Research On The Refinement Of Ionospheric VTEC Model Over China

With the continuous development of global navigation satellite system(GNSS)and the increasing demands of users,the accuracy of ionospheric delay correction model is required to be improved.Therefore,it is urgent to study and establish regional ionospheric precision model to improve the accuracy of ionospheric delay corrections.Centered around the regional ionospheric TEC monitoring and the elaboration of the ionospheric VTEC modeling for wide area augmentation system(WAAS),and aiming at the actual problem of complex ionospheric activities across the middle and low latitudes in China,in this dissertation,the Kriging interpolation method in geostatistics based on the properties which describe the randomicity and spatial structure change,but also the spatial correlation and variability of the regional variables,is applied to improve and refine the China regional ionospheric VTEC model for higher accuracy.At the same time,we propose a new Kriging spatial interpolation method with unknown variance components and collocation taking into account of random observational errors in TEC measurements and the signals of ionosphere.According to the models derived in this paper,we design and develop a software constructing the refinement regional ionospheric VTEC model over China based on the Crust Movement Observation Network of China(CMONOC)data,and on the basis of this study,the un-differenced un-combined precise point positioning(PPP)constrained by the China regional ionospheric VTEC refinement model is designed and implemented,and the verification and analysis are carried out at the same time.The main work and content of this dissertation are summarized as follows:1)Comprehensively considering of the regional characteristics of ionosphere vary obviously in different ares and the spatio-temporal variability of the ionospheric TEC,the distribution characteristics and variation features of the ionospheric TEC semivariogram over China are systematically studied and analyzed based on the statistical results of the experimental semivariogram utilizing the CMONOC observation data.Afterwards,on the basis of summarizing the spatial structure change and variability characteristics of ionospheric TEC,a construction method of semivariogram that fully takes into account the temporal variability and can be adjusted flexible according to ionospheric activities is designed,providing effective structure correlation and variability information of spatial variables for accurate ionospheric VTEC modeling over China.2)Present the Kriging spatial interpolation optimization method based on the semi-variogram optimization analysis.The regional ionospheric VTEC Kriging spatialinterpolation optimization algorithm is implemented,which can obtain the best linear unbiased estimation(BLUE)and provide the interpolation estimation accuracy.On this basis,the refinement of China regional ionospheric VTEC Kriging grid model in this paper is established,and a corresponding software is also designed.The results from comparative analysis with a variety of function base models and the global ionospheric TEC map(GIM)products of CODE show that,the Kriging spatial interpolation optimization model fully take into account the spatial structure and variability of TEC variables,which is more consistent with the theoretical and actual situation,with the fitting precision is within 2 TECU,the mean accuracy of gird VTEC estimation is around 1~4TECU,and the RMS of difference with CODE is about 3 TECU,and is more suitable and applicable to China and other regions with complex ionospheric activities.3)Propose a new Kriging spatial interpolation method with unknown variance components estimation,namely KVCE,for both the signals of ionosphere and the noise of TEC measurements.The new proposed model fully takes into account and coordinates the observation noise and random signal model,further improves the regional VTEC model over China.By conducting multi-angle comprehensive comparison and analysis with a variety of models,the results show that,during the experimental period,the local interpolation accuracy of KVCE method is about 1.37 TECU,which is about 1.2 and 0.7TECU higher than the ordinary Kriging spatial interpolation model and polynomial interpolation method,respectively.The regional modeling accuracy is about 1.5 TECU,which is about 1.0 TECU higher than the function-based model;The estimation accuracy of grid VTEC is basically within 6.0 TECU,which is better than the ordinary Kriging method and has more obvious advantages in the marginal region.4)The performance of the China regional ionospheric VTEC refinement model is verified and analyzed from different perspectives,such as the residuals of modeling,the mean square error(MSE)of grid VTEC estimation,and the verification of external rover stations under different solar activity levels.The statistical results of absolute and relative precision mentioned above show that the correction effect under high level of solar activity is still comparable to that in the quiet period of solar activity.On the whole,the correction accuracy of the middle latitude region is better than that of the low latitude region.At the peak and valley of solar activity,the mean epoch residual RMS and MSE are about 2.63,2.60 TECU and 4.52,4.35 TECU,respectively.From the perspective of external index,thedaily residuals of stations are distributed around-2~2 TECU,and the residual of low-latitude stations is about-8 TECU.The RMS at middle and low latitude stations are1~5 TECU and 5~15 TECU,respectively.The average correction on ionospheric delay of each station reaches more than 80%.5)According to single frequency SPP positioning validation and analysis,the 3D accuracy based on CRIM(China Regional Ionospheric Map)is better than 1.4 m and 3.2 m in middle latitudes and low latitudes,respectively,which outperforms CODE by 0.2-1.0 m.6)The un-differenced un-combined PPP(IC-PPP)model taking into account the ionospheric spatio-temporal changes based on the China regional ionospheric VTEC refinement model is realized.Compared with CODE,CRIM can help accelerate the convergence speed of dual-frequency IC-PPP and improve the initial positioning accuracy.In different latitude areas,the IC-PPP convergence time in horizontal and 3D convergence into 10/15/20 cm,improving 1.8/1.3/1.0 min and 3.9/3.0/1.8 min than CODE,respectively.And the IC-PPP constrained by CRIM improves the initial positioning accuracy of 12 cm,13 cm and 6 cm respectively in N,E and U directions.