Ionospheric response under the influence of the solar eclipse occurred on 20 March 2015: Importance of autoscaled data and their assimilation for obtaining a reliable modeling of the ionosphere

Abstract

This paper wants to highlight how the availability of measurements autoscaled at some reference ionospheric stations, and their assimilation by ionospheric models, was of crucial importance in determining, during the solar eclipse conditions occurred on 20 March 2015, a reliable representation of the ionosphere. Even though the solar eclipse falls in the recovery phase of the St. Patrick geomagnetic storm started on 17 March 2015, its influence on the ionospheric plasma seems undeniable. The reference ionospheric stations considered here are those of Rome (41°.8’ N, 12°.5’ E), and Gibilmanna (37°.9’ N,14°.0’ E), Italy. Specifically, in a time interval including that of the eclipse, the electron density profiles autoscaled by the Automatic Real-Time Ionogram Scaler with True-height (ARTIST) system at San Vito (40°.6′ N, 17°.8′ E), Italy, which are here considered as the truth profiles, were compared with both the electron density profiles calculated by the IRI-SIRMUP-Profiles (ISP) model, after assimilating data recorded at Rome and Gibilmanna, and the electron density profiles provided by the IRI-CCIR model. The ISP and IRI-CCIR performances were then evaluated in terms of the root mean square errors made on the whole electron density profiles. The three-dimensional (3-D) electron density mappings of the ionosphere provided by ISP and IRI-CCIR models were also considered as the ionospheric environment by the ray tracing software tool IONORT to calculate quasi-vertical synthesized ionograms over the short radio link San Vito – Brindisi (40°.4′ N, 17°.6′ E), Italy. The corresponding synthesized values of foF2 and fxF2, obtained by IONORT-ISP and IONORT-(IRI-CCIR) procedures, were compared with those autoscaled by ARTIST from the vertical ionograms recorded at the truth site of San Vito. Some examples of IONORT-ISP and IONORT-(IRI-CCIR)synthesized ionograms are shown and discussed. Finally, comparisons in terms of foF2 deduced by long-term prediction and nowcasting maps are also shown. The results achieved in this work demonstrate how the assimilation of autoscaled data into the ionospheric models turned out to be valuable in providing a better representation of the ionospheric electron density under very unusual conditions.