Using Low/High Orbit Satellite Based GNSS Observations For Derivation Of Thermosphere Mass Density And Electron Content

The modern GNSS systems has been successfully applied for users from ground to low earth orbit(LEO)as well as to high earth orbit(HEO)due its rapid development in recent decades;and thus,they have been playing an important role in LEO/HEO orbit determination as well as atmosphere derivation.For LEO/HEO onboard GNSS observations,they are not only affected by the atmosphere density through dynamic models,but also by the ionosphere refraction through observation models.However,this in another inspective gives opportunities in derivation of atmosphere properties using the LEO/HEO onboard observations,especially for the thermosphere mass density,ionosphere total electron content(TEC)as well electron density profiles(EDP).Based on the investigations about the current research status on these fields,we focused our study on LEO/HEO orbit determination(OD)and atmosphere derivation with onboard GNSS technique.The main contents and the results of this study is given as below:(1)LEO/HEO orbit determination with onboard GNSS by refiningobservation models and ambiguity resolutionTo improve orbit quality with onboard GNSS data,the antenna offset and variation errors are analyzed and then modeled for corrections while the single-differenced ambiguity resolution(AR)technique is also investigated.Compared to float solutions,the fixed GRACE-FO dynamic orbits show improvement about 28%.Correcting the aprior antenna errors can significantly improve residual statistics as well as orbit precision.For GRACE-FO the RMS(root mean squares)value of phase residuals is reduced to 7 mm level;the orbit precision by comparison to JPL products is 16 mm while by SLR validation is 10 mm.We analyzed the code errors of onboard BDS data and investigated its OD performance.The FY3C/FY3 D onboard observations both demonstrated onboard BDS data suffers from severe systematic code errors due to BDS satellites.The precision of BDS-based LEO OD is affected by the BDS product precision.When discarding the BDS GEO observations which are usually degraded by it poor orbit and clock products,the BDS only OD can reach 10 cm for LEO users.We compared WUM and GBM products for BDS OD,which suggested that the BDS GEO products from WUM is superior to GBM due to the upgraded GEO solar radiation model,and caution should be payed when using the GBM products since the clock products show significant discontinuities.We also analyzed HEO satellite OD using its onboard GNSS.The results from domestic GEO satellite TJS-2 showed an orbit precision at 2 m.(2)Thermosphere mass density derivation using LEO satellite onboard dataWe investigated the methods and applications of thermosphere density retrieve using LEO TLE(two line elements)data as well as onboard accelerometer data.The CHAMP,GRACE accelerometer was adopted for density retrieve,in which we calibrated the accelerometer data by adjusting the estimation intervals of calibration parameters instead of using empirical accelerations,thus the completence of calibration parameters was preserved.The derived densities along GRACE orbits was compared to results from other institutions,which show very good agreement with a bias of-0.2% and a deviation of 5.4%.The TLE-derived densities from CHAMP and GRACE satellites showed differences at 6% and 17%,respectively,when compared to the accelerometer-derived densities;however,the differences increased to 25% to 27% when compared to NRLMSISE00 model.This indicated the TLE-derived densities could be used for calibrating empirical density models.We developed different calibration schemes for NRLMSISE00 and Jacchia model with TLE-derived densities using 36 LEO objects,and results indicated an average improvement about 5-15%.(3)DCB and topside ionosphere TEC estimation using LEO onboard GNSSdataWe proposed a method for differential code bias(DCB)and topside ionosphere TEC(p TEC)estimation using LEO onboard GPS and BDS data and investigated the effects of systematic code errors on estimation precision.Using FY3 C onboard GPS/BDS data,we demonstrated that the DCB stability of GPS,BDS GEO,IGSO and MEO satellites can be improved by 5.0%,3.1%,16.2 and 13.6%,respectively,after correcting the systematic code biases.The stability of BDS MEO DCB using onboard data can even reach at 0.1 ns level,which is superior to results obtained from ground data.The p TEC estimates retrieved from FY3 C onboard data showed reasonable variations and its responses to geomagnetic storms gave very similar patterns compared to other studies.With the DCB and p TEC estimates,we also evaluated the high-order ionosphere delays for LEO GNSS observations.(4)GEO-based GNSS radio occultation analysis and EDP derivationWe have firstly analyzed the GEO-based GNSS radio occultation(RO)data and developed a method of estimating EDPs from it.We proposed a moving average filter to smooth the high-frequency errors in the TJS-2 single-frequency excess phase,which showed higher precision than the single-difference technique.With such method,we derived EDPs from TJS-2 data,and validated the precision by comparison with the ground-based digisonde,the IRI 2016 model and the COSMIC results.The TJS-2 ionospheric EDPs show good agreement with correlation coefficients all exceeding 0.8.The TJS-2 average Nm F2 differences compared to digisondes and COSMIC results are 12.9% and 1.4%,respectively,while the hm F2 differences are 1.65 km and 1.76 km,respectively.We also successfully derived EDP estimates to altitudes up to several thousand kilometers by using the main-lobe as well as side-lobe GEO-based GNSS RO signals.We also analyzed the temporal and spatial distribution patterns of GEO-based occultation events,and found these events occur at specific locations with daily repeatability.