The Type Ibn Supernova 2019kbj - Indications for Diversity in Type Ibn Supernova Progenitors

Type Ibn supernovae (SNe) are a rare class of stellar explosions characterized by a lack of hydrogen lines and the presence of narrow He I emission lines in their spectra (Pastorello et al., 2007). These events are thought to be SNe strongly interacting with H-poor, helium-rich circumstellar material (CSM; e.g. Smith, 2016, and references therein). Only a few dozen of such events are known (see Pastorello et al., 2016; Hosseinzadeh et al., 2017, for recent compilations), and their progenitor systems remain a mystery.

The H-rich analogs of Type Ibn events, (i.e. explosions interacting with a H-rich CSM), known as Type IIn SNe, show slowly evolving and diverse light curves (e.g. Kiewe et al., 2012). These traits are explained by the fact that CSM interaction injects extra luminosity through shocks, producing the observed prolonged emission, while diverse CSM density distributions produce the observed diversity in light-curve shapes. However, Hosseinzadeh et al. (2017) showed that many type Ibn SN light curves are strikingly similar and rapidly evolving, in contrast to the expectations from CSM-interaction-powered emission (but see also outliers to this uniformity discussed in Pastorello et al. 2016).

Even more puzzling is the discovery of a Type Ibn SN in a brightest cluster galaxy (Sanders et al., 2013), specifically in an environment with extremely low star formation, leading Hosseinzadeh et al. (2019) to conclude that some (if not all) Type Ibn SNe might not even be explosions of massive stars, as typically assumed.

Possible clues as to the progenitors of Type Ibn SNe can come from modeling their bolometric light curves. Gangopadhyay et al. (2020) fit the bolometric light curve of the Type Ibn SN 2019uo with the Chatzopoulos et al. (2012) model that includes luminosity from both ⁵⁶Ni decay and CSM interaction (after disfavoring ⁵⁶Ni decay as the sole power source). Their best fits require $\sim 16M_{\odot}$ of ejecta and just $0.01M_{\odot}$ of ⁵⁶Ni, with most of the luminosity at peak coming from interaction of the ejecta with a few tenths of a solar mass of CSM. They favor a uniform-density shell, rather than a steady wind, for the distribution of the CSM. Pellegrino et al. (2022), on the other hand, find a much smaller ejecta mass ($\sim 1M_{\odot}$) for the same event, while finding a similar ⁵⁶Ni mass, using the same models.

Here we present observations of SN 2019kbj, a well-observed member of the Type Ibn class, with multiband photometry and multiepoch spectroscopy. We analyze its light curve and spectra and model its bolometric light curve in a similar way to that of Gangopadhyay et al. 2020 for SN 2019uo to deduce its physical parameters. With this analysis we aim to increase the sample of Type Ibn events with deduced physical parameters. We assume the Planck18 (Planck Collaboration et al., 2020) cosmology throughout.

SN 2019kbj was discovered on 2019 July 1 (UT used throughout) by the Asteroid Terrestrial-impact Last Alert System (ATLAS; Tonry et al., 2018) transient survey (Smith et al., 2020) as ATLAS19ohl (Tonry et al., 2019), at R.A. 01:00:39.619 and decl. +19:37:03.5 (J2000)¹¹1The event was independently discovered on 2019 July 27 by the Panoramic Survey Telescope And Rapid Response System (Pan-STARRS; Chambers et al., 2016) as PS19dzw.. A faint (absolute magnitude $\sim-17$) and blue host galaxy is seen in archival PS1 images (Flewelling et al., 2020) at this position.

The event was initially classified on 2019 July 3 by Hiramatsu et al. (2019) as a possible young Type II SN at a redshift of $z=0.048$, based on the strong blue continuum, narrow H emission, and possible early flash-spectroscopy features (short-lived high-ionization emission lines indicative of a confined CSM; e.g. Khazov et al., 2016). However, it was later reclassified by Arcavi et al. (2022) as a Type Ibn SN based on narrow He I emission lines (and a lack of broad H features) seen in a spectrum taken one week later (the narrow H emission being attributed to the host galaxy rather than the SN). The redshift remained unrevised.

We obtained $BVgri$-band imaging of SN 2019kbj with the Las Cumbres Observatory (Brown et al., 2013) Sinistro cameras mounted on the network of 1-meter telescopes at the Cerro Tololo Inter-American Observatory (Chile), the South African Astronomical Observatory (South Africa), the Siding Spring Observatory (Australia), and the McDonald Observatory (United States), through the Global Supernova Project, from 2019 July 4 to 2019 September 20. Reference images were obtained on 2021 December 31, long after the SN faded. Standard image-reduction procedures were applied by the Las Cumbres Beautiful Algorithms to Normalize Zillions of Astronomical Images (BANZAI) pipeline²²2https://github.com/LCOGT/banzai (McCully et al., 2018). We then performed image subtraction and point-spread function (PSF) fitting using the PyRAF-based lcogtsnpipe³³3https://github.com/LCOGT/lcogtsnpipe pipeline (Valenti et al., 2016), which uses the High Order Transform of PSF ANd Template Subtraction (HOTPANTS; Becker, 2015) implementation of the Alard & Lupton (1998) algorithm. $BV$-band magnitudes are calibrated to the Vega system, and $gri$-band magnitudes to the AB system. We also obtained $c$- and $o$-band host-subtracted photometry of SN 2019kbj from the ATLAS Forced Photometry Server⁴⁴4https://fallingstar-data.com/forcedphot/ (Tonry et al., 2018; Smith et al., 2020). We find the last pre-explosion ATLAS 5$\sigma$ nondetection limit to be on 2019 June 29 at a magnitude of $19.89$ in the $c$ band, constraining the explosion time to a window of only $2$ days between 2019 June 29 and 2019 July 1.

We downloaded images of SN 2019kbj taken by the Ultraviolet Optical Telescope (UVOT; Roming et al. 2005) on board the Neil Gehrels Swift Observatory (Gehrels et al., 2004), obtained under a Target of Opportunity request (PI: Hiramatsu), from the High Energy Astrophysics Science Archive Research Center (HEASARC)⁵⁵5https://heasarc.gsfc.nasa.gov/. We performed aperture photometry with a 5$\arcsec$-radius circular region using the uvotsource package in HEAsoft v6.18, with version 20200925 of the calibration database (CALDB), following the standard guidelines from Brown et al. (2009). Host flux subtraction was performed using images taken on 2022 April 12 (PI: Grupe), long after the SN faded, following the prescriptions of Brown et al. (2014).

We correct all photometry for Milky Way extinction using the Schlafly & Finkbeiner (2011) calibrations of the Schlegel et al. (1998) maps, retrieved via the NASA/IPAC Extragalactic Database (NED)⁶⁶6https://ned.ipac.caltech.edu/. For the ATLAS $c$ and $o$ bands we use extinction data for the $g$ and $r$ bands, respectively. We neglect extinction in the SN host galaxy, as we find no evidence for strong Na I D absorption in a spectrum taken of the host (see below). Our photometry is presented in Table 1 and in Figures 1–3.

We obtained 12 spectroscopic observations with the Las Cumbres Observatory Floyds spectrographs mounted on the 2-meter Faulkes Telescope North (FTN) and South (FTS) at Haleakala (United States) and Siding Spring (Australia) observatories, respectively, through the Global Supernova Project. Spectra were obtained through a 2″ slit placed on the SN along the parallactic angle (Filippenko, 1982). One-dimensional spectra were extracted, and flux and wavelength calibrated using the floyds_pipeline⁷⁷7https://github.com/LCOGT/floyds_pipeline (Valenti et al., 2013). One of the spectra is of the host galaxy, obtained long after the SN faded. A log of the spectroscopic measurements is given in Table 2. All SN spectra are presented in Figure 4. The host spectrum is presented in Figure 6.

The multiband light curve of SN 2019kbj is shown in Figure 1. Using a parabolic fit to the $r$-band data around peak (from MJD 58668.4 to 58677.1), we determine the peak date to be MJD $58670.1\pm 0.26$, with an apparent peak magnitude of $17.67\pm 0.24$, corresponding to an absolute peak magnitude of $-18.99\mathbf{\pm 0.24}$ (errors are from the parabolic fit).

We calculate the post-peak decline rate using a linear fit to the magnitudes between MJD $58670$ and $58700$ for each band (except the $c$ band for which there are not enough epochs). Our results are presented in Table 3. We find a decline rate in all bands similar to the typical $r$-band 0.1 mag day^-1 measured for Type Ibn SNe by Hosseinzadeh et al. (2017).

Comparing the $r$-band light curve of SN 2019kbj to those of other Type Ibn SNe (Fig. 2), we find that it is rather typical and fits well within the template of Hosseinzadeh et al. (2017) around peak. SN 2019kbj shows excess emission compared to the template starting at around 20 days after peak, perhaps due to a larger amount of ⁵⁶Ni compared to other events (see below).

The color evolution of SN 2019kbj is shown in Figure 3. Both its $B-r$ and $B-V$ colors are roughly constant, as seen also in other Type Ibn SNe. SN 2019kbj is one of the bluest Ibn’s in the sample.

The spectroscopic evolution of SN 2019kbj is shown in the left panel of Figure 4. A blue continuum is seen in the early spectra, with no prominent flash-spectroscopy features. However, our earliest spectrum was obtained 2.41 days before peak, which is later than when flash features were observed in other Type Ibn events. In SN 2019uo (Gangopadhyay et al., 2020) prominent flash features were seen only up to 3.7 days before peak, while in SN 2010al (Pastorello et al., 2015a) flash features were seen 8 days before peak and disappeared four days later. Since our earliest spectrum is later than these times, we cannot rule out the existence of flash features for SN 2019kbj.

Prominent narrow He I emission lines can be seen throughout the evolution, together with Si II, Mg I and Ca II lines, which are seen in other Type Ibn SNe as well (right panel of Figure 4). A very prominent He I blend at $5015$ and $5047\text{\AA}$ develops shortly after peak. This blend is also seen in some other Type Ibn SNe (Fig. 5). It appears after peak, and at later times the $5047\text{\AA}$ component disappears. The $5015\text{\AA}$ component is further blended with O III $5007\text{\AA}$ which we attribute to the underlying host galaxy (Fig. 6).

Narrow H$\alpha$ emission is also seen in all epochs. This feature may be from the host galaxy or from H in the CSM surrounding the SN progenitor. Here, we attribute the narrow H$\alpha$ emission to the host galaxy for a few reasons. First, it becomes stronger relative to other features as the SN fades. Additionally, it is seen in our host-galaxy spectrum (Fig. 6) and as an extended feature in the two-dimensional spectra of our SNe (an example two-dimensional spectrum is shown in Figure 7). Since the host and SN spectra were each taken under different seeing conditions and with different slit orientations, it is not possible to accurately isolate the amount of H$\alpha$ or O III emission contributed by the host galaxy to each SN spectrum. Therefore, we can neither robustly associate nor rule out an association of a small amount of H$\alpha$ or O III with the SN.

We measure expansion velocites from the He I $5876$, $6678$, and $7065\textrm{\AA}$ lines as was done for SN 2019uo by Gangopadhyay et al. (2020). We first normalize the spectra with a parabolic fit to the continuum and then model each He P Cygni absorption line with a Gaussian. The offset between the best-fit Gaussian center and the line rest-frame wavelength is then translated to an expansion velocity. Our results are shown in Figure 8. All lines show expansion velocities of a few $10^{3}$ km s^-1 which increase with time during the first $\sim$30 days after peak. This is the same behavior seen in the sample of Ibn SNe analyzed by Gangopadhyay et al. (2020) and references therein.

SN 2019kbj is photometrically and spectroscopically similar to other Type Ibn SNe, with a post-peak luminosity decline rate that is similar to that of the average Type Ibn light curve (Fig. 2).

The blackbody temperature of SN 2019kbj is relatively constant around $15,000$ K, in contrast to the cooling seen in H-rich Type II SNe, for example (e.g. Valenti et al., 2014). This indicates that an ongoing power source, likely CSM interaction, continues to heat the ejecta as it expands. A constant color, indicative of constant temperature, is seen also in other Type Ibn’s (but not all; Fig. 3). The constant color and temperature might be an indication of a common CSM-interaction power source for this class of events. Determining why some events do not show constant color requires additional modeling of those data, which we leave to future work.

The bolometric light curve of SN 2019kbj yields extreme parameter values when fit by radioactive decay alone, but can be fit with much more reasonable values with the addition of CSM-interaction luminosity.

The steady-wind CSM ($s=2$) models require higher Ni masses and lower ejecta masses compared to a uniform-density CSM shell ($s=0$; Table 6). In fact, the steady-wind CSM models require most ($\sim$50–90%) of the ejecta to be Ni, while in the uniform-density CSM model less than 10% of the ejecta mass is Ni, as seen in most core-collapse SNe. This is the case regardless of the value of $n$ chosen. Therefore, we conclude that our models show a slight preference for a uniform-density CSM shell over a steady-wind CSM. This is consistent with the results of Karamehmetoglu et al. (2017) and Gangopadhyay et al. (2020) who also prefer a uniform-density CSM shell over a steady-wind CSM to explain the light curves of the Type Ibn SNe OGLE-2014-SN-131 and 2019uo.

In the uniform-density CSM case, the bolometric light curve of SN 2019kbj requires a ⁵⁶Ni mass of $0.08$–$0.1M_{\odot}$, which is an order of magnitude higher than the $0.01\,M_{\odot}$ derived by Gangopadhyay et al. (2020) and Pellegrino et al. (2022) for SN 2019uo (for both CSM cases). In the steady-wind CSM case, we derive an even higher value of $0.22\,M_{\odot}$ for the ⁵⁶Ni mass. The ejecta masses of $\sim 0.2$–$1.4\,M_{\odot}$ that we find are substantially lower than the $\sim 16\,M_{\odot}$ derived for SN 2019uo by Gangopadhyay et al. (2020), but overlap with the $\sim 1\,M_{\odot}$ found by Pellegrino et al. (2022) for that event.

The ejecta masses deduced are not highly sensitive to the value of $n$, and change only within a factor of two for $n=7$ vs. $n=10$. The ⁵⁶Ni and CSM masses are even less sensitive to $n$, and are in fact consistent within the errors for the different $n$ values tested.

The mass-loss rate, $\dot{M}$, that produced the CSM can be obtained from the continuity equation:

where $v_{w}$ is the CSM wind velocity, and we recall that $q=\rho_{CSM,in}r_{in}^{s}$. If we assume a constant wind velocity, then the mass-loss rate is

For the $s=2$ case, this results in a constant $\dot{M}$. Typical Wolf-Rayet wind velocities are of order 1000 km s^-1 (Crowther, 2007), which is consistent with the order of magnitude of the earliest He velocity we measure in Section 5 (before the CSM might have accelerated significantly due to impact from the SN ejecta). With our $s=2$ best-fit parameters we thus find a mass-loss rate on the order of $\sim$0.5$\left(\frac{v_{w}}{1000\,\textrm{km\,s}^{-1}}\right)$ $M_{\odot}$ yr^-1 (for both $n=7$ and $n=10$). At this mass-loss rate, to reach the total CSM mass from each fit, the mass-loss episode that shaped the light curve of SN 2019kbj would have lasted only $\sim$0.2–1.1 years.

The mass-loss rate found here is similar to the mass-loss rate derived for the Type IIn SN iPTF13z (0.1–2 ${M}_{\odot}$ yr^-1; Nyholm et al. 2017). Gangopadhyay et al. (2020) found a much higher value for SN 2019uo ($\sim$200$\left(\frac{v_{w}}{1000\,\textrm{km\,s}^{-1}}\right)$ $M_{\odot}$ yr^-1) which motivated them to rule out the $s=2$ model for that event.

For $s=0$, the mass loss rate is not constant for a constant $v_{w}$. Using the best fit parameters for this case, the mass loss rate at the inner CSM radius, $\dot{M}\left(r_{in}\right)$, is $\sim$16$\left(\frac{v_{w}}{1000\,\textrm{km\,s}^{-1}}\right)$ $M_{\odot}$ yr^-1 for $n=7$ and $\sim$2.5$\left(\frac{v_{w}}{1000\,\textrm{km\,s}^{-1}}\right)$ $M_{\odot}$ yr^-1 for $n=10$. We can not derive a timescale for the ejection of this shell, without knowing the difference between $r_{in}$ and the progenitor radius (which we can not constrain here). In addition to these uncertainties, $v_{w}$ might also not be constant in this type of CSM.

Although the light curve of SN 2019kbj is $\sim$1–2 magnitudes brighter than that of SN 2019uo, both are within the observed spread of Type Ibn SN luminosities (Fig. 2). Therefore, our results indicate a possible diversity in Type Ibn SN progenitor systems and explosions. However, this apparent diversity may be due, at least in part, to the different model implementation and fitting methods used for each event. As noted previously, Pellegrino et al. (2022) also fit a Ni decay + uniform-density CSM model to SN 2019uo, but find a very different ejecta mass than that found by Gangopadhyay et al. (2020), $\sim 1M_{\odot}$ vs. $\sim 16M_{\odot}$, using the same data. Comparing physical parameters between events requires fitting their light curves with the same models and methods. We plan to perform such systematic comparisons in future work (T. Ben-Ami et al. 2023, in preparation).

SN 2019kbj is similar both photometrically and spectroscopically to other Type Ibn SNe.

We show that the radioactive decay of ⁵⁶Ni is likely not enough to explain the light curve, but that an additional power source is needed. This is also evidenced by the roughly constant blackbody temperature of SN 2019kbj.

We fit the bolometric light curve of SN 2019kbj with a radioactive decay + CSM interaction model, and find that a uniform-density CSM shell produces more reasonable fits compared to a steady-wind CSM.

Other Type Ibn SNe show lower ⁵⁶Ni masses and higher ejecta masses compared to SN 2019kbj. This diversity might be intrinsic to Type Ibn SN progenitor systems and explosions, but it might also arise from differences in the way physical parameters have been inferred for different events. A systematic study of Type Ibn light curves could elucidate this issue and provide additional clues as to the puzzling nature of these events.

We thank the anonymous referee for useful comments that helped make the results of this paper more robust and insightful. T.B.-A acknowledges support from the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation program (grant No. 852097). IA is a CIFAR Azrieli Global Scholar in the Gravity and the Extreme Universe program and acknowledges support from that program, from the ERC under the European Union’s Horizon 2020 research and innovation program (grant No. 852097), from the Israel Science Foundation (grant No. 2752/19), from the United States - Israel Binational Science Foundation (BSF), and from the Israeli Council for Higher Education Alon Fellowship. The Las Cumbres group is supported by AST-1911151, AST-1911225, and NASA Swift grant No. 80NSSC19K1639. This work made use of data from the Las Cumbres Observatory global network of robotic telescopes, and of data from the Asteroid Terrestrial-impact Last Alert System (ATLAS) project. ATLAS is primarily funded to search for near earth asteroids through NASA grant Nos. NN12AR55G, 80NSSC18K0284, and 80NSSC18K1575; by-products of the NEO search include images and catalogs from the survey area. This work was partially funded by Kepler/K2 grant No. J1944/80NSSC19K0112 and HST GO-15889, and STFC grant Nos. ST/T000198/1 and ST/S006109/1. The ATLAS science products have been made possible through the contributions of the University of Hawaii Institute for Astronomy, the Queen’s University Belfast, the Space Telescope Science Institute, the South African Astronomical Observatory, and The Millennium Institute of Astrophysics (MAS), Chile. This work also made use of the NASA/IPAC Extragalactic Database (NED), which is funded by the National Aeronautics and Space Administration and operated by the California Institute of Technology, and of data, software, and web tools obtained from the High Energy Astrophysics Science Archive Research Center (HEASARC), a service of the Astrophysics Science Division at NASA/GSFC and of the Smithsonian Astrophysical Observatory’s High Energy Astrophysics Division.