Do Young Suns Produce Frequent, Massive CMEs? Results from Five-year Dedicated Optical Observations of EK Draconis and V889 Hercules

TL;DR

This work investigates flares and coronal mass ejections (CMEs) from two nearby young solar-type stars, EK Dra and V889 Her, via a five-year optical campaign using high-cadence H$\alpha$ spectroscopy and TESS photometry. By detecting $15$ H$\alpha$ superflares and analyzing line-profile signatures, the authors constrain flare heating, filament/prominence eruptions, and CME occurrences, deriving a lower-limit CME association rate of $\sim27\%$ and a direct CME-driven mass-loss rate for EK Dra of $\sim4\times(10^{-13}$--$10^{-12})\ M_{\odot}$ yr$^{-1}$. The results reveal strong chromospheric heating in superflares ($H\alpha$ broadening up to $14.1\pm2.4$ Å) and provide the first direct, multi-faceted observational view of CMEs/eruptions in young solar analogs, with implications for early planetary space weather and stellar evolution. Collectively, this study advances our understanding of the young Sun’s activity, offering empirical constraints on how intense flares and CMEs could have shaped planetary atmospheres and angular-momentum loss in the early solar system.

Abstract

We report results from a five-year (132-night) dedicated observational campaign targeting two nearby young solar-type stars, EK Draconis ($\sim$50-125 Myr age) and V889 Hercules ($\sim$30 Myr age), using the 3.8m Seimei Telescope and Transiting Exoplanet Survey Satellite. The aim is to observationally constrain statistical properties of flaring radiation/heating as well as coronal mass ejections (CMEs), through high time-cadence H$α$ spectroscopy. We obtained an unprecedented sample of 15 H$α$ superflares, including two blueshifted absorption, two blueshifted emission, one redshifted emission, and nine line broadening events. We obtain the following results: (1) Larger flares exhibit broader H$α$ line widths, up to 14.1$_{\pm 2.4}$ Å, indicating higher chromospheric heating than solar flares. (2) The long-lasting redshifted event at $\sim$100 km s$^{-1}$ may indicate dense post-flare loops. (3) H$α$ blueshifted absorptions/emissions provide evidence of massive filament/prominence eruptions, the core structures of CMEs. One newly identified event showed an unexpected rapid decrease in velocity. (4) The lower limit of the CME/eruption association rate with superflares is 27$_{-16}^{+25}$%, yielding occurrence rates of 0.21$_{\pm0.12}$ and $<$0.32$^{+0.46}_{-0.32}$ events per day for EK Draconis and V889 Hercules, respectively. (5) We derived the first direct estimate of the lower limit of the mass-loss rate driven by super-CMEs ($\gtrsim10^{33}$ erg) for EK Dra as $4 \times (10^{-13}$-$10^{-12})$ $M_{\odot}$ yr$^{-1}$, comparable to the stellar wind mass loss at a similar age. This study provides critical observational constraints on the radiation and plasma environment around young solar-type stars and the early Sun, which can drive planetary space weather and stellar mass/angular momentum loss.

Figures (19)

• Figure 1: H$\alpha$ spectra of the flare event EK3 on 21 April 2021. (a) Light curves from TESS (black) and H$\alpha$ equivalent width (red). Error bars are shown in the top right. The gray dashed line indicates the background level. The black arrow marks the time used to create the pre-flare template median spectrum. (b) Dynamic spectrum of the H$\alpha$ line after subtracting the pre-flare spectrum. The dotted line shows the line center. (c) Time evolution of the pre-flare subtracted H$\alpha$ spectra, shown from top to bottom. Each spectrum is time-averaged to improve S/N. Black error bars with points show uncertainties. Red lines are fitted spectra with the Gaussian function for data with sufficient S/N. Each fitting parameter of Doppler shift and line width is indicated in the upper right of each spectrum. The above explanations are applied to Figures \ref{['fig:ek3']}--\ref{['fig:v3']}.
• Figure 2: Same as Figure \ref{['fig:ek3']}, but for the flare event EK4 on 30 April 2021.
• Figure 3: Same as Figure \ref{['fig:ek3']}, but for the flare event EK5 on 23 February 2022. The TESS light curve shows three potential white-light flares around $t \sim 10$, $\sim 210$, and $\sim 380$ minutes, although their significance is low compared to the noise level. The H$\alpha$ data appear to exhibit corresponding peaks; however, the first and second flares do not return to the quiescent level and are instead connected. The final H$\alpha$ peak at $t \sim 400$ minutes falls below the flare threshold ($\Delta$EW $> 0.1$). Therefore, although multiple H$\alpha$ peaks are present, they are treated as a single, complex flare in this study.
• Figure 4: Same as Figure \ref{['fig:ek3']}, but for the flare event EK9 on 8 March 2023.
• Figure 5: Same as Figure \ref{['fig:ek3']}, but for the flare event EK10 on 20 March 2024.