Fading Echoes of Interaction: Probing Centuries of Mass-Loss in Four Old Type IIn Supernovae

TL;DR

This work investigates the late-time mass-loss histories of four old Type IIn supernovae (SN 2013L, SN 2014ab, SN 2015da, and KISS15s) by combining Chandra X-ray data and VLA/GMRT radio observations. The authors model the X-ray emission as reverse-shock–dominated thermal plasma and translate luminosities into mass-loss rates, revealing that centuries-before mass-loss rates are typically 1–3 orders of magnitude lower than early optical inferences, suggesting rapid evolution in progenitor winds prior to explosion. For KISS15s, they also identify a radio spectral inversion consistent with a two-component emission, hinting at a binary-related or complex CSM structure and possible secondary shocks. Collectively, the results show that X-ray and radio observations at late times can recover progenitor mass-loss histories when optical data have faded, refining our understanding of LBV-like or binary progenitors and the diversity of CSM geometries in SNe IIn.

Abstract

Supernovae characterized by enduring narrow optical hydrogen emission lines (SNe IIn) are believed to result primarily from the core-collapse of massive stars undergoing sustained interaction with a dense circumstellar medium (CSM). While the properties of SN IIn progenitors have relatively few direct constraints, the ongoing ejecta-CSM interaction provides unique information about late-stage stellar mass-loss preceding core-collapse. We present late-time X-ray and radio observations of four $\geq$3000 day-old SNe IIn: SN 2013L, SN 2014ab, SN 2015da, and KISS15s. The radio and X-ray emission from KISS15s indicate a mass-loss rate of $\rm{\dot M\sim4\times 10^{-3}~{M_{\odot}\,yr^{-1}}}$ at $\sim$450 years pre-supernova -- 2 orders of magnitude below earlier optical estimates (which probed the mass-loss immediately preceding the supernova). We find hints of a spectral inversion in the radio SED of KISS15s; a possible signature of a secondary shock due to a binary system or the emergence of a pulsar wind. For SN 2013L, we obtain a mass-loss rate of $\rm{\dot M\sim2 \times 10^{-3}~\rm{M_{\odot}\,yr^{-1}}}$ at $\sim$400 years pre-explosion based on the X-ray detection. For SN 2014ab and SN 2015da, we find a upper limits on the mass-loss rates of $\rm{\dot M<2\times10^{-3}~M_{\odot}\,yr^{-1}}$ explosion at $\sim$ 250 and 300 years pre-explosion, respectively. All four objects display mass-loss rates lower than estimates from earlier optical analyses by at least 1-3 orders of magnitude, necessitating a rapidly evolving progenitor process over the last centuries pre-explosion. Our analysis reveals how X-ray and radio observations can elucidate progenitor evolution when these objects have faded at optical wavelengths.

Figures (8)

• Figure 1: A (simplified and not to scale) model of circumstellar interaction in a Type IIn supernova. Narrow optical emission lines originate in the unshocked photoionized CSM, while broad optical emission lines are produced by supernova ejecta. The ejecta interaction with the CSM creates a forward and reverse shock. The hot forward shock can accelerate the CSM to very high speeds, producing X-ray and radio synchrotron emission, which may be reprocessed into UV and optical. At late times, X-ray and UV emission from the cooler reverse shock dominates. Based on Chandra2018.
• Figure 2: Chandra ACIS-S (0.2-10 keV) unsubtracted single epoch observations of SN 2013L (above) and KISS15s (below). SNe circled in blue.
• Figure 3: Best-fit models for the two X-ray detected supernovae, SN 2013L (above) and KISS15s (below). We show the combined and binned source spectra (red--Chandra ACIS-S) with best-fit models and residuals in black. Parameters for the models are given in Table \ref{['table:x_parameters']}.
• Figure 4: Normalization and column density contour plots for SN 2013L (above) and KISS15s (below). $\mathrm{1\sigma}$, $\mathrm{2\sigma}$, and $\mathrm{3\sigma}$ confidence regions are shown in red, green, and blue, respectively. Best-fit model values are denoted by the cyan cross. Flattening of curves at low $\mathrm{\mathrm{N_{H}}}$ values signifies lower limit due to line-of-sight galactic hydrogen.
• Figure 5: KISS15s in VLA C band (4-8GHz, left), X band (8-12GHz, center), and K band (18-26.5GHz, right). The ellipse in the bottom left of each image signifies the beamsize (resolution element size). Contours are scaled individually for each band --- for C band, contours appear at 10, 21, 32, and 43 $\mu$Jy/beam; for X band, contours appear at 13, 18.5, 24, and 29.5 $\mu$Jy/beam; and for K band, contours appear at 18, 24, 30, and 36 $\mu$Jy/beam.