Derivation of the ionospheric structure over extended geographical regions from azimuth and elevation scan backscatter ionograms: A modular neural network approach

This paper describes a new real‐time inversion method based upon multifrequency sounding schemes and variable azimuth and elevation angle backscatter ionograms for the derivation of the three‐dimensional structure of the ionosphere. The inversion technique which has been developed for this purpose employs neural network models in a modular structure and is fully explained. Tests on simulated data demonstrate the potential accuracy of the method for reconstructing the electron density profile over remote geographical regions.