Dual Empirical Orthogonal Functions Restrained Tomographic Model for Ionosphere Imaging

Extra constraints are always brought into an ionosphere tomographic model to strengthen its robustness. However, existent models using the constraints either show a weak restriction, amplifying the model’s uncertainties, or demonstrate a strong restriction, decreasing the model’s flexibility. Empirical orthogonal function (EOF), as a kind of basis function and generally being extracted from a prior dataset, is an ideal constraint for the tomographic model. By using EOFs as the vertical and slant constraints of the tomographic model, a dual-EOF restrained tomographic (DERT) model is proposed. In DERT, a smooth constraint, conducted by a moving window process, is first imposed, and then, a vertical constraint and a slant constraint, which are carried out by fitting the estimated vertical and slant profiles with their dominant EOFs via a least-squares method, are appended. To reduce the impacts of the errors of the slant total electron content (STEC) measurements on the processes of moving window and least-squares fit, an algorithm that can reject abnormal STEC values is employed before the two processes. Validations tests are conducted to validate the DERT and compare it with the Slepian method and the International Reference Ionosphere 2016 (IRI-2016) model. Results show that DERT outperforms the reference methods in aspects of vertical profiles, F2 layer peak density (foF2), STEC measurement, and vertical total electron content (VTEC). The root mean square error (RMSE) improvements over the Slepian method are 14.17%, 0.47%, 16.24%, and 40.46% (IGS map as the truth), respectively.