A Statistical Study of Solar Filament Eruptions
Abstract
Solar filaments are cool, dark channels of partially-ionized plasma that lie above the chromosphere. Their structure follows the neutral line between local regions of opposite magnetic polarity. Previous research (e.g. Schmieder et al. 2013, McCauley et al. 2015) has shown a positive correlation (70-80%) between the occurrence of filament eruptions and coronal mass ejections (CME’s). In this study, we attempt to use properties of the filament in order to predict whether or not a given filament will erupt. This prediction would help to better predict the occurrence of an oncoming CME. To track the evolution of a filament over time, a spatio-temporal algorithm that groups separate filament instances from the Heliophysics Event Knowledgebase (HEK) into filament tracks was developed. Filament features from the HEK metadata, such as length, chirality, and tilt are then combined with other physical features, such as the overlying decay index for two sets of filaments tracks - those that erupt and those that remain bound. Using statistical methods such as the Kolmogrov-Smirnov test and a Random Forest Classifier, we determine the effectiveness of the combined features in prediction. We conclude that there is significant overlap between the properties of filaments that erupt and those that do not, leading to predictions only ~5-10% above chance. However, the changes in features, such as a change in the filament's length over time, were determined to have the highest predictive power. We discuss the possible physical connections with the change in these features."This project has been supported by funding from the Division of Advanced Cyberinfrastructure within the Directorate for Computer and Information Science and Engineering, the Division of Astronomical Sciences within the Directorate for Mathematical and Physical Sciences, and the Division of Atmospheric and Geospace Sciences within the Directorate for Geosciences, under NSF award #1443061.”
- Publication:
-
AAS/Solar Physics Division Abstracts #47
- Pub Date:
- May 2016
- Bibcode:
- 2016SPD....47.0106S