the bisis of the three carrier ambiguity resolution（TCAR）algorithm，the two improved methods，ionospheric delay correction method and the method of geometryfree and ionospherefree combination，were compared and analyzed with the Beidou/Compass triplefrequency surveying data.The experiment results show that both methods can improve the accuracy of longbaseline ambiguity resolution，and a rapid longbaseline ambiguity resolution after data smoothing with some epochs can be realized.The two methods can be useful for achieving the longbaseline precise rapid navigation and position with Beidou/Compass triplefrequency.

The data of the Shaanxi continuous GPS reference stations from Jan 1st ,2011 to May 30th,2013 are computed and analyzed in this article. After common mode error being removed by the regional stacking, the mean RMS of coordinate time series are declined 69%,60% and 41% respectively in N,E,U directions. The spectral indexes analysis states that the afterfiltered coordinate time series contain the white noise and the color noise . The CATS software analysis indicates that the noise components of vertical direction are much larger than those of horizontal direction,and random noise appear in some stations along with flicker noise being decreased by the spatial filtering.

Considering the error of adjustment problem for independent variable in linear regression，an iteration algorithm of total least squares was derived according to indirect adjustment model and total least squares theory.The algorithm is simple and easy to programming.The result indicates that the algorithms is more effective and more feasible than other algorithms.

In GNSS control network adjustment，it is neccesary to convert GNSS baseline vector error between different coordinate system.A formula was derived to convert GNSS baseline vector error from space rectangular coordinate system to geodetic coordinates，and from geodetic coordinate system to the Gauss plane rectangular coordinate system.Due to unit of the parameters of GNSS baseline vector error in the space rectangular coordinate system is different fron the geodetic coordinate system，the angle errors in geodetic coordinate system is converted to length error with curvature radius of meridian and radius of parallel circle.On this basis，the conversion of GNSS baseline vector error from the space rectangular coordinate system to station rectangular coordinate system was carried.The test result shows that the formula is correct，and easy to form.

In order to solve the problem in large scale matrix inversion with least square solving，the parallel algorithm to inverse the highorder dense symmetrical positive define matrix with GaussJordan method on the basis of MPI was proposed.Considering the high memory demand on single processor for this algorithm，the MPI I/O interface was introduced，which can reduce communicational time between each node simultaneously.Introducing the parallel I/O efficiency，memory demand on single processor，acceleration ratio and relative efficiency，the efficiency of the parallel algorithm can be estimated comprehensively.The simulation result indicates that this parallel algorithm can improve the efficiency significantly，ant its relative efficiency can reach to 60%.Eventually，the earth gravity field complete to degree and order 120 and 240 are recovered on the Sugon Cluster，and the relative inversion time is only 229 s  and 7 395 s respectively.In addition，the memory demand on the single node is only 205 MB and 1.57 GB，while its inversion accuracy can reach to 10^-18，which indicates that the parallel algorithm in this study can be used to obtain optimal value in least square efficiently and stably.

A model to estimate shortterm prediction of satellite clock bias was established based on quadratic polynomial and twosteps minimization.The coefficient of covariance function was fitted with residuals of quadratic polynomial model.Comparison between result with the model and usual method indicates that the accuracy of the collocation model in shorttime prediction of satellite clock bias  improved.It can be also served as a new interpolation method by comparing with common algorithms of clock bias prediction and interpolation.

The premise of establishing GaussMarkov(GM）model based on least squares(LS）to fit GPS height anomaly is that plane coordinate of control points is errorfree，only elevation anomaly value contains observation error.A model applying robust estimation based on iteration method with variable weights was proposed in the paper.The result calculated with an example to prove the accuracy is improved with the model.

The principle and algorithm of gravity gradient calculated by EGM and DTM data are analyzed，and The numerical  of gravity gradient values are calculated by EGM2008 and SRTM3.Comparison of the correlation between DTM data and gravity gradient component  verifies the reliability of gravity gradient data calculation by DTM data.The principle and methods of gravity gradient aided inertial navigation are discussed，and its necessity and practicability are also verified by comparing the position errors generated by accelerometer bias and gravity anomaly changes separately of a simulated trajectory results.

The precipitable water vapor（PWV）reflects the total amount of water vapor of the atmosphere along the zenith direction，it is closely related to meteorological factors.The change of PWV shows a kind of regularity in the process of snowfall weather.Based on the consideration，a conversion model of PWV and ground water vapor pressure was established to analyze characteristics of PWV in Arctic Yellow River station of China with the data of GPS remote sensing PWV，meteorological data，sounding data and snowfall data.The results show that the PWV from GPS technique can be used to precast snowstorm process.

The deformation sinals of pylons were extracted and analyzed with Singular Spectrum Analysis（SSA）combined with wavelet analysis method.The optimal window length was selected according to the first and second principal component reconstructed with the original series minimum mean square error，combines with experience criteria.A method was proposed based on using peaks of curvature spectrum of singular value to determine the effective principal component and overcome the previously established subjectivity.Periodic component extracted combined with wavelet analysis shows that deformation of the pylons has the characteristics of multiperiod.The oscillation frequency of north and south pylon in X，Y direction is consistent in the main period time scales，and there is no discord oscillation.

The accurary of geomagnetic matching navigation can be improved a lot in a suitable geomagnetic matching area.Geomagnetic matching area characteristic parameters were analyzed based on statistical association and geomagnetic information entropy，and traditional geomagnetic information entropy was improved，combined with geomagnetic coarseness and accumulated gradient characteristics.Particle Swarm Optimization（PSO)algorithm was employed to obtain the suitable areas.The results show that the geomagnetic matching area improved characteristic parameters reduce runtime largely，and more suitable areas can be selected，which illustrate the efficiency of improved geomagnetic information entropy parameters.

A method of empirical mode decomposition was used to processe precipitable water vapor time series during 2000-2004 from GPS data of Crustal Movement Observation Network of China（CMONOC）and change trend of precipitable water vapor at every GPS station was obtained.The result of classification and comparison for the change trend indicates that the change trend in tropical monsoon climate zone，monsoon climate of subtropical zone，temperate continental climate zone and alpine climate zone is first up and then down；the change trend in temperate monsoon climate zone is first down and then up.

A method was presented，which is used to represent slope displacement process in graphs. The displacement of a single point can be represented in three axes component，Δx，Δy，and Δh，the horizontal displacement component (Δx，Δy)，and threedimensional space point
(Δx，Δy，Δh)，based on the slope displacement data.The different types of graphs can display the slope whole and local deformation characteristics of the slope as a whole or a local part.

A method of enhanced gridding with gradient of the minimum curvature gridding was proposed in the paper，base on analysis of features of several common grid methods and the background of enhancing minimum curvature gridding，and the magnetic model was presented to validate the ability of enhancing the grid.The experiment result with the method  proves the method is effective.

Based on the Robust Mahalanobis Distance，an algorithm for detecting and removing gross error in LiDAR point cloud is proposed.First，LiDAR point cloud is divided into several blocks.Then，detMCD (deterministic minimum covariance determinant) algorithm is performed to get robust estimation of the position and scale parameters of each block.With these parameters，Robust Mahalanobis Distance of each point within the block is calculated.As the square of Robust Mahalanobis Distance could be considered to obey Chisquare distribution，the gross error discrimination threshold could be obtained with certain confidence.Finally，gross errors might be detected and removed with the threshold and robust mahalanobis distance in each block.The proposed algorithm was tested in LiDAR data acquired from a typical gully in Qiaozi Valley，Tianshui City of Gansu Province.The algorithm was performed with various combinations of parameters，including the average point number per block in point cloud dividing step and the relative size of subset reserved in detMCD step.The results show that more points were removed as gross errors with increase of the average point number per block，while a larger relative size of subset led to less points detected as gross errors.The comparison of the TIN and profiles generated by LiDAR point cloud before and after the gross error detected and removed shows that the proposed algorithm can remove the gross error in LiDAR point cloud efficiently and achieve higher accuracy and robustness.

A RSSI algorithm without a prior knowledge of channel model parameters for node locating in Wireless Sensor Network was proposed in the paper.It can be realized using the algorithm realtime location estimation by coping with the channel model and TSVD to solve the equation.Extensive simulation results show that locating accuracy using the algorithm is higher than some of other RSSI algorithms，and the algorithm is especially suitable to the cases of absence or incorrectness of the channel model parameters.

For improving precision，a removingrecovering method is used to remove the long wave information of earth gravity field model to get more smooth residual gravity height anomaly，and compared the influence of different covariance function on the fitting results of least square collocation which take measurement error and model error into account.The results show that Gaussian recovering and resemble Gaussian function can achieve higher fitting precision for a large area with height anomaly value changes significance；the removerestore method can effectively improve the fitting precision to least square collocation method which depend on the covariance value of each points.