Abstract The total electron content (TEC) is a critical parameter of ionosphere morphology. Precise modeling and prediction of the TEC time series can significantly aid the functioning of the Global Navigation Satellite System (GNSS), satellite as well as shortwave communications. An improved Transformer-based TEC prediction model (ITM) is proposed in this study, in response to the high noise characteristics of the temporal data for TEC in the global ionosphere. The ITM model forecasts the global ionospheric TEC data provided by the Center for Orbit Determination in Europe (CODE) for the year 2018. Experimental results indicate that the model obtains an annual average root mean square error of 1.4 TECU, which represents a 6.67 % advancement over the daily forecast product C1PG published by the CODE center. The model also attains an annual average mean absolute error of 1 TECU, representing a 9.09 % improvement over the accuracy of C1PG. The annual average correlation coefficient is 0.98, which indicates a 0.62 % upgrade in C1PG’s accuracy. Additionally, the model exhibits a maximum increase of 23.53 % in monthly average root mean square error over C1PG, a maximum increase of 25 % in monthly average mean absolute error, and a maximum increase of 1.55 % in monthly average correlation coefficient. During the more geomagnetically active phases of 2018, the ITM model's forecast accuracy exhibited an overall improvement of 5 % in RMSE, 7.14 % in MAE, and 0.83 % in correlation coefficient R compared to the C1PG product.


    Zugriff

    Zugriff prüfen

    Verfügbarkeit in meiner Bibliothek prüfen

    Bestellung bei Subito €


    Exportieren, teilen und zitieren



    Titel :

    Forecast of global ionospheric TEC using an improved transformer model


    Beteiligte:
    Wu, Xuequn (Autor:in) / Fan, Cihang (Autor:in) / Tang, Jun (Autor:in) / Cheng, Yuesong (Autor:in)

    Erschienen in:

    Advances in Space Research ; 73 , 9 ; 4519-4538


    Erscheinungsdatum :

    2024-02-01


    Format / Umfang :

    20 pages




    Medientyp :

    Aufsatz (Zeitschrift)


    Format :

    Elektronische Ressource


    Sprache :

    Englisch




    Ionospheric Specification and Forecast Modeling

    Robert W. Schunk / Ludger Scherliess / Jan J. Sojka | AIAA | 2002


    Ionospheric specification and forecast modeling

    Schunk, R. / Scherliess, L. / Sojika, J. | AIAA | 2001



    Machine learning algorithm to forecast ionospheric time delays using Global Navigation satellite system observations

    Mallika I, Lakshmi / Ratnam, D. Venkata / Raman, Saravana et al. | Elsevier | 2020