Linking Magnetic Field Diagnostics with 3D CME Speeds in Solar Active Regions

TL;DR

This paper investigates how pre-eruption magnetic diagnostics relate to 3D CME speeds by comparing two Green’s-function potential-field extrapolation approaches: PIL-based and PIL-centered ROI-based. It demonstrates that the critical height $h_{ m crit}$, derived from decay-index profiles, is the strongest predictor of CME speed, with ROI-weighted $h_{ m crit}$ achieving correlations around $r\approx0.73$ for slower CMEs, and PIL-based $h_{ m crit}$ yielding similar performance. The mean transverse field $B_t$ at coronal heights contributes weakly on its own, while the product $B_t\times R_f$ related to ribbon flux modestly improves multi-parameter fits to $r_p\approx0.76$ when combined with $h_{ m crit}$. Overall, $h_{ m crit}$ emerges as the primary, robust predictor of CME speed across methods, with ROI-based diagnostics offering a scalable path toward automated space-weather forecasting that mirrors PIL-tracking results.

Abstract

Understanding how active-region properties influence coronal mass ejection (CME) dynamics is essential for constraining eruption models and improving space-weather prediction. Magnetic diagnostics derived above polarity inversion lines (PILs), including the critical height ($h_{\rm crit}$) of torus instability onset, the overlying field strength ($B_{\rm t}$), and ribbon flux ($R_{\rm f}$), provide physically motivated measures of eruption onset. The two main aims of this work are to (i) show that $h_{\rm crit}$ and $B_{\rm t}$ can equally well predict CME speeds when evaluated over the region of interest (ROI) not directly above the PIL, and (ii) assess the value of $h_{\rm crit}$, $B_{\rm t}$ and $R_{\rm f}$ in predicting CME speed. Photospheric magnetograms are modeled with potential-field extrapolations to obtain decay index profiles. Critical heights above PILs correlate strongly with 3D CME speed ($r = 0.71$). Using ROIs of $\approx$ 1.8, 3.7, and 7.3 Mm), centered on the PIL, weighted $h_{\rm crit}$ from the 7.3x7.3 ROI provides the strongest correlation ($r = 0.73$), while mean $B_{\rm t}$ at 150 Mm is weaker ($r = 0.33$). Combining both offers little improvement ($r = 0.74$), confirming $h_{\rm crit}$ as the dominant predictor. CME speed correlates moderately with $B_{\rm t} \times R_{\rm f}$ ($r = 0.44$), and highest when combined with $h_{\rm crit}$ ($r = 0.76$). Thus, in potential field models, ROI-based critical heights are as predictive as those above the PIL, indicating that the broader active-region field structure is equally valid as a diagnostic. When all parameters are considered together, $h_{\rm crit}$ alone consistently shows the highest predictive power for CME speed.

Figures (11)

• Figure 1: Full disk line-of-sight magnetogram observed by HMI onboard SDO on 2017 September 4 at 19:48:42 UT with zoomed in cutout of the active region used for the application of the extrapolation method.
• Figure 2: Potential-field extrapolation for one of the studied active regions. The background shows the radial photospheric field from HMI, and coloured curves are traced field lines.
• Figure 3: Figure shows HMI $B_r$ cutouts with the longest detected PIL overplotted in red. The temporal sequence shows how the PIL is tracked across magnetograms over time.
• Figure 4: (a) Mean decay index profile along the PIL as a function of height.(b) Temporal evolution of the critical height, with red and blue dashed lines marking CME onset time and the first appearance time in SOHO/LASCO-C2
• Figure 5: Photospheric $B_z$ map with polarity inversion line (PIL, yellow) and overlaid regions of interest (ROIs): 5×5 pixels (blue), 10×10 pixels (green), and 20×20 pixels (red).