Prediction of coronal mass ejection transit times using machine learning-assisted drag based model.
Saved in:
| Title: | Prediction of coronal mass ejection transit times using machine learning-assisted drag based model. |
|---|---|
| Authors: | Dani, Tiar1,2 (AUTHOR), Winarko, Edi1 (AUTHOR) ewinarko@ugm.ac.id, Heryawan, Lukman1 (AUTHOR), Muhamad, Johan2 (AUTHOR), Kesumaningrum, Rasdewita2 (AUTHOR), Sulistiani, Santi2 (AUTHOR), Nurzaman, Muhamad Zamzam2,3 (AUTHOR), Pangestu, Ayu Dyah2 (AUTHOR), Ratnasari, Elvina Ayu2 (AUTHOR), Juangsih, Mira2 (AUTHOR), Budi, Bakuh D.S.4,5 (AUTHOR), Najla, Farah6 (AUTHOR) |
| Source: | Advances in Space Research. Nov2025, Vol. 76 Issue 9, p5714-5730. 17p. |
| Subjects: | Coronal mass ejections, Machine learning, Space environment, Random forest algorithms, Drag (Aerodynamics), Solar wind, Time |
| Abstract: | Coronal mass ejections (CMEs) are the main drivers of space weather and significantly influence the near-Earth environment. Given the potentially severe impact of CMEs on Earth, it is essential to develop models that can accurately predict their arrival times. The Drag Based Model (DBM) is a physics-based model used to predict the propagation of CMEs through the interplanetary medium, specifically predicting CME transit time (TT) and arrival speed ( v as ) at 1 AU. DBM is known for its simplicity, computational efficiency, and ability to provide fairly accurate predictions, as demonstrated in several studies achieving mean absolute errors (MAE) in the range of 10–15 h for CME TT. However, DBM relies on constant solar wind speed (ω) and a static drag coefficient (γ), which may limit its predictive accuracy. This study introduces Random Forest Regressor models to predict dynamic ω and γ to improve the accuracy of the DBM. The models are trained on a dataset comprising CME parameters, including CME linear speed, CME speed at 20 R ⊙ , CME angular width, and CME measurement position angle (MPA). Additionally, sunspot number and solar wind speed at 1 AU, corresponding to the CME onset time, are incorporated as supplementary inputs. The target dataset consists of values ω and γ derived from analytical solutions. Furthermore, we propose a novel approach by predicting γ indirectly through a γ / ω model. Our results demonstrate that the γ / ω model combined with constant ω provides better accuracy for DBM in predicting TT and v as than other static or dynamic combinations of ω and γ. This machine learning-assisted DBM framework achieves the lowest MAE of 8.73 h for TT and 74 km/s for v as , which is comparable to various models on the same subject. [ABSTRACT FROM AUTHOR] |
| Copyright of Advances in Space Research is the property of Pergamon Press - An Imprint of Elsevier Science and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.) | |
| Database: | Engineering Source |
Be the first to leave a comment!
