Study On Empirical Model Of Ionospheric TEC Based On Multi-source Data

The ionosphere is an important part of the near-Earth space environment and can have a significant impact on space activities such as electromagnetic wave communications,satellite navigation and positioning,and aerospace.The construction of empirical ionospheric models to describe and predict Total Electron Content(TEC)can not only provide model support for in-depth study of the physical and dynamical mechanisms of the ionosphere,but also provide ionospheric delay correction services for communication,navigation,remote sensing and telemetry.However,the ocean and regions with ionospheric anomalies seriously affect the predictive capability of empirical ionospheric TEC models.This paper explores the fusion method of multi-source TEC data with encrypted oceanic regional data sets and ionospheric anomaly expressions,and presents the global TEC empirical model algorithm and functional expressions for subcomponent models,focusing on the Mid-latitude Summer Nighttime Anomaly(MSNA),Equatorial Ionization Anomaly(EIA),and the "hysteresis effect" in solar activity,and established the global empirical ionospheric TEC model(TEC Model of Multi-source Fusion,TECM-MF).The model can provide a new ionospheric delay correction method for GNSS single-frequency users,and is also an important reference for building other new and improving existing global ionospheric empirical models.The main research contents are as follows.(i)The variations and accuracy of the CODE GIM and IRI2016 models on a global scale are analyzed.For CODE GIM,the spatial and temporal variation characteristics of the global intra-ionospheric compliance accuracy are explored by applying CODE GIM data from January 1999 to June 2020.The results show that the TEC RMS is highly correlated with solar activity intensity;the root mean square of the difference between CODE TEC and single-station GNSS-TEC is less than 3 TECU;the TEC RMS is unevenly distributed globally and there is a "double hump" anomaly;the accuracy of land region is higher than that of ocean region,and the accuracy of northern hemisphere is higher.For the IRI2016 model,a statistical analysis of the model’s prediction accuracy on a global scale was completed.The results show that: the model accuracy is higher in the dense area of GNSS tracking stations;the solar activity intensity has a serious impact on the TEC accuracy of the single-station model,and the model accuracy is higher in the low solar activity year;the IRI2016 model has a stronger ability to predict TEC in summer;the accuracy of the land area model is higher,while the accuracy of the ocean area model is lower.(ii)The multi-source TEC data fusion method is studied.Applying the 15 th order 15 th spherical harmonic function model,solving the spherical harmonic function coefficients by the method of normal equation superposition,and performing the precise weighting of different TEC data by the method of Hermetic variance component estimation.The CODE global TEC grid data,Jason altimetric satellite TEC data,COSMIC occultation TEC data and IRI2016 model TEC data are fused from 01 January 2006 to 30 June 2020 in the oceanic region,which improve the accuracy of modeling datasets in the ocean region.(iii)The refined functional models of the diurnal,seasonal and geomagnetic components of the ionospheric TEC have been studied,and the MSNA corrected component,EIA corrected component and solar activity component(including the "hysteresis effect")have been studied based on the basic characteristics of MSNA,EIA and "hysteresis effect".The functional expressions of MSNA corrected component,EIA corrected component and solar activity component(including "hysteresis effect")are investigated.Based on the multi-source fused TEC data set,a new empirical model of global ionospheric TEC,TECM-MF,is developed by applying the nonlinear least squares method to fit the coefficients to be determined.The TECM-MF model was analyzed for residuals,internal compliance accuracy evaluation,external compliance accuracy evaluation,and impact on positioning accuracy.The results show that: the standard deviation and root mean square of the model residuals are 3.9 TECU,TECM-MF can fit the fused data set well and can show the "hysteresis effect" phenomenon;the model predicts the MSNA ionospheric anomalies better than the NTCM-GL model and Nequick2model;In the single-frequency standard single-point positioning,the model can correct the ionospheric error better than GPS Klob,and the positioning residuals in 3D direction are consistent with CODE GIM.