Automated Detection of Accelerating Solar Eruptions Using Parabolic Hough Transform

Abstract

Solar eruptions such as coronal mass ejections (CMEs) observed in the inner solar corona (up to 4 R⊙) show acceleration profiles that appear as parabolic ridges in height–time plots. Inspired by the white-light automated detection algorithms Computer Aided CME Tracking System (CACTus) and Solar Eruptive Events Detection System (SEEDS), we employ the parabolic Hough transform for the first time to automatically detect off-disk solar eruptions from height–time plots. Due to the limited availability of white-light observations in the inner corona, we use extreme ultraviolet (EUV) images of the Sun. In this article we present a new algorithm, CME Identification in Inner Solar Corona (CIISCO), that is based on Fourier motion filtering and the parabolic Hough transform, and we demonstrate its implementation using EUV observations taken by the Atmospheric Imaging Assembly (AIA) onboard the Solar Dynamics Observatory (SDO), Extreme Ultra Violet Imager (EUVI) onboard the STEREO-A and -B spacecraft, and Sun Watcher using Active Pixel System detector and Image Processing (SWAP) onboard the PRoject for On-Board Autonomy-2 (PROBA2) spacecraft. We show that CIISCO is able to identify off-disk, outward-moving features in EUV images. The use of automated detection algorithms, such as CIISCO, can potentially be used to provide early warnings of CMEs if an EUV telescope is located at ±90∘ from the Sun–Earth line, providing CME characteristics and kinematics close to the Sun. This article also presents the limitations of this algorithm and the prospects for future improvement.