Ionospheric TEC Forecast Using Bi-LSTM With the Adam Optimizer During X-Class Solar Flares Occurred in the Year 2024 and Validation With IRI-2020
Satellite communication and navigation systems have become more essential to everyday life, but at the same time, understanding the effect of solar activity on these systems is vital. Total electron content (TEC) is a key factor affecting satellite signals. Solar flares affect the TEC variations, an...
Saved in:
| Main Authors: | , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Wiley
2025-01-01
|
| Series: | Advances in Astronomy |
| Online Access: | http://dx.doi.org/10.1155/aa/7000070 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Satellite communication and navigation systems have become more essential to everyday life, but at the same time, understanding the effect of solar activity on these systems is vital. Total electron content (TEC) is a key factor affecting satellite signals. Solar flares affect the TEC variations, and this research examines the forecast of TEC during various X-class solar flares that occurred in February, March, May, June, July, and August 2024, employing a bidirectional long short-term memory (Bi-LSTM) coupled with the Adam optimizer (Bi-LSTM-AO). The forecasted results were validated with the IRI-2020. This study uses a robust dataset encompassing more than 1 year of TEC data from the IONOLAB database, along with key solar and geomagnetic parameters such as Kp, Ap, SSN, and F10.7 obtained from NASA OMNIWeb. These potent solar flares were scrutinized to evaluate the model’s performance in forecasting TEC variations under extreme solar activity. The Bi-LSTM-AO model exhibited exceptional accuracy in predicting TEC values across these dates, consistently outperforming the IRI-2020 model. For example, on May 14, 2024, coinciding with the X8.79 solar flare, the Bi-LSTM-AO model achieved impressive performance metrics, including a root-mean-square error of 3.52, a mean absolute percentage error of 6.88%, a mean absolute gross error of 2.97, and a centered mean square deviation of 9.93. In contrast, the IRI-2020 model showed significantly higher error metrics, with an RMSE of 13.18, a MAPE of 23.61%, and a MAGE of 10.93. This research provides the development of a more accurate space weather forecasting model to increase the positional accuracy in navigation systems. The improved predictions can enhance the reliability of satellite-dependent systems, which are increasingly important for global communication and navigation systems. |
|---|---|
| ISSN: | 1687-7977 |