Constraining the Mass Loss and the Kinetic Energy of Solar Coronal Mass Ejections with Far-Ultraviolet Flares

TL;DR

This study tackles the challenge of constraining stellar CME properties by using the Sun as a proxy and extending the established solar flare–CME mass relation into the far‑UV. By combining LASCO CME masses, GOES XRS flare data, and AIA1600/SDO FUV measurements across Solar Cycles 23–24 (1996–2019), it derives correlations between FUV flare peak flux at 1600 Å and CME mass, kinetic energy, and speed, enabling estimates of CME environments for Sun‑like stars. The analysis finds only weak but detectable log‑linear correlations (Pearson r around $0.2$–$0.4$), with halo CMEs biasing masses and correlations upward, and no strong solar‑cycle dependence. Collectively, the results provide a practical framework to infer CME properties from FUV flares, informing models of exoplanet atmospheric erosion around Sun‑like stars.

Abstract

Stellar eruptive events, such as flares and coronal mass ejections (CMEs), can affect planetary habitability by disturbing the stability of their atmospheres. For instance, strong stellar flares and CMEs can trigger atmospheric escape and, in extreme cases, may strip away the atmosphere completely. While stellar flares have been observed and explored at a wide range of wavelengths, the physical properties of stellar CMEs remain unconstrained due to the difficulty in observing them. In this context, the Sun provides our only window on the potential characteristics of CMEs on Sun-like stars. A correlation between solar X-ray flare peak flux and the mass of flare-associated solar CMEs has been reported using solar data collected during Solar Cycle 23 (Aarnio et al. 2011). Here, we build upon that work. We extend the correlation into the far-UV (FUV), where stellar flares are, and will continue to be, routinely detected with existing and future FUV observatories by incorporating data spanning two entire Solar Cycles (23 and 24; 1996-2019). Using three different space missions (CMEs from LASCO/SOHO, X-ray flare events from XRS/GOES, and FUV flares from AIA/SDO), we report a correlation between FUV flare peak flux and energy centered at 1600 Å and mass, kinetic energy, and linear speed of flare-associated CMEs. These empirical relations enable estimates of CME masses and kinetic energies from FUV flares on Sun-like stars. While direct stellar CME detections remain elusive, the correlations derived here are likely applicable to Sun-like stars and provide a working framework for evaluating exoplanet atmospheric erosion.

Figures (5)

• Figure 1: The mass distribution of solar CMEs in the Large Angle Spectrometric Coronagraph (LASCO) CME catalog dataset that occurred from 1996 to 2019. The mass values of CMEs that propagated towards or away from the line of sight of the instrument are labeled as halo CMEs in the LASCO CME database. The mass distribution of halo CMEs is presented as a dashed histogram. In total, we started with 21,945 LASCO CMEs, including 708 halo CMEs, in our analyses.
• Figure 2: The solar X-ray flare peak flux distribution of the GOES solar flare report from 1996 to 2019. We excluded flares without position values, since we require flares that are spatially and temporally coincident with CMEs.
• Figure 3: (Left) Histogram of solar CMEs occurred within $\pm2$ hours of the solar X-ray flare event start time in the GOES database as a function of angular separation between the corresponding CME and a flare. The dashed lines indicate spatial constraint used to pair flares and CMEs ($\pm45^{\circ}$ of angular separation). (Right) Histogram of selected CMEs within the spatial constraint among the flare-CME pairs that are within $\pm2$ hours start time differences as a function of time differences between the flare event start time and the CME detection time in minutes. The dashed lines indicate temporal constraint; CMEs that started 10 to 80 minutes after the start of a paired X-ray flare event. We found similar distribution peaks observed in aarnio2011solar, indicating that these constraints can also be applied to our extended LASCO CME and GOES X-ray flare database.
• Figure 4: One-hour averaged solar spectral irradiance at 1600 Å (left) and one-hour averaged detrended FUV flux from AIA1600 images (right). For clarity in this 10-year timeseries, one-hour averages are plotted; however, all detrending and quantitative analyses were performed using one-minute averaged data.
• Figure 5: The mass of the flare-associated LASCO CMEs with halo CMEs plotted along the X-ray peak flux of associated GOES X-ray flares that occurred during Solar Cycle 23 and 24 (a), Cycle 23 (b), and Cycle 24 (c), the kinetic energy of the corresponding CMEs along the X-ray peak flux during Cycle 23 and 24 (d), Cycle 23 (e), and Cycle 24 (f) and the linear speed of the CMEs along the X-ray peak flux during Cycle 23 and 24 (g), Cycle 23 (h), and Cycle 24 (i). The solar X-ray flare-associated regular CMEs are plotted as gray dots, while flare-associated halo CMEs are plotted as black dots. We binned the data with respect to the associated solar X-ray flares' peak flux to derive a log-linear correlation. Each bin, ordered from the lowest to highest flare peak flux, contains 94, 325, 744, 651, 372, and 274 samples for Cycle 23 and 24; 65, 217, 425, 429, 265, and 192 for Cycle 23; and 29, 108, 319, 222, 107, and 82 for Cycle 24. The median values of the bins are presented as blue dots. The error bars indicate the interquartile range (IQR) of the samples in the respective bins. The correlation derived from aarnio2011solar, which includes LASCO CMEs from 1996 to 2006, is presented as a green solid line in (a), (b), and (c), whereas the best fit derived from the blue dots is presented as a blue solid line. The shaded region around each best-fit line represents the 1$\sigma$ confidence interval of the regression model, reflecting the uncertainty in the slope and intercept estimates.