Radio Emission from a Nearby M dwarf Binary

TL;DR

We analyze archival VLA data at 4–8 GHz and complementary e-MERLIN observations to study radio emission from 2MASS J02132062+3648506 AB, a nearby M-dwarf binary with a wide T3 companion, and from 2MASS J04183483+213127, a chromospherically active L5 brown dwarf. The M-dwarf binary is detected with a peak flux density of ~356 μJy beam$^{-1}$ and shows quiescent, highly circularly polarized emission with $\alpha = -0.44 \pm 0.07$ and $f_c \approx 0.45$, giving $\log_{10}(L_R/L_{bol}) \approx -7.76$. Interpreting the emission as gyrosynchrotron radiation yields $B \lesssim 1.75\times 10^{2}$ G, emitting-region size $L \lesssim 1.54\,R_\star$, and $n_e \lesssim 2.91\times 10^{5}\ \mathrm{cm}^{-3}$; 2M0418 remains undetected, with upper limits set. A notable flux-density discrepancy between VLA and e-MERLIN is discussed in terms of variability and spatial filtering, highlighting the need for higher-resolution, time-domain follow-up to localize the emitter and quantify activity. These results expand the small sample of radio-active M-dwarf binaries and motivate continued high-resolution monitoring to probe magnetic activity and emission mechanisms in ultracool systems.

Abstract

We present the detection of the binary system 2MASS J02132062+3648506 AB using the Karl G. Jansky Very Large Array (VLA) archive data observed at 4-8 GHz. The system is a triple consisting of a tight binary ($\sim0.2"$) of two M dwarfs of spectral class M4.5 and M6.5 and a wide T3 brown dwarf companion ($\sim$16.4"). The binary displays coronal and chromospheric activity as traced by previously measured X-ray flux and H$α$ emission. We detect the unresolved binary at a peak flux density of $\sim356\ μ\mathrm{Jybeam}^{-1}$ at a signal-to-noise ratio (SNR) of $\sim36$ and determine a radio luminosity of $\mathrm{log}L_R/\mathrm{log}L_\mathrm{bol}\approx-7.76$. The radio emission is quiescent, polarised at a mean circular polarisation fraction $f_\mathrm{c}=45.20 \pm 1.58$ % and exhibits a spectral index $α=-0.44\pm0.07$ . We probe the binary using the Enhanced Multi-Element Remotely Linked Interferometer Network (e-MERLIN) with an angular resolution of $\sim40$ mas at 5 GHz and detect a component at a peak flux density of $\sim90\ μ$Jy $\mathrm{beam}^{-1}$ at a SNR $\sim5$ . We propose a gyrosynchrotron origin for the radio emission and estimate a magnetic field strength $B<174.86$ G, an emitting region of size $L<1.54$ times the radius of the M4.5 primary and a plasma number density $n_\mathrm{e}<2.91\times10^5\ \mathrm{cm}^{-3}$. The brown dwarf companion is not detected. Additionally, we have analysed observations of 2MASS J04183483+213127, a chromospherically active L5 brown dwarf which is also not detected. Accordingly, we place $3σ$ flux density upper limits at $36.9\ μ$Jy $\mathrm{beam}^{-1}$ and $42.3\ μ$Jy $\mathrm{beam}^{-1}$ for Stokes I and V respectively.

Figures (5)

• Figure 1: (a) Stokes I image of 2M0213 AB. The binary is unresolved and enclosed in the circle with a red outline. The binary is detected at a peak flux density of $\sim356\ \mu$Jy $\text{beam}^{-1}$. The $1\sigma$ r.m.s noise in the image is $\sim10\ \mu$Jy $\text{beam}^{-1}$ giving a SNR $\sim36$. The binary at the true coordinates at epoch 2017-11-15 is offset from the pointing centre by $\sim1.3$, that is, by $\sim2$ synthesised beamwidths due to proper motion effects described in Section \ref{['section:astrometry']}. The position of the wide T3 brown dwarf companion at a separation $\sim16.4$ is marked by the box. The red cross indicates the correlation centre ( pointing centre). (b) Stokes V image of 2M0213 AB. The binary is unresolved and enclosed in the circle with a red outline. The data has been phase-shifted to the true coordinates at epoch 2017-11-15. The binary is detected at a peak flux density of $\sim-174\ \mu$Jy $\text{beam}^{-1}$ where the negative indicates the left circular polarisation. The $1\sigma$ r.m.s noise in the image is $\sim4.6\ \mu$Jy $\text{beam}^{-1}$. The source is detected at SNR $\sim38$. Similarly, the position of the wide T3 brown dwarf companion is marked by the box. (c) Stokes I image of 2M0213 AB synthesised from e-MERLIN follow-up observations of the binary. The M4.5 component is detected and enclosed in a circle with a red outline. We do not show the position of the T3 dwarf since it is detected in the VLA data at a similar noise level to the e-MERLIN data at $\sim16\ \mu$Jy $\text{beam}^{-1}$. The synthesised beam in all images is indicated by the filled white circle with a black outline to the bottom left.
• Figure 2: Stokes I image of 2M0418. The position of the L5 brown dwarf is marked by the red square box in the image. A radio source was detected at $4\sigma$, where $\sigma$ is the thermal noise in the image, and at a peak flux density of $\sim16\ \mu$Jy $\text{beam}^{-1}$, and is enclosed in the red circle. The correlation centre (pointing centre) is indicated by the red-cross and the synthesised beam by the filled white circle with a black outline to the bottom left. We have not included Stokes V maps for observations of the L5 brown dwarf (2M0418) on account of the null detection.
• Figure 3: Flux light curves of 2M0213AB from the VLA data binned to a cadence of two minutes. The observation was $\sim55$ minutes long. The upper and lower subplots show the Stokes I and V light curves respectively. The error bars represent the $1\sigma$ thermal noise added in quadrature to the flux scaling error and the shaded region represent a $1\sigma$ standard deviation of $\sim133\ \mu\text{Jybeam}^{-1}$ and $\sim118\ \mu\text{Jybeam}^{-1}$ in the Stokes I and V respectively. The mean flux densities in the Stokes I and V are $\sim347\ \mu\text{Jybeam}^{-1}$ and $\sim-176\ \mu\text{Jybeam}^{-1}$ respectively and are indicated by the dashed lines in the respective subplots. Note: These values slightly differ from values obtained using imfit.
• Figure 4: Flux light curves of 2M0418 from the VLA data binned to a cadence of two minutes. The observation was $\sim55$ minutes long. The upper and lower subplots show the Stokes I and V light curves respectively. the shaded region indicate a 1,2 and 3$\sigma$ standard deviations from the mean flux density of $\sim-1.0\ \mu\text{Jybeam}^{-1}$ and $3.5\ \mu\text{Jybeam}^{-1}$ for the Stokes I and V is respectively. The mean values are represented by the dashed lines. The $1\sigma$ standard deviation for the Stokes I and V is $\sim16.3\ \mu\text{Jybeam}^{-1}$ and $\sim17.8\ \mu\text{Jybeam}^{-1}$ respectively which is consistent with the VLA's $1\sigma$ thermal noise estimated for a two minute cadence.
• Figure 5: An SED fit of the unresolved binary. The black curve represents the SED fit using different photometric catalogues described in Section \ref{['section:sed']}, with each data point representing a different catalogue. Flux error bars are included for each data point, although they are not visible due to the scaling. The two red data points depict VLA fluxes at 4 and 8 GHz, respectively. The green data point depicts e-MERLIN flux at 5 GHz. The blue broken line linking the SED data points to the blue data point at 8 GHz is a power law fit, which represents the expected flux for an SED extrapolated to 8 GHz.