Reevaluating the Predictive Power of Sunspot Area, Hale, and McIntosh Classifications for CME Speed in Solar Cycles 23 – 25
摘要
We investigate the long-assumed relationship between sunspot group properties and coronal mass ejection (CME) speed using a manually verified dataset of 1488 M- and X-class flare-CME associations spanning Solar Cycles 23, 24, and 25. For each event, the associated sunspot area, Hale magnetic classification, and McIntosh class were compiled based on the active region identified at the flare time, animated GOES X-ray flux plots, and movies from the Large Angle and Spectrometric Coronagraph. Contrary to operational forecasting expectations, we find that these traditional photospheric parameters are consistently poor predictors of CME speed. Across all three solar cycles and CME speed categories; Pearson and Spearman correlation coefficients and multivariable regression models show that these metrics only account for less than 4% variance in CME speed. These results demonstrate that even large and magnetically complex active regions do not reliably produce fast CMEs, while some of the fastest events originate from small or simple regions. These results highlight the limitations of traditional sunspot indicators, which exclude key magnetic structures such as plages, in capturing the full eruptive potential of active regions and fail to capture coronal conditions that govern CME acceleration. Improved CME forecasting may require greater emphasis on coronal magnetic topology and non-spot magnetic environments.