Research On Ionospheric Space-Based CT Based On Empirical Orthogonal Functions

The ionosphere is an important component of the sun and earth system.It is closely related to the lives of human beings.It has the effect of scattering and refraction of radio waves passing through the ionosphere.It is an important area of space physics to actively explore the ionosphere and its physical phenomena question.Computerized Ionospheric Tomography(CIT)is a new technique for radio telemetry of space environment.It is widely used to detect the spatial distribution of electron density at different scales.Studies on CIT have been underway for 30 years,but due to the complexity of the ionosphere,many problems still remain to be solved regarding CIT.The goal of this paper is to use the GRACE-GPS beacon data to reconstruct the electron density distribution from the top ionosphere to the plasma layer(450-5000 km).In this paper,TEC data from GRACE-GPS receivers and the source of error are firstly introduced.A study is conducted on the cycle-slip processing of GPS data,and a method for step-by-step processing of cycle-slip is proposed.Then,the Empirical Orthogonal Function(EOF)CT reconstruct algorithm was added to the space-based CT and compared with the traditional pixel-based CT algorithm.The results of two CT reconstructions were compared by simulation inversion and real inversion.Using the EOF reconstruct algorithm to reconstruct the electron density distribution during the geomagnetic storm from September,2003.The main study results are showed as follow:1.Using the GRACE-GPS observation data to simulate the cycle-slip,analysis these methods' advantages and disadvantagesThe results show that the phase-reduction method can detect large cycle-slip of more than 6 cycles.ionosphere remnant method of difference method and wide-lane combination method can detect small cycle-slip of more than 1 cycle.2.A combined step-by-step method for dealing with cycle-slip is concluded.The combined method was analyzed experimentally.The simulation experiment showed that the combined method can completely eliminate cycle-slip of more than 1 cycle.GPS phase data processed by this method is "clean" and reliable for CT inversion.3.Voxel CT inversion results and EOF inversion were compared by simulation inversion and real inversion.The simulation inversion results show that the addition of empirical orthogonal function to space-based CT inversion can effectively reduce the deviation rate of reconstructed electron density above 1000 km in the vertical direction.In addition,the use of empirical orthogonal function can greatly reduce the number of unknown equations,and the inversion convergence time is greatly shortened.The comparison between the real inversion results and the measured data shows that the EOF inversion can accurately reconstruct the ionospheric large-scale structures such as Equatorial Ionization Anomaly(EIA)and mid-latitude ionospheric trough(MIT).The statistical results of true inversion show that the EOF-based CT inversion and traditional pixel inversion results are significantly correlated with the DMSP satellites' actual measurement results.The correlation coefficients are 0.86 and 0.84,respectively.The new CT algorithm based on EOF is an effective method to observe the electron density distribution of the ionosphere over 450 km worldwide.4.The Equatorial Ionization Anomaly(EIA)and the mid-latitude trough structures were reconstructed and compared with the measured electron density data of the DMSP satellite and the CHAMP satellite.The results show that the EOF inversion used in this paper is very consistent with the measured data.In addition,the ionospheric electron density maps during the period of geomagnetic storms from September 15 to 21,2003 were reconstructed to analyze the response of the MIA during the storm.The results show that the MIA moved significantly towards the equator during the storm;After the storm,when the geomagnetic disturbance is small,the MIA moves to higher latitudes.