A New Method of Ionospheric Modeling Based on Semi‐Parametric Kernel Estimation

Objectives At present, in the modeling of regional ionospheric delay, the assumption error of thin layer height, the selection error of projection function, the estimation error of differential code bias and the modeling error of mathematical formula are seldom considered. In order to reduce the influence of these errors on the calculation accuracy, a semi-parametric model is introduced into the modeling of spherical harmonic function in the ionosphere, the results are substituted into total electron content (TEC) of the ionosphere in the modeling region by the fourth-order ionospheric harmonics function.

Methods The modeling accuracy of 24 international global navigation satellite system service (IGS) stations in the European continent and 48 stations in China is compared and analyzed by using the semi-parametric kernel estimation method and the traditional least square method.

Results and Conclusions Experimental results show that compared with the least square method, the accuracy of regional ionospheric modeling based on semi-parametric kernel estimation method is 12.2%-19.0% higher than that based on the ionospheric TEC published by Center for Orbit Determination in Europe (CODE) and 8.3%-13.6% higher than that based on the ionospheric TEC published by IGS in low solar activity years. Compared with the least square method, the semi-parametric kernel estimation method has the same accuracy in ionospheric modeling due to high solar activity.