An algorithm combined with EOFs analysis and MART for GPS-based ionospheric tomography

Ionospheric imaging usually involves solving an underdetermined inversion problem. The inversion involving additional constraints is performed to enforce realistic profiles in the vertical. A new algorithm combined with empirical orthonormal functions (EOFs) analysis and multiplicative algebraic reconstruction technique (MART) for GPS-based ionospheric tomography has been given in this paper. EOFs can be used to reconstruct the background ionospheric electron density effectively with lower distortion, while the MART can lead to more accuracy by iteration. Numerical simulations and its application on actual GPS-based TEC observations proved that this algorithm offers better performance in physical study of the anomaly, particularly under storm events when its morphology deviates from normal.