Investigating magnetic activity cycles in solar-like oscillators using asteroseismic data from the K2 mission

TL;DR

The paper investigates magnetic activity cycles in solar-like oscillators using two K2 campaigns, leveraging δν measured from cross-correlation and peakbagging of mode frequencies in 20 KEYSTONE stars. It finds seven stars with δν exceeding their uncertainties, with detections spanning both methods and some method-dependent cases, and notes that δν tends to rise with rotation and effective temperature but shows little dependence on large frequency separation or metallicity. The analysis combines K2 results with Kepler-based activity in a broader parameter space, revealing partial consistency with previous Kepler studies and highlighting correlations with Teff and P_rot when using fitted frequencies. The work emphasizes the need for longer, multi-epoch monitoring to confirm activity cycles, especially in evolved stars, and demonstrates the value of combining asteroseismic techniques with activity proxies to probe stellar dynamos.

Abstract

We present the results of an investigation into the possible presence of magnetic activity cycles in stars observed in two observational campaigns by the K2 mission. This study was based on the KEYSTONE asteroseismic sample of solar-like oscillators, which contained 20 stars for which we were able to determine whether the asteroseismic p-mode frequencies varied in time. These frequency shifts ($δν$) were determined using a cross-correlation method and using the individual mode frequencies, obtained by fitting power spectra. Three stars were found to exhibit $δν$ larger than their associated errors ($σ_{δν}$) using both methods, while two more stars exhibited $δν>σ_{δν}$ when the cross correlation was used and a further two stars exhibited $δν>σ_{δν}$ when the fitted frequencies were used. When considering the whole sample of 20 stars, the amplitude of $δν$ showed no dependence on the large frequency separation and metallicity. However, $δν$ was observed to increase with rotation rate and effective temperature. Our sample contained a number of evolved subgiant stars, allowing us to expand the parameter space usually considered when comparing $δν$ with stellar parameters. While $δν$ was small for all of the evolved stars, one was found to have $δν>σ_{δν}$, raising the possibility that these evolved stars may still exhibit variable magnetic activity.

Figures (4)

• Figure 1: Left: Power spectra observed for EPIC 212509747 in campaigns 6 (black) and 17 (grey), with a smoothed version of campaign 6 (cyan dashed). Here the smoothing was performed using a boxcar of width 22 bins. The spectrum is limited to the frequency range $\nu_{\textrm{max}}\pm4\Delta\nu$. The vertical dot-dashed line indicates $\nu_{\textrm{max}}$, while the vertical dashed lines are separated from $\nu_{\textrm{max}}$ by multiples of $\Delta\nu$. Right: An example of a cross-correlation (black) obtained using a realisation of the power spectra in the left panel, the Lorentzian fitted to the cross correlation (cyan, dashed) and the location of the maximum of the Lorentzian (purple, dot-dashed).
• Figure 2: Comparison of the frequency shifts obtained using cross correlations and the fitted frequencies (data points). A 1:1 line is included for guidance.
• Figure 3: Variation in frequency shift with different stellar parameters. In all figures, the hexagon symbols denote results obtained using the cross-correlation method, while the star symbols denote results obtained using the fitted frequencies. Open circles highlight those stars where the observed shift was larger than the uncertainty. All frequency shifts are the absolute values since it is the magnitude of the shift that is important, not whether the frequencies have increased or decreased between campaigns. The grey data were based on the results of santos2019signatures. Top left: Large frequency separation, $\Delta\nu$. Top right: Effective temperature, $T_{\mathrm{eff}}$. Middle left: Surface gravity, $\log\,g$. Middle right: Metallicity ([Fe/H]). Bottom left: $\log \Delta S_{\mathrm{ph}}$. Bottom right: Rotation period, $P_{\text{rot}}$, where the inset focuses on the faster rotating stars in our sample.
• Figure 4: Variation in frequency shift with $\Delta S_\text{ph}$. The hexagon symbols denote results obtained using the cross-correlation method, while the star symbols denote results obtained using the fitted frequencies. Open circles highlight those stars where the observed shift was larger than the uncertainties.