SNR 1987A : Spitzer data from days 6000 to 8000 revisited

TL;DR

This study revisits the Spitzer-derived SED of SN 1987A between days 6000–8000, testing whether the near-IR excess at 3–5 μm can be explained by self-absorbed free-free emission rather than warm amorphous carbon dust. By decomposing the IR–radio SED into five components (warm carbon dust, silicates, cold dust, free-free, and synchrotron) and fitting days 7200 and 7400, the authors find that a self-absorbed free-free model yields statistically competitive fits and a physically plausible, though tightly constrained, set of parameters (gas temperature ≈ 1.3–1.4×10^3 K, ionization proxy (1−ξ) ≈ 0.07–0.08, and a cold-dust mass around ∼0.16–0.18 M_⊙). The analysis reveals a linear growth of warm carbon and silicate dust masses over the period and suggests that collisionally heated, relatively cool gas in the ER could power the short-wavelength excess under specific conditions, with the free-free cut-off frequency tied to the total emitting gas mass. The findings motivate continued JWST observations and coordinated radio data to test the persistence and origin of the free-free hypothesis across later epochs. Overall, the work demonstrates that free-free emission can formally describe the 3–5 μm excess, though it requires particular parameter choices and remains subject to interpretation regarding the gas heating and mass in the ER.

Abstract

An excess emission has been observed by Spitzer in the [3, 5] micron range of the SNR 1987A spectrum. It is generally argued that this excess could be due to the presence of warm amorphous carbon dust in the equatorial ring (ER) around the supernova, but the proposed models all have problems. This prompted us to present an alternative view on the interpretation of the Spectral Energy Distribution (SED) of SNR 1987A from the near-IR wavelengths to the radio frequencies (from 3 micron up to 1.4 GHz), between 6000 and 8000 days after outburst. We argue that the origin of that excess could be attributed instead to a free-free emission. We show that under very specific conditions (the free-free is self-absorbed at a cut-off frequency imposed by the mass of the emitting region), it could be produced by collisional heating of the gas. We then discuss the time evolution of the various components of the SED. We establish a linear relationship between the growth of the warm carbon dust mass and that of the silicates dust during the analyzed period. Finally, we build the Spitzer light curves and we show that our models reproduce the observations pretty well, although our study clearly favors the free-free case. In conclusion, we argue that the free-free model provides a formally very good description of the data, however the model does require some very specific parameter choices, and results in an unusually low temperature for the ionized gas.

Figures (17)

• Figure 1: Light curves of SN 1987A from Spitzer observations prior to day 10500 Arendt2016. Purple, blue, green, orange, red = 3.6, 4.5, 5.8, 8, 24 $\mu$m. The dashed lines represent a model of the evolution as given by Eq. (7) of that work.
• Figure 2: $6\farcs4 \times 6\farcs4$ IRAC images of SNR 1987A. The first image in the top row is a single IRAC 3.6 $\mu$m exposure with the detector's $1\farcs2$ pixels. The following images show the improved resolution enabled by mosaicking all IRAC observations (including post-cryogenic data, i.e. day 6000-12000), collected at a wide variety of sub-pixel dither offsets and roll angles. The last image shows the application of a deconvolution procedure to the $0\farcs2$ mosaic. The second rows shows the same for 4.5 $\mu$m data. The bottom row shows 2 shorter wavelength HST images at their original resolutions and convolved to simulate the resolution of the deconvolved IRAC images. Adapted from Arendt2020.
• Figure 3: The correlations between both parameters of each component of the SED at day 7400 derived from our fitting procedure for the warm carbon case. The selected couple of parameters in each case are indicated by the red cross.
• Figure 4: SED of SNR 1987A at days 7200 (top) and 7400 (bottom) with our resulting fits. The short wavelengths range has been fitted with a warm carbon dust component. See text for the caution to be taken regarding the cold dust component fit. The measurement errors ($\sigma$) and the width of the filters are given by the height and width of the cross associated to each data point. $\chi^2$ and $\Delta$ as defined by Eq. (eq:chi2) and Eq. (eq:Delta) characterize the goodness of our fits. The symbols for the telescopes used are given in Table tab:telused and the color code is: red = warm carbon dust; blue = silicates dust; green = cold dust; yellow = radio synchrotron; black = sum of all the components. The lower panel below the SED displays $\delta(Flux) = Flux(measured) - Flux(fit)$. In that panel, red data points means that they lie outside the axis range. All the following figures representing the SED of SNR 1987A with our fits obey the same colors and legends.
• Figure 5: SED of SNR 1987A for the Spitzer data acquired during the period 6000 -- 8000 days with a warm carbon dust component to account for the excess emission at near-IR wavelengths. Symbols and colors follow the conventions used in Figure \ref{['fig:sed_wc']}. The two bottom panels are the interpolated data for the two epochs analysed in more detail, shown for comparison.