Radio studies of supernovae 1979C, 1986J and 2006X with LOFAR

TL;DR

This study leverages LOFAR's low-frequency capabilities to probe the long-term radio evolution of three supernovae in two nearby galaxies, focusing on ejecta–CSM interaction and the energy partition between magnetic fields and relativistic particles. By combining LoTSS 0.146 GHz data with ILT imaging and multi-instrument follow-up, the authors derive tight constraints on the circumstellar density for SN 2006X, detect and characterize the late-time radio emission of SN 1979C, and resolve the complex structure of SN 1986J, including its central component. Hydrodynamic modeling suggests an outer ejecta density slope of $n\approx8$ for SN 1979C between 22–42 years, with a total ejecta mass of $M\approx5.2\,M_\odot$ and kinetic energy $E\approx(1.6$–$2.2)\times10^{51}$ erg, while spectral breaks around 1.5 GHz are consistent with synchrotron cooling rather than central-object emission. SN 1986J shows a rapidly decaying shell and a relatively constant central component, possibly indicating a pulsar-wind nebula or shocked central ejecta, highlighting the importance of high-resolution, low-frequency observations to disentangle multiple emission components. Overall, the work demonstrates LOFAR's capability to illuminate the late-time, low-frequency evolution of SNe and motivates ongoing monitoring to refine ejecta structures and CSM environments in aging remnants.

Abstract

We present LOw Frequency ARray (LOFAR) studies of supernovae SN 1979C, SN 1986J, and SN 2006X, focusing on new observations from the LOFAR Two-metre Sky Survey (LoTSS) and the International LOFAR Telescope (ILT). For Type Ia SN 2006X, we derive a 3$σ$ upper limit of 0.7 mJy at 0.146 GHz, and using radio emission models based on the CS15DD2 explosion model, we constrain the circumstellar density to $n_{\rm H} \lesssim 10$ cm$^{-3}$ for the microphysical parameters $ε_{\rm rel} = ε_{\rm B} = 0.01$. SN 1979C is clearly detected in the LoTSS image with a flux density of $4.6 \pm 0.36$ mJy nearly 40 years postexplosion. Modeling its radio evolution suggests a steep flux decay ($F_ν \propto t^{-2.1}$) between 22 and 42 years, a break in the spectrum near 1.5 GHz possibly due to synchrotron cooling, a progenitor mass of $\sim 13$ solar masses, and a progressive steepening with velocity for the density slope of the supernova ejecta. Our findings for SN 1979C contradict scenarios involving central compact object emission, and we obtain X-ray temperatures close to those derived from recent observations. For SN 1986J, we present the first ILT image showing a flux density of $6.77\pm0.2$ mJy at 0.146 GHz. The spectral index of the shell emission is found to be $0.66\pm0.03$, consistent with previous estimates, although variations at low frequencies warrant further investigation. Our results highlight the power of LOFAR for studying long-term radio evolution in supernovae.

Figures (9)

• Figure 1: LOFAR image of M100 at 6$^{\prime\prime}$ resolution at a central frequency of 146 MHz. Left panel shows the galaxy while the right panel is a zoom-in of SN 1979C. The purple area to the lower left in both panels marks the resolution element of the images.
• Figure 2: LOFAR image of NGC 891 at 6$^{\prime\prime}$ resolution at a central frequency of 146 MHz. Left panel shows the galaxy while the right panel is a zoom-in of SN 1986J. The purple area to the lower left in the panel to the right marks the resolution element of the image.
• Figure 3: ILT image of NGC 891 at 0$.\!\!^{\prime\prime}$54 $\times$ 0$.\!\!^{\prime\prime}$28 resolution at a central frequency of 146 MHz. Left panel shows the central part of the galaxy while the right panel is a zoom-in of SN 1986J. The purple area to the lower left in the panel to the right marks the resolution element of the image.
• Figure 4: Modeled radio light curves for SN 2006X for five frequencies (0.146, 1.5, 4.8, 6.1 and 8.4 GHz) along with all data listed in Table \ref{['tab:06xfluxes']}. The model parameters are $\epsilon_{\rm rel} = \epsilon_{\rm B} = 0.01$, $n=10$, $p=2.8$, and n$_{\rm H}$=10.1 cm$^{-3}$. For this value of n$_{\rm H}$ the predicted luminosity at 1.5 GHz after 14.57 years equals the observed $3\sigma$ upper limit on 2020 Aug 30, while the modeled luminosities fall below the observed $3\sigma$ upper limits of all other epochs/frequencies in Table \ref{['tab:06xfluxes']}. For the distance to SN 2006X we have used 17.1 Mpc. See text for further details.
• Figure 5: Derived values of $n_{\rm H}$ as a function of $\epsilon_{\rm rel}$ for various models with fixed values of $\epsilon_{\rm B}$ between $0.001-0.1,$ and for $n=10$ and $n=12$. The criteria for the models are that they should produce a 1.5 GHz luminosity at 14.57 years that agrees with the observed e-MERLIN $3\sigma$ upper limit, and that $t_{\rm b} \geq 14.57$ years. For the three upper lines, $t_{\rm b} = 14.57$ years at their left ends, and the solution for lower values of $\epsilon_{\rm B}$ becomes much less constraining. Note that for $\epsilon_{\rm rel} = \epsilon_{\rm B} = 0.01$ and $n=10$ the value of $n_{\rm H}$ used in Figure \ref{['fig:06_lightcurves']} has been marked. The upper limit at $\epsilon_{\rm B} = 0.1$ marked in blue is the derived upper limit by 2016ApJ...821..119C for $\epsilon_{\rm rel} = \epsilon_{\rm B} = 0.1$, $p=3$, $n=10.18$, and using the 4.8 GHz VLA data at 0.786 years.