SN 2017egm: A Helium-rich Superluminous Supernova with Multiple Bumps in the Light
Curves
Jiazheng Zhu1,2 , Ning Jiang1,2 , Subo Dong3 , Alexei V. Filippenko4 , Richard J. Rudy5,6 , A. Pastorello7 ,
Christopher Ashall8 , Subhash Bose9,10 , R. S. Post11, D. Bersier12 , Stefano Benetti7 , Thomas G. Brink4 , Ping Chen13 ,
Liming Dou14 , N. Elias-Rosa7 , Peter Lundqvist15 , Seppo Mattila16,17 , Ray W. Russell5, Michael L. Sitko18 ,
Auni Somero17 , M. D. Stritzinger19 , Tinggui Wang1,2 , Peter J. Brown20 , E. Cappellaro7 , Morgan Fraser21 ,
Erkki Kankare17 , S. Moran17, Simon Prentice12 , Tapio Pursimo22 , T. M. Reynolds17,23 , and WeiKang Zheng4
1 CAS Key Laboratory for Research in Galaxies and Cosmology, Department of Astronomy, University of Science and Technology of China, Hefei, 230026, Peopleʼs
Republic of China; jiazheng@mail.ustc.edu.cn, jnac@ustc.edu.cn
2 School of Astronomy and Space Sciences, University of Science and Technology of China, Hefei, 230026, Peopleʼs Republic of China
3 Kavli Institute for Astronomy and Astrophysics, Peking University, Yi He Yuan Road 5, Hai Dian District, Beijing 100871, People’s Republic of China
dongsubo@pku.edu.cn
4 Department of Astronomy, University of California, Berkeley, CA 94720-3411, USA
5 Physical Sciences Laboratory, The Aerospace Corporation M2-266, P.O. Box 92957, Los Angeles, CA 90009, USA
6 Kookoosint Scientific, 1530 Calle Portada, Camarillo, CA 93010, USA
7 INAF-Osservatorio Astronomico di Padova, Vicolo dell’Osservatorio 5, I-35122 Padova, Italy
8 Department of Physics, Virginia Tech, Blacksburg, VA 24061, USA
9 Department of Astronomy, The Ohio State University, 140 W. 18th Avenue, Columbus, OH 43210, USA
10 Center for Cosmology and AstroParticle Physics (CCAPP), The Ohio State University, 191 W. Woodruff Avenue, Columbus, OH 43210, USA
11 Post Observatory, Lexington, MA 02421, USA
12 Astrophysics Research Institute, Liverpool John Moores University, 146 Brownlow Hill, Liverpool L3 5RF, UK
13 Department of Particle Physics and Astrophysics, Weizmann Institute of Science, 234 Herzl Street, 7610001 Rehovot, Israel
14 Department of Astronomy, Guangzhou University, Guangzhou 510006, Peopleʼs Republic of China
15 Department of Astronomy and The Oskar Klein Centre, AlbaNova University Center, Stockholm University, SE-10691 Stockholm, Sweden
16 School of Sciences, European University Cyprus, Diogenes street, Engomi, 1516 Nicosia, Cyprus
17 Tuorla Observatory, Department of Physics and Astronomy, University of Turku, FI-20014 Turku, Finland
18 Center for Exoplanetary Systems, Space Science Institute, 4750 Walnut Street, Suite 205, Boulder, CO 80301, USA
19 Department of Physics and Astronomy, Aarhus University, Ny Munkegade 120, DK-8000 Aarhus C, Denmark
20 George P. and Cynthia Woods Mitchell Institute for Fundamental Physics & Astronomy, Texas A. & M. University, Department of Physics and Astronomy, 4242
TAMU, College Station, TX 77843, USA
21 School of Physics, O’Brien Centre for Science North, University College Dublin, Belfield, Dublin 4, Ireland
22 Nordic Optical Telescope, Apartado 474, E-38700 Santa Cruz de La Palma, Spain
23 Cosmic Dawn Center (DAWN), Niels Bohr Institute, University of Copenhagen, Jagtvej 128, DK-2200 København N, Denmark
Received 2022 December 29; revised 2023 March 2; accepted 2023 March 2; published 2023 May 23
Abstract
When discovered, SN 2017egm was the closest (redshift z= 0.03) hydrogen-poor superluminous supernova (SLSN-
I) and a rare case that exploded in a massive and metal-rich galaxy. Thus, it has since been extensively observed and
studied. We report spectroscopic data showing strong emission at around He I λ10830 and four He I absorption lines
in the optical. Consequently, we classify SN 2017egm as a member of an emerging population of helium-rich
SLSNe-I (i.e., SLSNe-Ib). We also present our late-time photometric observations. By combining them with archival
data, we analyze high-cadence ultraviolet, optical, and near-infrared light curves spanning from early pre-peak
(∼−20 days) to late phases (∼+300 days). We obtain its most complete bolometric light curve, in which multiple
bumps are identified. None of the previously proposed models can satisfactorily explain all main light-curve features,
while multiple interactions between the ejecta and circumstellar material (CSM) may explain the undulating features.
The prominent infrared excess with a blackbody luminosity of 107–108 Le detected in SN 2017egm could originate
from the emission of either an echo of a pre-existing dust shell or newly formed dust, offering an additional piece of
evidence supporting the ejecta–CSM interaction model. Moreover, our analysis of deep Chandra observations yields
the tightest-ever constraint on the X-ray emission of an SLSN-I, amounting to an X-ray-to-optical luminosity ratio
10−3 at late phases (∼100–200 days), which could help explore its close environment and central engine.
Unified Astronomy Thesaurus concepts: Supernovae (1668)
Supporting material: data behind figures, machine-readable tables
1. Introduction
Thanks to the rise of wide-field time-domain surveys, a rare
class of supernovae (SNe) that are ∼2–3 mag more luminous
than normal SNe have been revealed in the past one and a half
decades, known as the superluminous supernovae (SLSNe; see
a recent review by Gal-Yam 2019). As with normal SNe
(Filippenko 1997), SLSNe are categorized into two classes
based on their spectra around maximum light: hydrogen-poor
(SLSNe-I) and hydrogen-rich events (SLSNe-II; Gal-
Yam 2019).
Since the discovery of the first SLSN-I over 15 yr ago (SN
2005ap; Quimby et al. 2007), more than 100 such events have
been found, and identifications of diverse properties among
The Astrophysical Journal, 949:23 (16pp), 2023 May 20 https://doi.org/10.3847/1538-4357/acc2c3
© 2023. The Author(s). Published by the American Astronomical Society.
Original content from this work may be used under the terms
of the Creative Commons Attribution 4.0 licence. Any further
distribution of this work must maintain attribution to the author(s) and the title
of the work, journal citation and DOI.
1
them have motivated subclassifications. Subclasses with
rapidly and slowly evolving light curves have been proposed
(Gal-Yam 2012; Inserra et al. 2018; Quimby et al. 2018), but
whether clear divisions can be made is debated (e.g., Nicholl
et al. 2017b; Lunnan et al. 2018). Spectroscopically, their early-
time spectra often show a characteristic W-shaped feature near
4200Å produced by a pair of broad absorption features
associated with O II (Quimby et al. 2011). SLSNe-I have long
been linked to SNe-Ic due to their similarities in post-peak
spectra and a lack of both H and He (Pastorello et al. 2010;
Quimby et al. 2011). During the past few years, several SNe
initially classified as SLSNe-I have been found with He
spectral features. PTF10hgi shows both H and He lines
(Quimby et al. 2018), and is followed by a small sample of
He-rich SLSNe-I classified as SLSNe-Ib from the Zwicky
Transient Facility (ZTF; Yan et al. 2020). However, the studies
of SLSNe-Ib still lack high-quality multiwavelength photo-
metric and spectroscopic observations covering the early phase
to the late stages.
The mechanisms powering SLSNe-I remain elusive (Moriya
et al. 2018; Gal-Yam 2019). Several possible mechanisms
beyond the conventional means such as radioactive decay of
56Ni have been proposed to explain their extreme radiative
power, including energy injection from a rapidly rotating,
highly magnetic neutron star (e.g., Kasen 2010; Woosley 2010),
the interaction of the SN ejecta with circumstellar material
(CSM; e.g., Chevalier & Irwin 2011; Chatzopoulos et al. 2013;
Sorokina et al. 2016), and the pair-instability mechanism (e.g.,
Heger & Woosley 2002; Gal-Yam et al. 2009). Magnetar
models are often invoked in fitting the light curves. However,
some observational evidence already indicates that the
magnetar model cannot be the only process affecting the
luminosity and morphology for SLSNe-I, and multiple
processes may contribute to the optical emission. For example,
the light-curve undulations and sometimes prominent post-peak
bumps seen in some SLSNe-I light curves cannot be explained
by a simple magnetar model (e.g., Nicholl et al. 2016; Yan
et al. 2017a, 2017b; Anderson et al. 2018; Lunnan et al. 2020).
The CSM interaction model offers a natural explanation for the
bumps and fluctuations in light curves (e.g., Blinnikov 2017;
Inserra et al. 2017; Moriya et al. 2018), but the absence of
relatively narrow emission lines in the spectra of most SLSNe-I
is often regarded as a major issue for the CSM model.
Gaia17biu/SN 2017egm (with a redshift z= 0.03063) was
discovered by the Photometric Science Alerts Team of the
Gaia mission on 2017 May 23 (UT dates are used throughout
this paper) and subsequently classified as an SLSN-I (Dong
et al. 2017); it was the closest SLSN-I discovered by then. In
contrast to the typical dwarf and metal-poor host galaxies for
SLSNe-I, its host galaxy, NGC 3191, is massive (Dong et al.
2017) and has a mean metallicity near solar (Nicholl et al.
2017a; Bose et al. 2018; see also spatially resolved observa-
tions by Chen et al. 2017; Izzo et al. 2018). Its pre-peak near-
ultraviolet (NUV) to optical colors (Bose et al. 2018) are
similar to those of Gaia16apd (Kangas et al. 2017) and among
the bluest observed for an SLSN-I, while its peak luminosity
(Mg=−21 mag) is substantially lower than that of Gaia16apd.
Moreover, SN 2017egm holds the tightest upper limit of radio
luminosity among SLSNe-I (Bose et al. 2018). The sharp
peaks in the light curves of SN 2017egm are shown to be well
fitted by ejecta–CSM interaction (CSI) models, but are
difficult to explain by only a magnetar or radioactive decay of
56Ni (Wheeler et al. 2017). Hosseinzadeh et al. (2022) also
concluded that SN 2017egm cannot be explained well by the
magnetar model alone, given the multiple bumps in its
extended post-peak light curve.
As one of the closest SLSNe-I and having extensive
multiwavelength observations, SN 2017egm appears to be an
ideal target for further detailed studies, which may help us
understand the explosion mechanism of SLSNe-I. In this paper,
we present the most complete photometric and spectroscopic
analysis of SN 2017egm, taking our observational data and all
public data together. We present the observations and data
reduction in Section 2. In Section 3, we analyze its light curves
and spectra, and we classify it as an SLSN-Ib and compare it to
other SLSNe-I. Section 4 presents a discussion and our
conclusions. In this work, we adopt a luminosity distance of
DL= 138.7± 1.9 Mpc assuming a standard Planck cosmology
(Planck Collaboration et al. 2016).
2. Observations and Data
Multiband photometry from UV to mid-infrared (MIR)
wavelengths, as well as optical spectra spanning ∼300 days
since the discovery of SN 2017egm, were taken by us or
gathered from publicly available archives. Here we describe the
relevant observations and data reduction. In addition, analysis
of the four epochs of Chandra visits is presented for the
first time.
2.1. Ground-based Optical and Near-infrared Photometry
We obtained late-time multiband images with a set of
ground-based instruments. Optical data in the BVgriz bands
were taken with 0.6 m telescopes at Post Observatory (PO) at
the Sierra Remote Observatories (SRO; CA, USA), the 2.0 m
Las Cumbres Observatory Global Telescope network
(LCOGT), the 2.0 m Liverpool Telescope (LT) at La Palma,
and the ALFOSC mounted on the 2.6 m Nordic Optical
Telescope (NOT). The JHK near-infrared (NIR) images were
obtained with the NIR Camera mounted on the NOT
(NOTCam) and the NIR Wide-Field Camera mounted on the
United Kingdom Infrared Telescope (UKIRT).
We found that the transient signals associated with
SN 2017egm were undetectable after ∼360 days in all of our
optical and NIR images. Thus, we generate the reference
images with these late-stage data, which are dominated by host-
galaxy emission around the position of SN 2017egm. All
photometric measurements were performed on the difference
images, which are obtained by subtracting the reference image
from a current science image by matching the point-spread
function (PSF). Specifically, we used HOTPANTS (Becker
2015) for image subtraction. Before subtraction, we removed
cosmic rays and aligned the images using Astrometry.net. Most
instruments have good references with high signal-to-noise
ratios (S/Ns) and Gaussian-like PSF profiles, so the subtrac-
tions were conducted directly with their corresponding
references. However, the quality of the final-epoch LT image
did not satisfy the requirement for a reference image owing to
clouds; thus, the 2.0 m LCOGT reference images were adopted
instead.
After image subtraction, PSF photometry was performed on
the difference image with the Photutils package (Bradley et al.
2022) of Astropy (Astropy Collaboration et al. 2022) for the
optical and NIR data. The photometric data were calibrated
2
The Astrophysical Journal, 949:23 (16pp), 2023 May 20 Zhu et al.
using PS1 standards (Chambers et al. 2016) in the field of view
with Pan-STARRS color-term corrections (Tonry et al. 2012)
applied for the Johnson BV filters and Sloan Digital Sky Survey
griz filters, and the JHK NIR data were calibrated using Two
Micron All Sky Survey standards. Swift/UVOT and optical
griz photometric data were calibrated for the AB magnitude
system and BVJHK for the Vega magnitude system.
We reprocessed early-time optical data observed by the
above-mentioned instruments and presented by Bose et al.
(2018) for consistency according to the same procedure in this
work. The phases in the paper are given in the rest frame of
SN 2017egm, and the reference epoch is set to be the g-band
peak at JD= 2,457,926.3.
2.2. Swift/UVOT Photometry
UV images were obtained with the Neil Gehrels Swift
Observatory (hereafter Swift) with the Ultraviolet Optical
Telescope (UVOT). The Swift/UVOT data covered ∼140 days
of NUV data after discovery, which provided good sampling for a
blue bump before the NUV became undetectable (see Figure 1).
The Swift/UVOT photometry was measured with the UVOT-
SOURCE task in the Heasoft package using 5″ apertures after
subtracting the galaxy background using a Swift/UVOT image
taken on 2018 June 26 (+360 days). It was placed in the AB
magnitude system (Oke & Gunn 1983), adopting the revised zero-
points and sensitivity from Breeveld et al. (2011) as done in the
previous work analyzing the early-time UVOT data (Bose et al.
2018). All of the photometric data are presented in Table 1.
2.3. Archival MIR Data
The Wide-field Infrared Survey Explorer (WISE) conducted
a full-sky survey at 3.4, 4.6, 12, and 22 μm (labeled as W1–
W4, respectively) from 2010 February to August (Wright et al.
2010). The solid hydrogen cryogen used to cool the W3 and
W4 instruments was depleted later and placed in hibernation in
2011 February. WISE was reactivated and renamed NEO-
WISE-R in late 2013, using only W1 and W2, to search for
asteroids that could pose potential impact hazards to Earth
(Mainzer et al. 2014). Up to 2021 December, the NEOWISE
survey has visited every region of the sky 16 times every six
months, with an average of 12 single exposures within each
epoch (typically within one day).
Previous works show that the IR echoes (UV/optical light
absorbed and re-radiated in the IR by dust) of optical transients
are detectable on timescales of months to years while the
intraday variability is negligible (e.g., Jiang et al. 2016, 2021;
Sun et al. 2022). To acquire high-S/N measurements, we
choose to perform photometry on the time-resolved WISE/
NEOWISE coadds, and each coadd per band per epoch is
produced by stacking exposures within typically ∼1 day
intervals (Meisner et al. 2018).25
2.4. X-Ray Observations
Four epochs of X-ray observations of SN 2017egm (PI R.
Margutti) were obtained with the Chandra X-ray Observatory
on 2017 June 26 (24 ks exposure), 2017 September 17 (25 ks),
2017 November 9 (25 ks), and 2018 May 21 (24 ks); they
correspond to +4, +85, +136, and +323 days since the optical
peak. We processed the Chandra data with CIAO (v.4.14)
using the latest calibration files (v.4.9.6). The Level-2 event
files were recreated by the script of chandra_repro.
SN 2017egm is not detectable at any individual epoch. Then
we analyzed the stacked images of all four epochs with a total
exposure time of 96.9 ks, still yielding a nondetection.
Assuming a power-law spectrum with photon index Γ= 2
and Galactic column density NH= 1.08× 10
20 cm−2, the
inferred 3σ unabsorbed flux upper limit is 8.4× 10−16
erg s−1 cm−2 in the 0.5–10 keV band, corresponding to
1.9× 1039 erg s−1 at the distance of SN 2017egm. Alterna-
tively, if we assume a thermal emission, e.g., a blackbody with
temperature of kT= 1 keV, the inferred unabsorbed flux upper
limit is 7× 10−16 erg s−1 cm−2, corresponding to a luminosity
of 1.8× 1039 erg s−1.
The nondetection from the deep Chandra observations gives
the most stringent constraint on the X-ray emission of an
SLSN-I to date. Figure 2 shows this limit compared with a
large sample of X-ray observations of other SLSNe-I from
Margutti et al. (2018). The X-rays from SN 2017egm could be
heavily absorbed by CSM, i.e., a neutral hydrogen column
density of NH = 10
24 cm−2 will lead to a nondetection of
intrinsic X-ray luminosity of 1041 erg s−1.
2.5. Spectroscopic Observations
Optical spectra from +84 to +261 days were obtained using
the Kast Spectrograph mounted on the 3 m Shane telescope at
Lick Observatory (CA, USA; Miller & Stone 1993), the
Double Spectrograph for the Palomar 200 inch Hale telescope
(P200), the MODS1 multiobject double spectrographs mounted
at the Large Binocular Telescope Observatory (LBT), and the
OSIRIS instrument located in the Gran Telescopio CANAR-
IAS (GTC). We also obtained spectra of SN 2017egm in the
NIR at +105 days using the Aerospace Corporation’s Visible
and Near-Infrared Imaging Spectrograph (VNIRIS) mounted
on the Shane telescope and +143 days using the Spex medium-
resolution spectrograph (0.7–5.3 μm; Rayner et al. 2003) on the
NASA Infrared Telescope Facility (IRTF). All of the spectra
except for the one observed with LBT/MODS were taken at or
near the parallactic angle to minimize differential slit losses
caused by atmospheric dispersion (Filippenko 1982). Regard-
ing the LBT observation, the slit width was 1″ and the slit
position angle was set to the default 0° (north–south), which
corresponds to a parallactic angle of 146° at the start of
observation. However, due to a low airmass (∼1.0) and short
exposure time (600 s), the effect of differential atmospheric
refraction should be small (see Filippenko 1982). The LBT/
MODS data were processed with the modsCCDRed program
(Pogge 2019) and the MODSIDL pipeline (Croxall &
Pogge 2019). The raw measurements of SpeX data were
converted to absolute flux using the Spextools software
package (Vacca et al. 2003; Cushing et al. 2004). The VNIRIS
data were processed using software customized for VNIRIS
and written in the Interactive Data Language (Rudy et al.
2021). All of the other spectra were reduced and calibrated
using standard procedures in IRAF.26 See a summary of our
late-time spectra in Table 2.
25 https://portal.nersc.gov/project/cosmo/temp/ameisner/neo8
26 IRAF is distributed by the National Optical Astronomy Observatories,
which are operated by the Association of Universities for Research in
Astronomy, Inc. (AURA) under cooperative agreement with the National
Science Foundation.
3
The Astrophysical Journal, 949:23 (16pp), 2023 May 20 Zhu et al.
3. Analysis and Results
3.1. Photometric and Color Evolution
All light curves are shown in Figure 1, including our data
presented in this work along with early-phase data from Bose
et al. (2018) and late-phase gri data published by Hosseinzadeh
et al. (2022). The magnitudes are corrected for Galactic
extinction (RV = 3.1 and E(B− V )= 0.0097± 0.0005 mag;
Schlafly & Finkbeiner 2011) following Bose et al. (2018) and
K-corrections based on the optical spectra.
During ∼2 weeks following the season gap
(∼+30–100 days), the light curves are in a relatively smooth
decline, which appears to extend the post-peak linear decline in
magnitude prior to the gap. The post-peak decline rate of SN
2017egm is one of the slowest among SLSNe-I (Bose et al.
2018), but its peak luminosity (Mg,peak≈−21 mag) is close to
the mean for SLSNe-I, defying the empirical correlation
reported by Inserra & Smartt (2014) that the more slowly
declining SLSNe-I tend to be more luminous.
At ∼+115 days, the optical light curves show a small bump
in all bands, followed by a rapid decline (∼2 mag) at
∼+130–150 days. Meanwhile, the U-band light curve also
exhibits a bump followed by possibly an even more rapid drop
of NUV fluxes given the nondetection around +150 days. The
bump is reminiscent of that found in SN 2015bn (Nicholl et al.
2016), which was more pronounced in the bluer bands. A sharp
decline in all bands a few months after the peak is an expected
common feature for the ejecta–CSM interaction model, and this
occurs when the shock wave reaches the outer edge of the
extended envelope with no more material ahead, with no source
of radiation energy remaining (Sorokina et al. 2016). Interest-
ingly, there is a second bump at ∼+150–240 days, which is
most clearly seen in the g-band light curve. The second bump
appears nearly flat, and the variations are particularly small in
redder bands, suggesting an extra power supply, such as CSM
interaction.
As shown by Bose et al. (2018), the UV-to-optical color
(uvw2− r) evolution near the peak of SN 2017egm is similar to
that of Gaia16apd (Kangas et al. 2017), which starts with a very
blue color and becomes redder rapidly as the SN cools. By
contrast, the optical color evolution is much less dramatic (see
the top panel of Figure 3). We compare the g− r color
Figure 1. Multiband light curves of SN 2017egm from NUV to NIR with Galactic extinction corrected. The light curves are shifted vertically at various offsets for
different bands for display. The reference epoch is set to be the g-band peak at JD = 2,457,926.3. All optical light curves show a rapid descent during 120−150 days
and multiple bumps distinctly. The hollow triangles with downward arrows represent upper limits of the corresponding bands.
4
The Astrophysical Journal, 949:23 (16pp), 2023 May 20 Zhu et al.
evolution of SN 2017egm with that of other well-observed
SLSNe-I and find it is most similar to SN 2015bn (see the
bottom panel of Figure 3) during all available phases. They
both show less evolution than others. Their optical colors are
roughly constant near the peak with g− r≈ r− i≈−0.2 mag,
and they become gradually redder until ∼+100 days. After
∼+100 days, the g− r color curve of SN 2017egm appears to
flatten out, probably because of multiple bumps and the strong,
varying emission-line features in the corresponding wavelength
regimes.
3.2. Pseudo-bolometric Light Curve
The densely sampled multiband light curves of SN 2017egm
allow us to analyze its spectral energy distributions (SEDs) at
different epochs. We use the PhotoFit package (Soumagnac
et al. 2020) to fit the blackbody temperature (Tbb), luminosity
(Lbb), and effective radius (Rbb). The fitting results are given in
Table 3. The blackbody model fits the SEDs well near
maximum light. The observing cadence after +150 days in B
and V is sparse except for the g band; thus, we interpolate the g-
band light curve to match with the BV epochs. We require at
Table 1
Photometry of SN 2017egm
MJD Phasea B g V r i z Telescopeb Sourcec
(days) (mag) (mag) (mag) (mag) (mag) (mag) /Inst.
57,904.90 −20.28 ... ... ... ... ... 16.74 ± 0.04 LT Bose et al. (2018)
57,905.18 −20.01 ... ... 15.89 ± 0.02 ... ... ... PO Bose et al. (2018)
57,906.20 −19.02 15.82 ± 0.04 15.76 ± 0.03 15.81 ± 0.02 16.03 ± 0.02 16.21 ± 0.03 ... PO Bose et al. (2018)
K K K K K K K K K K
58,213.37 279.02 ... 20.72 ± 0.27 ... 21.63 ± 0.32 >20.96 ... KeplerCam Hosseinzadeh et al. (2022)
58,218.46 283.96 ... ... ... ... ... 19.29 ± 0.51 LT This work
58,221.20 286.62 ... >20.73 20.55 ± 0.21 ... ... ... PO This work
58,238.90 303.79 ... ... ... ... ... >21.45 AF This work
58,293.20 356.48 ... >21.48 >20.70 ... ... ... PO This work
MJD Phasea J H K Telescopeb Source
(days) (mag) (mag) (mag) /Inst.
57,906.94 −18.30 16.06 ± 0.06 16.07 ± 0.06 16.03 ± 0.06 NC Bose et al. (2018)
57,924.96 −0.82 14.87 ± 0.06 14.92 ± 0.06 15.03 ± 0.06 NC Bose et al. (2018)
57,937.90 11.74 14.86 ± 0.06 14.88 ± 0.06 14.91 ± 0.06 NC Bose et al. (2018)
58,043.24 113.95 16.13 ± 0.07 15.82 ± 0.06 15.72 ± 0.06 NC This work
58,122.26 190.62 17.72 ± 0.21 17.38 ± 0.08 17.13 ± 0.07 NC This work
58,195.62 261.80 19.23 ± 0.20 18.78 ± 0.14 ... UKIRT This work
MJD Phasea UVU UVW1 UVM2 UVW2 Telescopeb Source
(days) (mag) (mag) (mag) (mag) /Inst.
57,906.32 −18.90 15.48 ± 0.05 15.60 ± 0.03 15.67 ± 0.03 15.94 ± 0.03 UVOT Bose et al. (2018)
57,908.69 −16.60 15.38 ± 0.05 15.56 ± 0.03 15.66 ± 0.03 15.96 ± 0.03 UVOT Bose et al. (2018)
57,912.07 −13.32 15.18 ± 0.05 ... ... 15.82 ± 0.03 UVOT Bose et al. (2018)
K K K K K K K K K K
58,058.91 129.15 19.07 ± 0.29 21.48 ± 0.86 ... ... UVOT This work
58,083.36 152.88 ... >20.90 >21.34 >21.47 UVOT This work
58,096.30 165.43 >20.42 ... ... ... UVOT This work
Notes.
a Rest-frame time (days) relative to the epoch of the g-band peak brightness.
b The abbreviations of telescope/instrument used are as follows: LT—2.0 m Liverpool Telescope; PO—0.6 m telescopes of Post Observatory; LCOGT—Las
Cumbres Observatory Global Telescope Network; AF—ALFOSC mounted on 2.0 m NOT telescope; KeplerCam—KeplerCam mounted on 1.2 m telescope at Fred
Lawrence Whipple Observatory; NC—NotCAM IR imager on 2.0 m NOT telescope; UKIRT—NIR Wide-Field Camera mounted on the United Kingdom Infrared
Telescope; UVOT—Ultraviolet Optical Telescope on board the Swift satellite.
c We have reprocessed the data with the source of Bose et al. (2018) using the same procedure with this work.
(This table is available in its entirety in machine-readable form.)
Figure 2. X-ray observations of SN2017egm along with other SLSNe-I from
Margutti et al. (2018) spanning the phase range ∼10–2000 days since
explosion. Upper limits for SN 2017egm are shown as black diamonds, other
upper limits are shown as red circles, and detections or possible detections for
SCP 06F6 (Barbary et al. 2009), PTF12dam (Nicholl et al. 2013) and
ASASSN-15lh (Dong et al. 2016) are shown as black circles. Black horizontal
line with arrows is the SN2017egm average weighted by exposure time. The
explosion time of SN2017egm is estimated as the time of its first detection.
5
The Astrophysical Journal, 949:23 (16pp), 2023 May 20 Zhu et al.
least three photometry bands in the fitting. We do not use the
NIR JHK data in the single blackbody fitting, and their analysis
is shown in next section.
In Figure 4, we show the temporal variations of inferred
blackbody parameters and compare them with those of
PTF12dam (Nicholl et al. 2013), Gaia16apd (Kangas et al.
2017), and SN 2015bn (Nicholl et al. 2016). SN 2017egm
evolves in temperature like Gaia16apd but with smaller radii
and luminosities. The total observed radiation energy is close to
7.8× 1050 erg. Distinct from the smooth decline in the other
three, the bolometric light curve of SN 2017egm features
several slope changes, including high temperature initially with
a steady cooling, a small bump followed by a steep decline, and
then a late-time broad bump accompanied by rising temper-
ature. It is worth noting that the late-time bolometric light
curves at 150 days, which exhibited variations including a
broad bump, were synthesized without NUV photometry.
Thus, the derived blackbody parameters were less certain than
those at earlier phases. However, such a multitude of variations
with complicated characteristics suggests that it may have
undergone multiple physical processes, which can hardly be
explained by a simple magnetar or 56Ni decay model.
Recent work suggests that the undulations and late-stage
bumps appear to be a common phenomenon among SLSNe-I
(Inserra et al. 2017; Chen et al. 2023; Hosseinzadeh et al.
2022). There are two main interpretations. One attributes them
to the temporal variation of the central engine, which could be
modulated by photon diffusion in the ejecta, or the change in
opacity that causes the variations in photospheric emission (see
discussions in Hosseinzadeh et al. 2022 and references therein).
The other mechanism invokes the CSI model, in which the
ejecta–CSM interaction is an effective process to convert
mechanical energy into thermal emission. In the case of
SN 2017egm, the latter scenario seems to be preferable. First,
Wheeler et al. (2017) find that the linear rise and decline around
the sharp peak of SN 2017egm can be well explained by the
CSI model but are difficult to fit with the magnetar model, and
they also expected that the late-time light curves could
discriminate between the two models. The magnetar model
(Vurm & Metzger 2021; Hosseinzadeh et al. 2022) has
significant issues in fitting the bumps. We discuss the models
further in Section 4.
3.3. MIR Excess and Dust Emission
The NEOWISE survey visited the sky region of
SN 2017egm every six months since 2014 and acquired a total
of 14 epochs of observations. The images at the first
Table 2
Summary of Late-time Spectroscopic Observations of SN 2017egm
UT date JD − Phasea Exp. Time Airmass Telescope
2,458,000 (days) (s) /Instrument
2017-09-16.50 13.00 84.0 900 2.9 P200/DBSP
2017-09-27.54 24.04 94.8 2100 1.9 Shane/Kast
2017-10-08.50 35.00 105.5 720 1.5 Shane/AeroSpOpIR
2017-10-19.55 46.05 116.2 1200 1.3 Shane/Kast
2017-10-25.54 52.04 122.0 2100 1.3 Shane/Kast
2017-10-30.47 56.97 126.8 900 1.3 P200/DBSP
2017-10-30.54 57.04 126.9 3000 1.2 Shane/Kast
2017-11-16.58 74.08 143.4 2965 1.3 IRTF/SpeX
2017-11-21.50 79.00 148.2 1980 1.1 Shane/Kast
2017-12-12.48 99.98 169.0 3600 1.1 Shane/Kast
2017-12-18.50 106.00 174.4 3600 1.0 Shane/Kast
2017-12-28.43 115.93 184.0 3 × 1000 1.1 P200/DBSP
2018-01-13.50 132.00 199.6 3600 1.0 Shane/Kast
2018-03-14.29 191.79 257.6 600 1.0 LBT/MODS
2018-03-17.99 195.49 261.2 2 × 1800 1.1 GTC/OSIRIS
Note.
a Rest-frame time (days) relative to the epoch of the g-band peak brightness.
Figure 3. Optical color evolution of SN 2017egm. Top: evolution of a
selection of optical colors. Bottom: comparison of rest-frame g − r color with
that of other well-observed SLSNe-I.
6
The Astrophysical Journal, 949:23 (16pp), 2023 May 20 Zhu et al.
NEOWISE epoch are used as references since they were
observed securely before the SN explosion. We then perform
image subtraction with HOTPANTS and PSF photometry on the
difference images following Jiang et al. (2021). We take the
epochs displaying >3σ detection at the position of
SN 2017egm in at least one band as robust MIR excess (see
the light curve in Figure 5). The excesses are clearly detectable
at +137, +292, and +492 days after the g-band peak.
SN 2017egm remained luminous in the optical band at +137
and +260 days, indicating a non-negligible contribution from
the Rayleigh–Jeans tail of the optical blackbody. Therefore, we
try to fit the overall optical–NIR–MIR SEDs with double-
blackbody functions for these two epochs (see Figure 5). The
fitting results show that the MIR excesses are mainly
dominated by the blackbody component peaked in the MIR,
which can be naturally attributed to reprocessed dust emission.
The dust blackbody temperatures at the three epochs are
520± 70, 565± 83, and 827± 45 K with luminosities of
(3.54± 1.50), (3.89± 2.01), and (1.65± 0.40)× 1041 erg s−1,
respectively (see Table 4 for the fitting results). The last epoch
has a significantly bluer W1−W2 color and subsequently
higher inferred blackbody temperature than the previous two
epochs, suggesting that the dust might be heated up while the
SN luminosity decreased. Our measurements of the W1 and
W2 excesses are consistent with those found by Sun et al.
(2022), but they did not infer blackbody parameters of dust
emission owing to a lack of contemporaneous optical data.
The origin of dust in SLSNe remains poorly explored until
very recently. Chen et al. (2021) has concluded that the
prominent IR excess detected after +230 days in SN 2018bsz
can only be well explained by newly formed dust within the
SNe ejecta after careful analysis. Moreover, Sun et al. (2022)
performed a systematic study of the MIR light curves of 10
SLSNe at z< 0.12, including SN 2017egm, and suggested that
the MIR excess of SN 2017egm may also be explained by the
newly formed dust model like SN 2018bsz, owing to their
similarities. One notable characteristic that makes SN 2017egm
distinct from other slowly declining SLSNe-I is the rapid drop
in luminosity that occurred between +120 and +150 days,
which is reminiscent of the sudden drop starting at +98 days
observed in SN 2018bsz. However, as claimed by Chen et al.
(2021), the drop is yet likely due to the ejecta transitioning to
the nebular phase rather than dust extinction, although robust
Table 3
Best-fit Blackbody Parameters
Phasea Temperature Tbb Radius Rbb Luminosity Lbb
(days) (103 K) (1015 cm) ( -Log erg s10 1( ))
−19.02 17.36 ± 0.15 1.18 ± 0.01 43.95 ± 0.03
−18.05 17.03 ± 0.13 1.22 ± 0.01 43.95 ± 0.02
−17.08 16.45 ± 0.14 1.32 ± 0.02 43.96 ± 0.03
−16.11 16.24 ± 0.11 1.37 ± 0.01 43.97 ± 0.03
−15.39 16.13 ± 0.12 1.43 ± 0.02 43.99 ± 0.02
−10.29 15.29 ± 0.10 1.79 ± 0.02 44.10 ± 0.02
−8.35 14.86 ± 0.09 1.95 ± 0.02 44.12 ± 0.02
−7.38 14.53 ± 0.09 2.09 ± 0.02 44.14 ± 0.02
−6.40 14.39 ± 0.08 2.17 ± 0.02 44.16 ± 0.02
−5.42 14.77 ± 0.10 2.15 ± 0.03 44.20 ± 0.02
−4.45 14.53 ± 0.09 2.26 ± 0.03 44.21 ± 0.02
−1.55 13.63 ± 0.09 2.55 ± 0.03 44.20 ± 0.02
−0.57 13.40 ± 0.07 2.61 ± 0.02 44.19 ± 0.02
0.40 12.95 ± 0.06 2.70 ± 0.02 44.17 ± 0.01
1.37 12.73 ± 0.06 2.79 ± 0.03 44.16 ± 0.02
2.34 12.68 ± 0.06 2.71 ± 0.02 44.13 ± 0.01
3.31 12.12 ± 0.09 2.95 ± 0.05 44.13 ± 0.02
4.28 12.03 ± 0.06 2.91 ± 0.03 44.10 ± 0.02
5.26 11.80 ± 0.05 2.95 ± 0.03 44.08 ± 0.02
6.22 11.63 ± 0.06 2.94 ± 0.03 44.05 ± 0.02
7.19 11.34 ± 0.06 3.05 ± 0.03 44.04 ± 0.02
8.16 11.18 ± 0.05 3.07 ± 0.03 44.02 ± 0.02
9.13 11.02 ± 0.06 3.11 ± 0.03 44.01 ± 0.02
10.10 10.83 ± 0.05 3.19 ± 0.03 44.00 ± 0.02
11.07 10.79 ± 0.06 3.14 ± 0.03 43.98 ± 0.02
11.97 10.58 ± 0.06 3.22 ± 0.03 43.97 ± 0.01
12.04 10.57 ± 0.08 3.22 ± 0.05 43.96 ± 0.03
13.98 10.07 ± 0.18 3.42 ± 0.08 43.93 ± 0.06
15.91 10.24 ± 0.14 3.29 ± 0.08 43.92 ± 0.06
17.85 9.88 ± 0.13 3.38 ± 0.08 43.89 ± 0.05
19.79 9.86 ± 0.12 3.41 ± 0.07 43.89 ± 0.04
20.76 9.73 ± 0.11 3.49 ± 0.07 43.88 ± 0.05
21.73 9.72 ± 0.10 3.44 ± 0.06 43.87 ± 0.04
24.65 9.57 ± 0.08 3.47 ± 0.05 43.86 ± 0.04
27.56 9.41 ± 0.08 3.48 ± 0.10 43.83 ± 0.06
28.53 9.34 ± 0.07 3.46 ± 0.12 43.81 ± 0.07
29.50 8.94 ± 0.12 3.70 ± 0.14 43.79 ± 0.09
99.65 6.35 ± 0.10 3.51 ± 0.12 43.15 ± 0.04
103.55 5.97 ± 0.08 3.98 ± 0.14 43.16 ± 0.05
106.83 6.15 ± 0.13 3.39 ± 0.17 43.07 ± 0.07
111.30 6.11 ± 0.14 3.33 ± 0.19 43.04 ± 0.07
115.17 6.58 ± 0.13 2.78 ± 0.12 43.01 ± 0.04
121.96 6.20 ± 0.11 3.17 ± 0.13 43.03 ± 0.05
126.82 6.91 ± 0.10 2.38 ± 0.07 42.96 ± 0.03
128.75 6.59 ± 0.14 2.34 ± 0.11 42.87 ± 0.04
135.50 5.78 ± 0.34 2.17 ± 0.30 42.57 ± 0.16
137.45 6.45 ± 0.70 1.54 ± 0.35 42.47 ± 0.19
139.35 5.72 ± 0.51 1.87 ± 0.39 42.43 ± 0.22
140.41 5.81 ± 0.35 1.69 ± 0.22 42.36 ± 0.13
143.41 5.07 ± 0.35 1.92 ± 0.28 42.24 ± 0.12
144.38 5.36 ± 0.34 1.72 ± 0.22 42.24 ± 0.11
145.35 5.22 ± 0.42 1.68 ± 0.32 42.17 ± 0.20
148.26 4.78 ± 0.39 1.95 ± 0.40 42.15 ± 0.26
151.17 5.17 ± 0.55 1.61 ± 0.39 42.12 ± 0.22
152.13 6.30 ± 1.37 0.99 ± 0.43 42.04 ± 0.28
159.85 6.02 ± 1.23 1.42 ± 0.59 42.28 ± 0.29
165.63 7.19 ± 0.63 1.03 ± 0.17 42.31 ± 0.13
167.57 7.85 ± 1.06 0.82 ± 0.20 42.26 ± 0.21
168.54 7.44 ± 1.02 0.91 ± 0.22 42.26 ± 0.18
171.45 7.89 ± 0.93 0.82 ± 0.16 42.27 ± 0.21
204.45 9.46 ± 1.02 0.62 ± 0.11 42.34 ± 0.26
208.33 9.20 ± 1.06 0.63 ± 0.12 42.31 ± 0.28
211.00 7.32 ± 0.40 0.99 ± 0.09 42.30 ± 0.10
Table 3
(Continued)
Phasea Temperature Tbb Radius Rbb Luminosity Lbb
(days) (103 K) (1015 cm) ( -Log erg s10 1( ))
212.04 6.34 ± 0.48 1.26 ± 0.17 42.26 ± 0.12
213.21 7.96 ± 0.65 0.85 ± 0.11 42.31 ± 0.16
214.02 8.47 ± 0.76 0.72 ± 0.11 42.28 ± 0.17
215.04 5.58 ± 0.53 1.54 ± 0.30 42.22 ± 0.11
217.79 7.22 ± 1.40 0.96 ± 0.30 42.25 ± 0.28
225.75 7.67 ± 0.66 0.86 ± 0.14 42.26 ± 0.12
233.55 6.90 ± 1.36 0.89 ± 0.32 42.11 ± 0.23
248.99 7.19 ± 1.45 0.59 ± 0.26 41.83 ± 0.28
Note.
a Rest-frame time (days) relative to the epoch of the g-band peak brightness.
Data after +148 days were fitted without UV limit.
(This table is available in machine-readable form.)
7
The Astrophysical Journal, 949:23 (16pp), 2023 May 20 Zhu et al.
evidence for dust formation appears later (see also Pursiainen
et al. 2022). Actually, an IR echo of a proper pre-existing
spherical dust shell can equally reproduce the observed MIR
emission. Note that Rbb,IR is around (7.3–8.2)× 10
16 cm, which
is much larger than Rbb,Opt (∼10
15 cm) in our double-
blackbody fits at both epochs. The IR emission comes much
further out than the region of ejecta, and thus the excess may be
dominated by pre-existing dust emission. It is worthwhile to
note that the inferred dust temperature shows a rising trend
albeit with large errors, which is unusual in the echo scenario.
Nevertheless, an asymmetric dust bubble shell, i.e., the shell on
the far side is closer to the SNe than that in our line of sight,
can produce such a trend. In spite of this, we cannot definitively
exclude a dust formation process at later epochs, particularly
considering a higher blackbody temperature and smaller radius
for the third MIR epoch (+492 days, see Table 4), which may
represent the dust formed during ejecta–CSM collision. In fact,
the newly formed and pre-existing dust could work
simultaneously, as has been demonstrated in SN 2006jc
(Mattila et al. 2008) and some other SNe II (e.g., Pozzo et al.
2004; Meikle et al. 2007).
3.4. Spectroscopic Analysis
The optical spectra are presented chronologically in Figure 6,
with notable O I, Ca II, and Mg I] emission lines marked. For
∼200 days after the peak, its blue continuum faded rapidly, and
in contrast, the emission-line features evolve slowly. We
identify the lines by comparing with other SLSNe-I (Nicholl
et al. 2016, 2019; Kangas et al. 2017; Quimby et al. 2018).
Two NIR spectra are presented in Figure 7. There is a strong
feature around the region of He I λ10830 at +105 days. There
was controversy in the past over the identification of He I
λ10830 in SLSNe-I spectra. The proposed He I λ10830 in the
spectra of SN 2012il (Inserra et al. 2013) was later found to be
a broad emission feature redshifted by ∼1500 km s−1 relative
to λ10830, and may instead be blueshifted nebular Paγ
emission (Quimby et al. 2018). In our case, the feature around
Figure 5.MIR light curves of SN 2017egm and the double-blackbody fitting of
the UV–optical–IR SEDs. Top panel: W1 (3.4 μm, blue diamonds) and W2
(4.6 μm, red diamonds) light curves of SN 2017egm measured from the
NEOWISE survey, in which the triangles mark the 3σ upper limits. Middle and
bottom panels: SED fitting at +137 and +260 days. Two blackbody
components are required to account for the overall SED, which is peaked at
optical and MIR bands, respectively.
(The data used to create this figure are available.)
Figure 4. Evolution of the blackbody radius (top panel), temperature (middle
panel), and luminosity (bottom panel) of SN 2017egm. The evolution of three
other well-observed SLSNe-I (PTF12dam (red squares), SN2015bn (green
triangles), and Gaia16apd (orange diamonds)) is overplotted for comparison.
The open symbols denote that the blackbody fitting is performed without NUV
photometry.
Table 4
Blackbody Fitting Results for Epochs with MIR Data
Phasea TIR,bb RIR,bb LIR,bb TOpt,bb ROpt,bb LOpt,bb
(days) (K) (1015 cm) (1041 erg s−1) (K) (1015 cm) (1041 erg s−1)
+137 520 ± 70 82.1 ± 12.0 3.54 ± 1.50 7100 ± 500 1.2 ± 0.2 25.30 ± 9.70
+292 565 ± 83 73.1 ± 14.7 3.89 ± 2.01 4800 ± 420 1.5 ± 0.5 5.51 ± 2.08
+492 827 ± 45 22.2 ± 2.5 1.65 ± 0.40 L L L
Note.
a Rest-frame time (days) relative to the epoch of the g-band peak brightness.
8
The Astrophysical Journal, 949:23 (16pp), 2023 May 20 Zhu et al.
10830Å at +105 days also shows a strong P Cygni profile with
a velocity of ∼−8000 km s−1, which is similar to SN 2019hge,
the only other SLSN-I with He I 1.08 μm, so it cannot be a
nebular Paγ line. The nondetection of He I at ∼2.05 μm might
be simply due to the low S/N, since it is generally much
weaker than He I λ10830 (Yan et al. 2020). Interestingly, the
He I feature faded rapidly with the detection of a weak emission
line at +143 days. On the other hand, it is worthwhile to note
Figure 6. Rest-frame spectral evolution of SN 2017egm. Strong O I , Ca II, and Mg I] λ4571 features are marked by different colors at the top. The ⊕ symbol marks
the positions of the strongest telluric absorption bands, which have been masked. Each spectrum is labeled by the instrument employed and the rest-frame phase from
maximum light. The data used to create this figure are made available digitally.
(The data used to create this figure are available.)
9
The Astrophysical Journal, 949:23 (16pp), 2023 May 20 Zhu et al.
that there is no significant He I λ10830 in the early-time NIR
spectra (−2.7 and −0.5 days) according to Bose et al. (2018),
indicating that the He I could be short-lived.
Yan et al. (2020) reported six He-rich SLSNe-Ib from the
ZTF Phase-I SLSN-I sample, suggesting that the key to
identifying He-rich events is the presence of multiple He
features. We cross-check our optical spectra for He features. In
Figure 8, we compare the spectra of SN 2017egm at ∼+100
days with the well-studied SLSN-Ic SN 2015bn in detail. Four
He I absorption lines (He I λλ5876, 6678, 7065, 7281) are
detected in SN 2017egm with a velocity of −7000 km s−1. He I
λ5876 is the strongest of the four, with a deep absorption
feature that could be caused by possible contamination from
Na I λ5890. The other three lines are weaker, while the He I
λ7065 absorption is well isolated. Since the four He I lines are
detected in multiple spectra of SN 2017egm between +95 and
+127 days, the existence of helium is firmly established.
Further comparisons between SN 2017egm and other known
SLSNe-Ib are presented in the next section.
3.5. Spectroscopic Evolution
Apart from the strong W-shaped O II absorption features at
3700–4500Å, the early-stage spectra of SN 2017egm are
mostly devoid of other prominent features, while heavily
blended metallic lines start to appear at ∼+26 days (Bose et al.
2018; see their Figure 6 for details). Some common line
features in SLSNe-Ic are identified, such as a few Fe II lines and
Na I D at 4900–5600Å, the Ca II λ3750, the Ca II λλ3934,
3969 H&K doublet, the Ca II λλ8498, 8542, 8662 NIR triplet,
and O I λ7774. At +84 days, the spectrum is full of strong,
blended lines and P Cygni profiles, with also a strong Mg I]
λ4571 line, which usually emerges at 30 days post-peak
(Inserra et al. 2013; Nicholl et al. 2016). Other prominent
features identified are also shown in Figure 8 and compared
with those in SN 2015bn. The Fe II lines between 4200 and
5000Å, including contributions from Mg II, become stronger in
the high-quality +95 days spectrum, in which the previously
dominant O II lines become weaker and finally disappear as the
ejecta cool and these species recombine. At the same time, the
strengths of [O I] λλ6300, 6364, Na I λ5890, and the calcium
lines increase substantially. The Ca II λ3750 is weaker than the
prominent Ca II λλ3934, 3969 H&K doublet in both early-time
(+26 days; Bose et al. 2018) and late-time (up to +127 days,
see Figure 6) spectra. There is an unidentified line at ∼6100Å,
and a similar line seen for SN 2015bn was interpreted as a
possibly detached, high-velocity component of Si II (Nicholl
et al. 2016). The asymmetric O I λ7774 line profile may be
contaminated by Mg II λλ7877, 7896. By +148 days, the blue
continuum has faded, and thus the features at the blue end
(<6000Å; e.g., Mg I] λ4571, Ca II λλ3969, 3750, and a few
Fe II lines), which were previously overwhelmed by the
continuum, become quite visible. It is notable that the inferred
velocity of He I λ10830 (−8000 km s−1) is larger than that of
the optical He I lines (−7000 km s−1). A similar velocity
difference was also found in SN 2019hge (Yan et al. 2020; see
their Figure 1 for optical He I lines at −6000 km s−1 and their
Figure 2 for He I λ10830 at −8000 km s−1), which was
attributed to either the difference in excitation levels of the NIR
and optical lines or the He I λ10830 being blended with a
possible blueshifted C I λ10691 line.
The He I λ10830 fades concurrently in the +143 days NIR
spectra (see Figure 7) as the He I λλ7065, 7281 appears to fade
in the +148 days optical spectra (see Figure 9 for the zoomed-
in view of the spectra normalized to [Ca II] λ7300). Curiously,
the apparent fading in He I happens concurrently when the light
curve drops rapidly to a local minimum. Interestingly, He I
slightly emerges again at +200 days, coinciding with the
second bump in the light curve. If the He I features and the light
curve indeed coevolve, it could indicate that the source of
nonthermal emission needed to produce the He I features (see,
e.g., Yan et al. 2020) may also power the optical emission. We
note that the above-mentioned interpretation is complicated by
the possible blending of the [Ca II] λ7300 line with oxygen
features (Nicholl et al. 2019). We also note that the He I λ5876
feature is always present in most optical spectra, though it is
heavily blended with Na I λ5890. Most line features can be
well fitted by a broad Gaussian and the line velocities decline
slowly. For instance, the velocity of [Ca II] λ7300 drops from
10,900 km s−1 (+95 days) to 8000 km s−1 (+200 days).
The features described above are common for slowly
declining Type Ic SLSNe, so we try to compare the spectral
evolution of SN 2017egm with that of PTF12dam and
SN 2015bn in Figure 10. However, SN 2017egm differs greatly
from the other two objects at early phases; for example, the O II
absorption features disappear faster and the O I λ7774 is
weaker near the time of peak brightness. Their later evolution is
similar except for the presence of strong absorption features at
5876Å as well as other weaker He I lines in SN 2017egm.
Yan et al. (2020) classified SN 2019hge as an SLSN-Ib
based on its NIR spectrum and the other five events as good
SLSN-Ib/IIb candidates. Here we compare the optical spectra
of SN 2017egm with those of SN 2019hge and PTF10hgi
(Figure 11). SN 2019hge is redder with strong Si II λ6355, C II
λ6580, and He I at an early phase, while these features are
weaker or absent in SN 2017egm at similar epochs. Similar to
SN 2017egm, the He I in SN 2019hge appeared most promi-
nently when the temperature was relatively low
(Tbb∼ 7000–8000 K) but were suppressed at higher temper-
ature in the earlier phases (Yan et al. 2020). The He I
Figure 7. NIR spectra at two epochs. The prominent He I line λ10830 is
detected at +105 days. Strong telluric and unreliable regions in the spectra are
masked out. The dashed lines mark the rest-frame wavelengths of He I lines at
10830 and 20586 Å. The data used to create this figure are made available
digitally.
(The data used to create this figure are available.)
10
The Astrophysical Journal, 949:23 (16pp), 2023 May 20 Zhu et al.
absorption features of SN 2017egm, with a velocity of around
−7000 km s−1, are similar to both SN 2019hge and PTF10hgi
at late phases. Other nebular-phase features of those objects are
similar to those in SLSNe-Ic. Meanwhile, the light curves of
the SLSNe-Ib in the ZTF sample (see Figure 12) show blue
bumps, strong undulations, and mostly relatively low peak
luminosities in comparison with SLSNe-Ic (Chen et al. 2023).
Interestingly, with a peak luminosity Mr≈−21 mag,
SN 2017egm is the most luminous among the known SLSNe-
Ib, while the peak magnitudes Mr of other SLSN-Ib/IIb are all
dimmer than −20.5. As discussed earlier, the sharp peak of
SN 2017egm may suggest a constant-density CSM shell
(s= 0), as previously noted by Chatzopoulos et al. (2013)
and Wheeler et al. (2017). Chen et al. (2023) also concluded
that three SLSNe-Ib strongly prefer the CSM+Ni model over
the magnetar model. However, none of the SLSNe-Ib shows
narrow nebular emission-line features indicative of ejecta–
CSM interaction like in SNe-Ibn or SNe-IIn.
4. Summary and Discussion
We have conducted a comprehensive photometric and
spectroscopic analysis of SN 2017egm, a hydrogen-poor SLSN
that exploded in the rare environment of a massive and metal-
rich galaxy, and we identify it as the closest SLSN-Ib found to
date. The main results obtained from our analysis are
summarized below.
1. The long-term multiband light curves of SN 2017egm
spanning ∼300 days have a sharp peak, rapid decline,
and multiple late-time bumps, but the object can be
broadly categorized into the subclass of slowly evolving
SLSNe-I according to the luminosity and color evolution
near peak brightness.
2. The spectra of SN 2017egm resemble those of SLSNe-Ic
except for the presence of strong He I λ10830 emission
and multiple optical He I absorption lines. Given the
presence of these helium features, we classify
SN 2017egm as an SLSN-Ib, which is a rare yet
accumulating subpopulation of SLSNe-I (Yan et al.
2020).
3. SN 2017egm is one of a handful of SLSNe-I with a
significant IR excess (LIR≈ 10
8 Le) indicative of dust
echoes from pre-existing dust or emission from newly
formed dust (Sun et al. 2022).
4. SN 2017egm was not detectable in each of the four-epoch
Chandra observations or even in the stacked images. This
yields the tightest constraint on the X-ray emission of an
SLSN-I to date, with a 3σ upper limit of 1.9× 1039
erg s−1 in the 0.5–10 keV range.
Figure 8. Optical spectrum of SN 2017egm at +95 days post-maximum and NIR spectrum at +105 days. The optical spectrum of SLSN-Ic SN 2015bn at +106 days
is plotted for comparison. The identified lines are labeled by different colors at the top. The dashed lines mark the locations of He I λλ5876, 6678, 7065, and 7281,
with a velocity of −7000 km s−1 (black) and He I λ10830 with a velocity of −8000 km s−1 (red).
11
The Astrophysical Journal, 949:23 (16pp), 2023 May 20 Zhu et al.
The observational features discussed above serve as the most
complete characterization of SN 2017egm to date in terms of
the unprecedented temporal and wavelength coverage. Before
this work, several models were proposed to account for the
observations of SN 2017egm, and its light curve seems more
likely to be explained by CSM interaction (Wheeler et al. 2017;
Hosseinzadeh et al. 2022). On the basis of our more complete
characterization, we try to further explore this scenario with
TigerFit (Chatzopoulos et al. 2016; Wheeler et al. 2017)
assuming a hybrid model by combining radioactive decay with
a constant density (camrads0) or a wind-like density
(csmrads2) profile (Chatzopoulos et al. 2013). The best-fit
model has an ejecta mass of 10.7Me and a
56Ni mass of
0.15Me, colliding with a CSM with mass of 2.7Me.
So far, no simple models, including our best-fit one or others
(Wheeler et al. 2017; Hosseinzadeh et al. 2022), can reproduce
the bolometric light curve of SN 2017egm satisfactorily (see
Figure 13). The model powered solely by radioactive decay is
not viable because it requires substantially more 56Ni than that
allowed by the ejecta mass. The magnetar-only model needs a
rather smaller ejecta mass, but it fails to reproduce the peak
light-curve profile (Wheeler et al. 2017; see their Table 1 for
the parameters) and late-time fluctuations. The model will
release a total energy significantly exceeding the observed
value, even counting in the reprocessed energy radiated in the
MIR, which is yet too low to compensate for the sharp drop in
optical luminosity (see Figure 13). The CSM model with a
constant-density shell could fit the sharp peak well, but a single
ejecta–CSM interaction model is inadequate to produce the
multiple bumps seen in SN 2017egm.
Liu et al. (2018) has shown that a multiple ejecta–CSM
interaction model can fit the undulating bolometric light curve
of iPTF15esb, which displays two prominent peaks and a
plateau. In their scenario, the appearance of late-time broad Hα
emission can be naturally interpreted as the ejecta running into
a neutral H shell (Yan et al. 2017a). Massive stars could
experience mass losses in the form of eruptions in the final
stage of their lives (Vink 2022), and the progenitors of SNe
could sometimes expel two or more shells and/or episodic
winds. Furthermore, the steep decline (∼0.1 mag day−1) in
SN 2017egm is expected in the CSM interaction scenario
(Sorokina et al. 2016).
Thus, these observations of light curves may indicate the
existence of extended and multiple CSM shells or CSM clumps
at different radii around SN 2017egm. The ejecta–CSM
interactions could be the sources of the nonthermal excitation
needed to produce the He I lines (see, e.g., Yan et al. 2020), so
this might explain why the He I features appear to coevolve
with the late-time light curve as shown in Figure 9. Similar
coevolution of light curve and spectra was seen in the case of
hydrogen-rich SN 1996al, which was attributed to CSM
interaction (Benetti et al. 2016). One major sign from
interaction with the CSM is broad emission lines from shocked
photoionized ejecta (see, e.g., Blinnikov 2017), and the
“hump” near He I λ7065 in spectra from +174 to +200 days
(see the bottom panel of Figure 9) may suggest the existence of
such features blended by the strong Ca II lines. Detailed
modeling will be needed to further model the data and explore
the underlying mechanism, but this is beyond the scope of
this work.
A common issue with the interpretation in terms of ejecta–
CSM interaction for SLSNe-I is the lack of narrow emission
lines expected from the photoionized CSM. A hydrogen-
deficient CSM can lead to nondetection of Balmer lines, but the
lack of emission from other elements could remain proble-
matic. However, Chatzopoulos et al. (2013) suggested that the
absence of narrow emission lines in the optical spectra of
SLSNe-I does not necessarily argue against CSM interaction.
In that scenario, emission due to CSM interaction may appear
as a blue continuum, which has been observed in the +200
days spectra of SN 2017egm with an H-deficient CSM. They
argued that the intermediate-width emission such as [O I]
λλ6300, 6364 and O I λ7774 features found in SN 2007bi
could be evidence for CSM interaction (Gal-Yam et al. 2009).
In fact, Nicholl et al. (2019) noticed a suggested narrow core of
[O I] λ6300 in the late-time spectra of SN 2017egm, and their
latest spectrum at 353 days had the clearest indication of such a
core. Our late-time spectra (∼200 days) show similar profiles
(see the bottom panel of Figure 14), albeit wider than the
profile in their 353 days spectrum. Interestingly, like the He I
features, the [O I] feature also changes significantly during the
drastic light-curve evolution. Its primary peak was considerably
redshifted before +122 days (see the top panel Figure 14), and
the line profiles of [O I] λλ6300, 6364 change considerably
after the light curve drops rapidly. Since detailed, non-LTE
radiation hydrodynamics models of H-poor CSM interaction
models are not available, the nature of these lines and their
formation sites await further investigation.
We have shown that the color and spectral evolution of
SN 2017egm are broadly similar to those of other SLSNe-I
such as SN 2015bn and also SLSN-Ib SN 2019hge except for
He I lines, yet they have substantial differences in light-curve
Figure 9. The He I λλ7065, 7281 evolution of SN 2017egm. All spectra, with
continuum subtracted, have been smoothed using the Savitzky–Golay filtering
algorithm and normalized to [Ca II] λ7300. All spectra are overlaid onto the
original spectral data, which are in the same color. The dashed vertical lines
mark He I λλ7065, 7281, with a velocity of −7000 km s−1.
12
The Astrophysical Journal, 949:23 (16pp), 2023 May 20 Zhu et al.
evolution. Previous spectroscopic observations of SLSNe-I
have not captured the drastic spectral change in the bumpy
phase and thus offered us few clues to the light-curve evolution
(Hosseinzadeh et al. 2022). However, our high-S/N spectra
obtained at different bumpy phases of SN 2017egm suggest
clearly not only the variations in continuum flux but also
significant changes in spectroscopic profiles, potentially
shedding new light on the formation of lines in the ejecta.
In addition, the X-ray and MIR observations may offer
useful insights into SN 2017egm. SN 2017egm holds the
tightest X-ray upper limit among SLSNe-I, which might
provide strict constraints on the subparsec environment
and properties of central engines in SLSNe-I (see, e.g.,
Chevalier & Fransson 2003; Levan et al. 2013; Huang &
Li 2018; Margutti et al. 2018; Moriya et al. 2018; Vurm &
Metzger 2021). Our deepest X-ray observations in SLSNe-I
rule out the densest environments typical of eruptions of
luminous blue variable and Type IIn SNe (Margutti et al.
2018), but the parameter space of SN 2017egm (LX< 10
40
erg s−1 and X-ray-to-optical luminosity ratio LX/LOpt 10−3)
is almost an entirely uncharted territory to explore. Andrews &
Smith (2018) invoked the possibility that strong CSM
interaction is hidden behind the ejecta over a wide range of
viewing angles to explain the absence of narrow emission lines
and X-ray emissions in the case of the peculiar iPTF14hls, and
it may be of interest to consider similar scenarios for SLSNe-Ib
Figure 10. The comparison of optical spectra between SN 2017egm and some other slowly declining Type Ic SLSNe at similar phases, including PTF12dam (Nicholl
et al. 2013) and SN 2015bn (Nicholl et al. 2016). Their spectroscopic evolution is nearly identical.
13
The Astrophysical Journal, 949:23 (16pp), 2023 May 20 Zhu et al.
in the context of CSM interactions. On the other hand, the
detection of an unambiguous dust emission has not been
reported in SLSNe until very recently (Chen et al. 2021; Sun
et al. 2022), and SN 2017egm belongs to the population with
clear evidence of pre-existing dust or dust formation. The MIR
echo agrees nicely with the ejecta–CSM interaction model, and
reveals a new promising method to probe the subparsec
environment of SLSNe.
In the upcoming golden era of wide-field time-domain
surveys, such as the Legacy Survey of Space and Time (Ivezić
et al. 2019) with the Rubin Observatory and the Wide-Field
Survey Telescope (Lin et al. 2022), SLSNe are expected to
soon be discovered at a high rate. SN 2017egm, as a rare SLSN
studied in great detail, will serve as a valuable case for
comparison and be re-evaluated statistically in the context of
large samples.
We thank the anonymous referee for carefully reading our
manuscript and providing valuable comments. This work is
supported by the National Natural Science Foundation of China
(grants 11833007, 12133005, 12073025, 12192221) and the
Fundamental Research Funds for the Central Universities
(WK3440000006). S.D. acknowledges support from the
Xplorer Prize. J.Z and N.J. gratefully acknowledge the support
of Cyrus Chun Ying Tang Foundations. S.M. acknowledges
support from the Magnus Ehrnrooth Foundation and the Vilho,
Figure 11. Comparison of optical spectra of SN 2017egm and other well-studied SLSNe-Ib/IIb, SN 2019hge (Yan et al. 2020) and PTF10hgi (Quimby et al. 2018), at
similar phases. The dashed lines mark the locations of He I λλ5876, 6678, 7065, and 7281, with a velocity of −7000 km s−1.
14
The Astrophysical Journal, 949:23 (16pp), 2023 May 20 Zhu et al.
Yrjö, and Kalle Väisälä Foundation. A.P., N.E.R., S.B., and E.
C. are partially supported by the PRIN-INAF 2022 project
“Shedding light on the nature of gap transients: from the
observations to the models.” A.V.F.ʼs supernova group at UC
Berkeley is grateful for financial support from the Christopher
R. Redlich Fund and numerous individual donors. S.M.
acknowledges support from the Academy of Finland project
350458.
Based in part on observations made with the Nordic Optical
Telescope, owned in collaboration by the University of Turku
and Aarhus University, and operated jointly by Aarhus
University, the University of Turku, and the University of
Oslo (respectively representing Denmark, Finland, and Nor-
way), the University of Iceland, and Stockholm University, at
the Observatorio del Roque de los Muchachos, La Palma,
Spain, of the Instituto de Astrofisica de Canarias.
This research uses data obtained through the Telescope
Access Program (TAP). Observations with the Hale telescope
at Palomar Observatory were obtained as part of an agreement
between the National Astronomical Observatories, Chinese
Academy of Sciences, and the California Institute of Technol-
ogy. We thank the Swift PI Brad Cenko, the Observation Duty
Scientists, and the science planners for approving and
executing our Swift/UVOT observations.
A major upgrade of the Kast spectrograph on the Shane 3 m
telescope at Lick Observatory, led by Brad Holden, was made
possible through generous gifts from the Heising–Simons
Foundation, William and Marina Kast, and the University of
California Observatories. Research at Lick Observatory is
partially supported by a generous gift from Google. This work
is partly based on observations made with the Gran Telescopio
Canarias (GTC), installed at the Spanish Observatorio del
Roque de los Muchachos of the Instituto de Astrofísica de
Canarias, on the island of La Palma.
The LBT is an international collaboration among institutions
in the United States, Italy, and Germany. LBT Corporation
partners are The University of Arizona on behalf of the Arizona
Board of Regents; Istituto Nazionale di Astrofisica, Italy; LBT
Beteiligungsgesellschaft, Germany, representing the Max-
Planck Society, The Leibniz Institute for Astrophysics
Potsdam, and Heidelberg University; The Ohio State
Figure 12. The r-band light curves of five SLSNe-Ib/IIb with well-sampled
observations.
Figure 13. The bolometric light curve is plotted together with the best-fit
csmrads0 model and a previous CSM0/magnetar model from Wheeler et al.
(2017). The inset shows a zoomed-in view around the peak.
Figure 14. Observed [O I] λλ6300, 6364 features. Top panel: the considerably
redshifted features of [O I] λλ6300, 6364 before 122 days. Bottom panel: a
suggested narrow core near 6300 Å in late-time spectra was previously reported
by Nicholl et al. (2019; a spectrum at +353 days from Nicholl et al. 2019 is
marked by * in the legend). Our late-time spectra show similar profiles, while
the widths of the profiles in our +174 and +200 days spectra are broader than
that of the +353 days spectrum from Nicholl et al. (2019). The spectra are
smoothed using the Savitzky–Golay filtering algorithm and have been
continuum-subtracted and converted to velocity space. The strong host-galaxy
Hα emission was masked.
15
The Astrophysical Journal, 949:23 (16pp), 2023 May 20 Zhu et al.
University, representing OSU, University of Notre Dame,
University of Minnesota, and University of Virginia. This
paper made use of the modsCCDRed data reduction code
developed in part with funds provided by NSF grants AST-
9987045 and AST-1108693. This paper made use of the
modsIDL spectral data reduction pipeline developed in part
with funds provided by NSF grant AST-1108693 and a
generous gift from OSU Astronomy alumnus David G. Price
through the Price Fellowship in Astronomical Instrumentation.
ORCID iDs
Jiazheng Zhu https://orcid.org/0000-0003-3824-9496
Ning Jiang https://orcid.org/0000-0002-7152-3621
Subo Dong https://orcid.org/0000-0002-1027-0990
Alexei V. Filippenko https://orcid.org/0000-0003-
3460-0103
Richard J. Rudy https://orcid.org/0000-0003-3096-759X
A. Pastorello https://orcid.org/0000-0002-7259-4624
Christopher Ashall https://orcid.org/0000-0002-5221-7557
Subhash Bose https://orcid.org/0000-0003-3529-3854
D. Bersier https://orcid.org/0000-0001-7485-3020
Stefano Benetti https://orcid.org/0000-0002-3256-0016
Thomas G. Brink https://orcid.org/0000-0001-5955-2502
Ping Chen https://orcid.org/0000-0003-0853-6427
Liming Dou https://orcid.org/0000-0002-4757-8622
N. Elias-Rosa https://orcid.org/0000-0002-1381-9125
Peter Lundqvist https://orcid.org/0000-0002-3664-8082
Seppo Mattila https://orcid.org/0000-0001-7497-2994
Michael L. Sitko https://orcid.org/0000-0003-1799-1755
Auni Somero https://orcid.org/0000-0001-6566-9192
M. D. Stritzinger https://orcid.org/0000-0002-5571-1833
Tinggui Wang https://orcid.org/0000-0002-1517-6792
Peter J. Brown https://orcid.org/0000-0001-6272-5507
E. Cappellaro https://orcid.org/0000-0001-5008-8619
Morgan Fraser https://orcid.org/0000-0003-2191-1674
Erkki Kankare https://orcid.org/0000-0001-8257-3512
Simon Prentice https://orcid.org/0000-0003-0486-6242
Tapio Pursimo https://orcid.org/0000-0002-5578-9219
T. M. Reynolds https://orcid.org/0000-0002-1022-6463
WeiKang Zheng https://orcid.org/0000-0002-2636-6508
References
Anderson, J. P., Pessi, P. J., Dessart, L., et al. 2018, A&A, 620, A67
Andrews, J. E., & Smith, N. 2018, MNRAS, 477, 74
Astropy Collaboration, Price-Whelan, A. M., Lim, P. L., et al. 2022, ApJ,
935, 167
Barbary, K., Dawson, K. S., Tokita, K., et al. 2009, ApJ, 690, 1358
Becker, A. 2015, HOTPANTS: High Order Transform of PSF and Template
Subtraction, Astrophysics Source Code Library, ascl:1504.004
Benetti, S., Chugai, N. N., Utrobin, V. P., et al. 2016, MNRAS, 456, 3296
Blinnikov, S. 2017, Handbook of Supernovae (Berlin: Springer), 843
Bose, S., Dong, S., Pastorello, A., et al. 2018, ApJ, 853, 57
Bradley, L., Sipőcz, B., Robitaille, T., et al. 2022, astropy/photutils: 1.5.0,
v1.5.0, Zenodo, doi:10.5281/zenodo.6825092
Breeveld, A. A., Landsman, W., Holland, S. T., et al. 2011, in AIP Conf. Proc.
1358, Gamma Ray Bursts 2010, 373
Chambers, K. C., Magnier, E. A., Metcalfe, N., et al. 2016, arXiv:1612.05560
Chatzopoulos, E., Wheeler, J. C., Vinko, J., et al. 2013, ApJ, 773, 76
Chatzopoulos, E., Wheeler, J. C., Vinko, J., et al. 2016, ApJ, 828, 94
Chen, T.-W., Brennan, S. J., Wesson, R., et al. 2021, arXiv:2109.07942
Chen, T.-W., Schady, P., Xiao, L., et al. 2017, ApJL, 849, L4
Chen, Z. H., Yan, L., Kangas, T., et al. 2023, ApJ, 943, 42
Chevalier, R. A., & Fransson, C. 2003, in Supernovae and Gamma-Ray
Bursters, ed. K. Weiler (Berlin: Springer), 171
Chevalier, R. A., & Irwin, C. M. 2011, ApJL, 729, L6
Croxall, K. V., & Pogge, R. W. 2019, rwpogge/modsIDL: modsIDL Binocular
Release, v1.0, Zenodo, doi:10.5281/zenodo.2561424
Cushing, M. C., Vacca, W. D., & Rayner, J. T. 2004, PASP, 116, 362
Dong, S., Bose, S., Chen, P., et al. 2017, ATel, 10498, 1
Dong, Subo, Shappee, B. J., Prieto, J. L., et al. 2016, Sci, 351, 257
Filippenko, A. V. 1982, PASP, 94, 715
Filippenko, A. V. 1997, ARA&A, 35, 309
Gal-Yam, A. 2012, Sci, 337, 927
Gal-Yam, A. 2019, ARA&A, 57, 305
Gal-Yam, A., Mazzali, P., Ofek, E. O., et al. 2009, Natur, 462, 624
Heger, A., & Woosley, S. E. 2002, ApJ, 567, 532
Hosseinzadeh, G., Berger, E., Metzger, B. D., et al. 2022, ApJ, 933, 14
Huang, Y., & Li, Z. 2018, ApJ, 859, 123
Inserra, C., Nicholl, M., Chen, T.-W., et al. 2017, MNRAS, 468, 4642
Inserra, C., Prajs, S., Gutierrez, C. P., et al. 2018, ApJ, 854, 175
Inserra, C., & Smartt, S. J. 2014, ApJ, 796, 87
Inserra, C., Smartt, S. J., Jerkstrand, A., et al. 2013, ApJ, 770, 128
Ivezić, Ž., Kahn, S. M., Tyson, J. A., et al. 2019, ApJ, 873, 111
Izzo, L., Thöne, C. C., García-Benito, R., et al. 2018, A&A, 610, A11
Jiang, N., Dou, L., Wang, T., et al. 2016, ApJL, 828, L14
Jiang, N., Wang, T., Hu, X., et al. 2021, ApJ, 911, 31
Kangas, T., Blagorodnova, N., Mattila, S., et al. 2017, MNRAS, 469, 1246
Kasen, D. 2010, ApJ, 708, 1025
Levan, A. J., Read, A. M., Metzger, B. D., et al. 2013, ApJ, 771, 136
Lin, Z., Jiang, N., & Kong, X. 2022, MNRAS, 513, 2422
Liu, L.-D., Wang, L.-J., Wang, S.-Q., et al. 2018, ApJ, 856, 59
Lunnan, R., Chornock, R., Berger, E., et al. 2018, ApJ, 852, 81
Lunnan, R., Yan, L., Perley, D. A., et al. 2020, ApJ, 901, 61
Mainzer, A., Bauer, J., Cutri, R. M., et al. 2014, ApJ, 792, 30
Margutti, R., Chornock, R., Metzger, B. D., et al. 2018, ApJ, 864, 45
Mattila, S., Meikle, W. P. S., Lundqvist, P., et al. 2008, MNRAS, 389, 141
Meikle, W. P. S., Mattila, S., Pastorello, A., et al. 2007, ApJ, 665, 608
Meisner, A. M., Lang, D., & Schlegel, D. J. 2018, AJ, 156, 69
Miller, J. S., & Stone, R. P. S. 1993, Lick Observatory Technical Reports, 66
Lick Observatory, Santa Cruz, CA
Moriya, T. J., Sorokina, E. I., & Chevalier, R. A. 2018, SSRv, 214, 59
Nicholl, M., Berger, E., Blanchard, P. K., et al. 2019, ApJ, 871, 102
Nicholl, M., Berger, E., Margutti, R., et al. 2017a, ApJL, 845, L8
Nicholl, M., Berger, E., Smartt, S. J., et al. 2016, ApJ, 826, 39
Nicholl, M., Guillochon, J., & Berger, E. 2017b, ApJ, 850, 55
Nicholl, M., Smartt, S. J., Jerkstrand, A., et al. 2013, Natur, 502, 346
Oke, J. B., & Gunn, J. E. 1983, ApJ, 266, 713
Pastorello, A., Smartt, S. J., Botticella, M. T., et al. 2010, ApJL, 724, L16
Planck Collaboration, Ade, P. A. R., Aghanim, N., et al. 2016, A&A,
594, A13
Pogge, R. 2019, rwpogge/modsCCDRed 2.0, v2.0, Zenodo, doi:10.5281/
zenodo.2550741
Pozzo, M., Meikle, W. P. S., Fassia, A., et al. 2004, MNRAS, 352, 457
Pursiainen, M., Leloudas, G., Paraskeva, E., et al. 2022, A&A, 666, A30
Quimby, R. M., Aldering, G., Wheeler, J. C., et al. 2007, ApJL, 668, L99
Quimby, R. M., De Cia, A., Gal-Yam, A., et al. 2018, ApJ, 855, 2
Quimby, R. M., Kulkarni, S. R., Kasliwal, M. M., et al. 2011, Natur,
474, 487
Rayner, J. T., Toomey, D. W., Onaka, P. M., et al. 2003, PASP, 115, 362
Rudy, R. J., Subasavage, J. P., Mauerhan, J. C., et al. 2021, AJ, 161, 291
Schlafly, E. F., & Finkbeiner, D. P. 2011, ApJ, 737, 103
Sorokina, E., Blinnikov, S., Nomoto, K., et al. 2016, ApJ, 829, 17
Soumagnac, M. T., Ganot, N., Irani, I., et al. 2020, ApJ, 902, 6
Sun, L., Xiao, L., & Li, G. 2022, MNRAS, 513, 4057
Tonry, J. L., Stubbs, C. W., Lykke, K. R., et al. 2012, ApJ, 750, 99
Vacca, W. D., Cushing, M. C., & Rayner, J. T. 2003, PASP, 115, 389
Vink, J. S. 2022, ARA&A, 60, 203
Vurm, I., & Metzger, B. D. 2021, ApJ, 917, 77
Wheeler, J. C., Chatzopoulos, E., Vinkó, J., et al. 2017, ApJL, 851, L14
Woosley, S. E. 2010, ApJL, 719, L204
Wright, E. L., Eisenhardt, P. R. M., Mainzer, A. K., et al. 2010, AJ, 140, 1868
Yan, L., Lunnan, R., Perley, D. A., et al. 2017a, ApJ, 848, 6
Yan, L., Perley, D. A., Schulze, S., et al. 2020, ApJL, 902, L8
Yan, L., Quimby, R., Gal-Yam, A., et al. 2017b, ApJ, 840, 57
16
The Astrophysical Journal, 949:23 (16pp), 2023 May 20 Zhu et al.