Multi-GNSS Global Ionosphere Monitoring And The Unified Handling Of The Differential Code Bias

The Global Ionospheric Maps(GIM)are the most direct products for global users to obtain the ionospheric Total Electron Content(TEC).The accuracy of the GIM depends both on the ionospheric observable and the mathematical model to represent the temporal and spatial distribution of the ionosphere.On the one hand,Multi-GNSS(Global Navigation Satellite System),including the Global Positioning System(GPS)from America,GLObal NAvigation Satellite System(GLONASS)from Russia,Galileo from Europe and BeiDou navigation System(BDS)from China,means more satellites and signal types,as well as different constellation configurations,which brings both opportunities and challenges for the ionospheric observable retrieval and ionosphere modelling.On the other hand,on behalf of the more and more mature high-precision GNSS geodesy technique,the Precise Point Positioning(PPP)technique provides another useful method to monitor the ionosphere precisely.The Differential Code Bias(DCB),which is included in the ionospheric observable,is estimated and coupled along with the ionosphere mathematical model.Taking the DCB apart from the ionospheric observable,the remaining part is the ionosphere TEC.Therefore,the DCB,which is related to both the ionospheric observable and ionosphere mathematical model,is one of the main error resources for the ionospheric observable retrieval and ionosphere modelling.Aimed at the ionospheric observable retrieval and global ionosphere modelling,as well as the generalized handling of the DCB,the following parts are investigated systematically in the thesis:(1)Multi-GNSS Global Ionospheric Maps,namely SDU in the format of IONosphere EXchange format(IONEX),based on the carrier phase leveled pseudorange ionospheric observable and the spherical harmonic model.As the GIM products provided by the International GNSS Service(IGS)Ionosphere Analysis Centers(ACs)are mainly based on the carrier phase leveled pseudorange ionospheric observable and the spherical harmonic model,the Multi-GNSS GIM,using carrier phase leveled pseudorange ionospheric observable from GPS,as well as GLONASS,Galileo and BDS,are produced in ShanDong University(SDU)and compared with the other GIMs provided by the ionosphere ACs.Taking the CODE(Center of Orbit Determination in Europe)GIM as reference from July 1st to 31st,2018,we evaluate the GIMs from SDU,JPL(Jet Pulse Laboratory),ESA(Europe Space Agency)and UPC(Technical University of Catalonia).The results demonstrate that the mean bias and Root Mean Squares(RMS)of GIM products from EAS/JPL/UPC/SDU with regard to that of CODE is 0.07/-1.99/-0.84/-0.13 TEC Unit(TECU)and 1.1 5/2.1 9/1.45/1.64 TECU.(2)Accuracy evaluation of the carrier phase leveled pseudorange and PPP ionospheric observable,as well as the single-layer and two-layer GIMs.Firstly,the accuracy of the carrier phase leveled pseudorange,or smoothed geometry-free combination observable,"SP4" for short,and the undifferenced ambiguity-fixed ionospheric observable,"FL4" for short are evaluated from Day of Year(DOY)150 to 180,2014.The results present that the accuracy of the FL4 observable is more than one order of magnitude better than that of the SP4 observable.Specifically,the accuracy of the FL4 observable ranges from 0.05 to 0.11 TECU while that of the SP4 observable is 0.65 to 1.65TECU.Secondly,the contribution of the varying Between-Receiver DCB(BR-DCB)to the Single-Difference ionospheric observable is quantified.Besides,it is induced that the DCB variation is the main error resources of the ionospheric observable.Thirdly,the GIMs based on different ionospheric observable and different mathematical models,including the EHRG based on SP4 and the spherical harmonic model,the GAG1 based on the FL4 and the spherical harmonic model,as well as the FPPP based on the FL4 and the two-layer model,under different solar activities are evaluated.The results demonstrate that?the accuracy of the FPPP model is always the best whenever during the low solar activity or the high solar activity(the FPPP RMS is 0.73TECU when solar activity is low and 2.12TECU when solar activity is high),followed by the GAG 1 and EHRG model;?the comparison between the GAG1 and EHRG proves that the accuracy of the FL4 observable is higher than that of the SP4 observable;?Though both the GAG1 and FPPP model are based on the FL4 observable,the higher accuracy of the two-layer based FPPP model shows the priority of the two-layer model over the single-layer model.Furthermore,the Standard Point Positioning(SPP)experiments using the EHRG,GAG 1 and FPPP model under different solar activities proves that the positioning accuracy using the FPPP model is the best,followed by the GAG1 and EHRG.Specifically,compared to the EHRG,the improvement of the positioning accuracy using FPPP under low and high solar activity period can reach to 32%and 38%.Therefore,the above conclusions indicate that the accuracy of the GIMs are depend on both the ionospheric observable and the mathematical model.The high accuracy ionospheic observable should match the high accuracy ionosphere model to show the best accuracy of the GIM.Specifically,the accuracy of the spherical harmonic model is comparable to that of the SP4 observable,however,cannot reflect the accuracy the FL4 observable.Compared with the spherical harmonic model,the two-layer model is more suitable to the FL4 observable.(3)The estimation of the GPS/GLONASS intra-frequency bias and inter-frequency bias.The DCB,including the intra-frequency bias and inter-frequency bias,is the largest error resources to retrieve the ionosphere TEC.The accuracy and stability of the GPS/GLONASS satellites are analyzed using the IGS stations.Taking the monthly mean DCB products from CODE in July 2018 as reference,the monthly stability of GPS P1C1,P2C2 and P1P2 estimated by SDU is 0.18,0.08,0.03 ns while the RMS of GPS P1C1,P2C2 and P1P2 estimated by SDU is 0.09,0.13 and 0.04ns.As for GLONASS,the monthly stability of P1C1,P2C2 and P1P2 is 0.18,0.08 and 0.09ns while the RMS is 0.29,0.27 and 0.19ns.(4)The analysis of the short-term stability of the GPS DCB,as well as the impacts of the ionosphere mathematical model and temperature on the DCB variation.Firstly,the hourly GPS DCBs from DOY 150 to 180,2014 are estimated by the Fast PPP technique.The results indicate that the mean daily stability(Standard deviation,STD)of the GPS satellite is about 0.20ns.And all the GPS satellites show the same level stability.The monthly stability,however,affected by the day boundary discontinuity effects,is from 0.60 to 0.75 ns,with a mean value of 0.70ns.Secondly,in the aspect of the coupling of receiver DCB variation and mathematical model error,the presence of the seasonal and latitudinal dependence of both the DCB variation and mathematical model error proves the coupling effects between them.The seasonal effects,to a great extent,are attributed to the solar activity.That means the diurnal and annual variation of the solar activity lead to the periodical variation of the ionosphere mathematical model error.Because of the diurnal effects of the solar activity,the hourly receiver DCB is more affected than the daily DCB.The results in 2014 present that the correlation between hourly DCB and the solar activity index Global Electron Content(GEC)can reach to as much as 83.4%.Thirdly,in the aspect of the temperature dependence of the receiver DCB,the correlation of the Between-Receiver DCB(BR-DCB)and temperature varies at the different receiver configuration,that different receiver and antenna types.In our experiments,the highest correlation can reach to 84%and the correlation improves when the temperature rises.It is noted that due to the fact that the BR-DCB may counteract the trend term of the undifferenced DCB variation,the correlation results in our experiment can be regarded as the lowest limit relationship.The correlation of the undifferenced DCB and temperature should be higher than that present in the thesis.Besides,the DCB variation with regard to the ionosphere mathematical model error,should be regarded as the pseudo-variation,which can not reflect the actual change of hardware based DCB.(5)The estimation of the Multi-GNSS DCB,including GPS,GLONASS,Galileo and BDS,as well as the concept and processing of the differential code bias to the pseudo-absolute observable bias.Based on the Multi-GNSS EXperiment(MGEX)network,the accuracy and stability of the GPS/GLONASS/Galileo/BDS DCB are analyzed.Taking the daily Multi-GNSS DCB products from Chinese Academy of Science(CAS)in July,20 1 8 as reference,for GPS,the monthly stability of C1CC2W,C1CC5X and C1CC5Q are 0.07,0.12 and 0.09ns while the RMS are 0.09,0.07 and 0.08ns;for GLONASS,the monthly stability of C1CC2P and C1CC2C are both 0.09ns while the RMS are 0.14 and 0.20ns;for BDS,the monthly stability of C2IC7I and C2IC6I are both 0.10ns while the RMS are 0.17 and 0.29ns;for Galileo,the monthly stability of C1XC5X,C1XC7X,C1XC8X,C1CC5Q,C1CC7Q and C1CC8Q are 0.07,0.07,0.26,0.09,0.09 and 0.09ns while the RMS are 0.07ns,0.08ns,0.28,0.08,0.09 and 0.08ns.Furthermore,the necessity of the generalized handling of the Multi-GNSS DCB is clarified by the standard point positioning example in the form of ionosphere-free combination using GPS C1X and C5Q observable.The principle of the generalized handling is then introduced,i.e.,through the satellite clock estimation datum,the differential code bias can be parameterized to the pseudo-absolute observable code bias.Afterwards,both the processing flowchart and results of the pseudo-absolute observable code bias are presented.Finally,by the same experiment using GPS C1X and C5Q observable,the availability and handleability of the pseudo-absolute observable code bias is demonstrated.