SN 2023gfo: A Peculiar Type IIP Supernova with High Luminosity and Normal Plateau Duration

Abstract: We present near-infrared (NIR) and optical observations of the highly reddened Type IIP supernova (SN) 2023gfo in the nearby galaxy NGC 4995 ($d = 26.4 \pm 3.2$ Mpc), which reached a high peak luminosity of $M_V = -18.6$ mag. The SN was initially detected as a faint red event, with $B-V = 0.8$ mag at the beginning of the plateau phase. By comparison with template, we estimate a total extinction of $A_V = 2.1$ mag. After correcting for this extinction, we derive a peak quasi-bolometric luminosity of $(5.9 \pm 1.5)\times10<sup>{42}$ erg s$<sup>{-1}$, placing this event among the most luminous SNe IIP, while its plateau duration remains within the normal range. The early-phase optical spectrum exhibits a P-Cygni profile of H$α$, with a broad absorption of $V$ = $13{,}800$ km s$<sup>{-1}$, which is among the highest observed for SNe IIP at comparable epochs. The high luminosity and the normal plateau duration suggest that this event represents an outlier. Applying an analytical model, we infer an unusually large progenitor radius. This may indicate that the progenitor experienced an extreme energy injection from the core to the envelope shortly before explosion, resulting in a substantially inflated radius. While ejecta-circumstellar matter (CSM) interaction could in principle account for the high luminosity, we find no observational evidence supporting strong interaction.

• The paper finds that SN 2023gfo reaches M_V = −18.6 mag and a peak pseudo-bolometric luminosity of (5.9 ± 1.5) × 10^42 erg s−1 while maintaining a normal 108 ± 8 day plateau.
• The paper combines optical and near-infrared photometry with early spectroscopy, measuring an Hα velocity of approximately 13,800 km s−1 and deriving robust extinction through multi-band color comparisons.
• The paper infers a near-canonical explosion energy but an unusually extended progenitor radius of roughly 2,600–3,900 R☉, favoring envelope inflation from late-stage energy injection while acknowledging that weak circumstellar interaction remains possible.

Overview and observational context

SN 2023gfo was discovered by ATLAS on 20 April 2023 (MJD 60054.2) in NGC 4995, with a pre-discovery detection at 19.2 mag on MJD 60051.2 and a last non-detection on MJD 60050.0; the authors adopt MJD 60050.6 as the explosion epoch. The host redshift is $z = 0.00594$, and the distance is taken from the Tully–Fisher method as $26.4 \pm 3.2$ Mpc ($m-M = 32.06 \pm 0.42$ mag), consistent with the NED Hubble-flow distance corrected for Virgo, Great Attractor, and Shapley contributions. The dataset comprises $BVRI$ photometry from HOWPol on the 1.5-m Kanata telescope, $griz$ photometry from the GROWTH-India Telescope, simultaneous $JHK_s$ imaging with kSIRIUS on the Kagoshima 1-m telescope, and optical spectra at $t = 4$, 5, and 11 d from KOOLS-IFU on the 3.8-m Seimei telescope.

Extinction correction

The event is heavily reddened: the host-galaxy Na I D equivalent width of $\sim$2.4 Å indicates substantial extinction, though the Poznanski et al. relation is unreliable at this EW regime. The authors therefore rely on color-curve methods under the assumption that plateau-phase photospheric temperatures—and hence intrinsic colors—are nearly uniform among SNe IIP because the plateau is governed by hydrogen recombination. The $B-V$ color at 30 d gives $E(B-V) = 0.52 \pm 0.19$ mag ($26.4 \pm 3.2$0 mag), but a more robust estimate comes from $26.4 \pm 3.2$1 color excesses against template colors built from SNe 2003hn, 2008in, 2012A, 2012ec, and 2013ej, yielding $26.4 \pm 3.2$2 mag for $26.4 \pm 3.2$3. Because NIR extinction is small ($26.4 \pm 3.2$4 mag), the derived luminosity is insensitive to $26.4 \pm 3.2$5 uncertainties. Independent support comes from the extinction-corrected Balmer decrement matching Case B recombination. This robustness matters: the exceptional luminosity claimed below rests directly on this extinction scale.

Photometric properties

The $26.4 \pm 3.2$6-band light curve peaks at 13.5 mag at $26.4 \pm 3.2$7 d, and the ATLAS $26.4 \pm 3.2$8-band evolution lies within the CSP sample distribution. Following Anderson et al.'s methodology, the plateau duration is $26.4 \pm 3.2$9 d—close to the mode of the CSP distribution. After correcting for extinction and distance, SN 2023gfo reaches $m-M = 32.06 \pm 0.42$0 mag, exceeding even the most luminous events in the Anderson et al. sample ($m-M = 32.06 \pm 0.42$1 mag). In the NIR it is brighter than all comparison objects, peaking at $m-M = 32.06 \pm 0.42$2 mag. The combination of extreme luminosity with a normal plateau duration places SN 2023gfo, like SN 2009kf, at the luminous tail while remaining typical in duration—an outlier relative to the standard luminosity–duration correlations of SNe IIP.

Bolometric light curve and spectroscopy

Integrating the extinction-corrected $m-M = 32.06 \pm 0.42$3 SEDs yields a peak pseudo-bolometric luminosity of $m-M = 32.06 \pm 0.42$4 erg s$m-M = 32.06 \pm 0.42$5 at $m-M = 32.06 \pm 0.42$6 d, significantly above comparison objects at similar phases. The $m-M = 32.06 \pm 0.42$7 d spectrum shows a normal Type IIP configuration: P-Cygni H$m-M = 32.06 \pm 0.42$8 with absorption minimum at $m-M = 32.06 \pm 0.42$913,800 km s$BVRI$0—among the highest velocities reported at comparable epochs—plus broad H$BVRI$1 and He I $BVRI$2 absorption. Narrow host-galaxy lines yield a site metallicity of 8.7 via the Pettini–Pagel method, consistent with the average for Type II SNe.

Physical parameters from scaling relations

Applying the Popov and Kasen & Woosley analytic scalings—with radiation-hydrodynamic calculations by Ouchi & Maeda supporting their validity even for expanded-envelope models—the authors anchor on SN 2009kf as primary reference and cross-check against SNe 1999em, 2012A, 2012aw, and 2017eaw:

Reference SN	$BVRI$3 (foe)	$BVRI$4 ($BVRI$5)	$BVRI$6 ($BVRI$7)
SN 2009kf	3.0	19	3500
SN 1999em	1.8	15	3900
SN 2012A	0.6	11	3600
SN 2012aw	2.0	13	3000
SN 2017eaw	2.2	7	2600
Mean	1.9 ± 0.9	13 ± 4.7	3320 ± 508

The velocity at $BVRI$8 d ($BVRI$97000 km s$griz$0) was not measured directly but extrapolated by scaling the exponential velocity evolution of SN 2017eaw to the observed H$griz$1 minimum—a modeling step that introduces systematic uncertainty into all three inferred quantities. Nevertheless, the progenitor radius remains systematically large ($griz$2–$griz$3) regardless of reference object, whereas the explosion energy is only modestly above canonical values. The large radius is therefore presented as the robust inference, implying that the progenitor was substantially inflated relative to typical red supergiants.

Interpretation: inflated envelope versus CSM interaction

The favored interpretation follows Ouchi & Maeda's treatment of SN 2009kf: pre-supernova energy injection from the core into the envelope—possibly via convectively excited waves during final nuclear burning—inflates the radius, boosting plateau luminosity without proportionally extending its duration. The alternative, ejecta–CSM interaction, is argued against on three observational grounds: no flash-ionized emission lines in the $griz$4 d spectrum (only a blue continuum); broad ejecta absorption lines at $griz$5 d, which interaction-powered events lack because continuum dilution suppresses absorption; and no rapid rise or bumpy structure in the blue-band light curves, with color evolution consistent with normal SNe IIP. The paper concedes that the physical mechanism of energy deposition remains uncertain and that the inflation interpretation "is not unique."

Limitations and open questions

Several caveats bear directly on the conclusions. First, the extinction estimate assumes uniform intrinsic plateau colors among SNe IIP; if SN 2023gfo's intrinsic colors deviate from the template set, the luminosity scale shifts. Second, the scaling analysis relies on a single mid-plateau velocity point extrapolated from another object, and analytic relations collapse the explosion into a small set of observables—the residual dispersion across reference objects ($griz$6) reflects both reference diversity and these simplifications. Third, no spectra exist between $griz$7 d and later phases, so velocity evolution during the plateau is unconstrained. Fourth, the absence of detected flash features does not strictly exclude a tenuous or already-swept-up confined CSM. The paper leaves open whether radiation-hydrodynamic models with inflated envelopes can simultaneously reproduce the high luminosity, normal plateau duration, and high expansion velocity, and what observational discriminants would separate inflated-envelope progenitors from weak or hidden CSM interaction.

Conclusion

SN 2023gfo combines a peak pseudo-bolometric luminosity of $griz$8 erg s$griz$9, $JHK_s$0 mag, an H$JHK_s$1 expansion velocity of 13,800 km s$JHK_s$2, and a normal 108-d plateau. Scaling analyses consistently indicate a near-canonical explosion energy but an unusually large progenitor radius of roughly $JHK_s$3. Together with SN 2009kf, it constitutes a second candidate example of a Type IIP event whose progenitor envelope was inflated by late-stage energy injection, although the mechanism and the exclusion of CSM interaction remain open issues requiring dedicated hydrodynamic modeling.