Radio and Optical Flares on the dMe Flare Star EV Lac

TL;DR

This study analyzes a four-day, multi-wavelength campaign on EV Lac to connect radio gyrosynchrotron emission with optical flare spectra. By combining 27 hours of radio and 29 hours of optical data, the authors find that only 4 of 9 optically detected flares with radio coverage show a radio response, and optical photometry alone does not predict radio association. Time-resolved spectroscopy suggests that radio-associated flares require electron distributions with higher low-energy cutoffs, implying harder spectra (e.g., $E_c$ up to ≈150 keV) and beam fluxes $F\,\ge\,10^{12}$ erg cm$^{-2}$ s$^{-1}$; optical peaks typically precede radio peaks by about $1$–$7$ minutes, consistent with multiple loops or trap versus precipitating electrons. The analysis also identifies a frequency-dependent delay in radio peaks for large flares, which may arise from free-free opacity increases in a chromospheric evaporation front, constraining plasma conditions to densities $n_e$ around $10^9$–$10^{10}$ cm$^{-3}$ and scales on the order of 0.01–1 Mm, rather than loop lengths. Overall, the work advances our understanding of particle acceleration and transport in stellar flares and highlights the value of high-cadence, broad-band, time-resolved spectroscopy for disentangling flare physics across atmospheric layers.

Abstract

We present the results of a coordinated campaign to observe radio and optical stellar flares from the nearby M dwarf flare star EV~Lac. From a total of 27 hours of radio and 29 hours of optical observations, we examine the correspondence of the action of accelerated electrons of different energies in two distinct regions of the stellar atmosphere. We find that out of 9 optical flares with suitable radio coverage, only four have plausible evidence for a radio response. Optical photometric properties cannot predict which flares will have a radio response. From flares with time-resolved optical spectroscopy available, optical-only flares have similar implied electron distributions, while those with radio responses better correlate with higher low-energy cutoffs. The optical flares with a radio response all exhibit a delay between the optical and radio peaks of $\approx$1-7 minutes, with the optical flare peaking earlier in all cases. This likely indicates multiple loops are involved in the event, and/or the different impacts on electrons trapped in a magnetic loop (producing radio emission), versus those directly precipitating from the loop (producing the optical flare). We also remark on the radio spectral index behavior at early times for the largest radio flare observed in this study, which we interpret as evidence for increased opacity from a chromospheric evaporation front.

Figures (11)

• Figure 1: Radio light curves. Top panel of each subfigure displays the 3.6 cm light curve and 1$\sigma$ errors, middle panel displays 6 cm light curve and 1$\sigma$ errors, and bottom panel shows value of $\phi$ calculated for each 60s time bin. $\phi$ is a discriminator of correlated multi-band variability, and is defined in Equation \ref{['eqn:phi']}. Dotted line shows $\phi=6$ line, and dashed line indicates $\phi=10$ line. Grey shaded region indicates the time of optical observation coverage.
• Figure 2: (Left panel) Scatter plot of $\phi$ against 3.6 cm flux density in $\mu$Jy; only positive values are shown. Each flux density measurement is accompanied by error bars. (Right panels) Distribution of flux densities in several bins of $\phi$. The vertical dotted line indicates the average flux density in each histogram. The distribution for $\phi<6$ is reproduced in the other $\phi$ distribution plots with a renormalization in grey. While the flux density variations in the uppermost panel indicate noise, and those in the bottom panel are clear measurements of correlated variability, the intermediate case demonstrates situations of correlated variability at lower S/N.
• Figure 3: Light curves and spectral index variations during two remarkable multi-frequency radio flares. Top panels display 3.6 cm (6 cm) band light curves in black (red), with 1$\sigma$ errors, with flux densities in mJy (=1000 $\mu$Jy). The black dotted line indicates the level of quiescent emission before the start of the flare. Bottom panels show variation of spectral index $\alpha$ during the flare. Black dashed line indicates the time of maximum observed flux at 3.6 cm, while red dashed lines indicate the time of 6 cm flux density maximum. During each flare the light curves are shown with time bins of 10s (60s) for the 2001 Sept. 20 (2009 Sept. 21=R1) flare, while before and after a larger time bin of 300s is employed. Grey shaded regions indicate times of calibrator scans. The large flare on EV Lac analyzed in ostenetal2005 is shown on the left for comparison with the largest radio flare in our sample (R1) on the right. Both display a frequency-dependent delay in the time of peak flux density, with the delay increasing to lower frequencies.
• Figure 4: Constraints on loop height as a function of observed radio flare decay time, under assumptions of strong diffusion and a single loop. Dotted lines show decay time of the radio flares considered here (taken from Table \ref{['tbl:radio']}).
• Figure 5: Overlap of optical photometry and radio light curves on the night of 21 Sept. 2009, during times of optical flare activity. Specific optical flares are labeled in the top plot. Top panel displays optical photometric brightness variations, while bottom panels display 3.6 cm (6 cm) light curves in black (red), along with the $\phi$ parameter in blue, as discussed in the text. Dotted lines connect time bins in the radio light curve with $\phi > 6$ with the optical light curve. The quiescent flux density values at 3.6 (6) cm are indicated in dotted lines in black (red), respectively. Light gray regions indicate gaps in data stream due to phase calibrator observations.