Optical Observations of the Nearby Type Ia Supernova 2021hpr

The $BVRI$-band photometric observations of SN 2021hpr were carried out by using the 60cm telescope at XingLong Observatory, National Astronomical Observatories, Chinese Academy of Sciences (NAOC). The telescope is equipped with a 2048 $\times$ 2048 pixel CCD, giving a field of view (FOV) of $37.4^{\prime}\times 37.4^{\prime}$ (pixel scale $\sim 1^{\prime\prime}.1\text{pixel}^{-1}$). The photometric data were obtained from 2021 April 3 (two days after discovery) to 2021 June 3, covering the peak phase and including 26 observing nights. The observations of SN 2021hpr followed the sequence of $B$, $V$, $R$, and $I$ filters and usually were repeated three times in each observation night. Since SN 2021hpr is located in one spiral arm of its host galaxy, the contamination due to the galaxy should be removed carefully. The template images of the host galaxy were obtained about 240 days after the SN explosion, when the SN was faint.

The ccdproc package (Craig et al., 2021) was employed to do the bias subtraction and flat field correction of the SN images. Generally, the combined bias frames and combined twilight flat fields taken on the same night were adopted. Cosmic rays were removed using LACosmic (van Dokkum, 2001). After these processes, astrometric calibrations of SN images were conducted by using Astrometry.net (Lang et al., 2012). Then, the SN images observed each day with the same filter (usually three images) were combined using the reproject package (Robitaille et al., 2020) and the ccdproc package. The template images obtained on 2021 December 24 were subtracted from the SN images after aligning. The point-spread functions (PSFs) of these two images were matched by adopting a Gaussian convolution, and the fluxes of the template images were scaled to the SN image by utilizing the fluxes of the same field stars in these two images.

Aperture photometry was performed on the template-subtracted images to acquire the instrumental magnitude of SN 2021hpr and of the local reference stars. These reference stars were used to calibrate the instrument magnitude to the standard Johnson $BV$ (Johnson et al., 1966) and Kron$-$Cousins $RI$ (Cousins, 1980) systems. The $BVRI$ magnitudes of the six reference stars (see Figure 1) are given in Table 2 of Biscardi et al. (2012). Here, we give the final photometry result of SN 2021hpr in Table 1.

The magnitudes errors take into account the Poisson noise of the target signal, the noise from image combination, the noise from the photometric calibration, and the noise of the sky background.

The optical spectra of SN 2021hpr were obtained with the Beijing Faint Object Spectrograph and Camera (BFOSC) attached to the 2.16m telescope at XingLong Observatory, NAOC (Fan et al., 2016). A grism of G4 + 385LP with a wavelength range of $3700$ to $8800~{}\text{\AA}$ and a slit width of $2^{\prime\prime}.3$ was used for all the observations. The spectral resolution is about 20 $\text{\AA}$, determined from the sky emission lines.

The spectroscopic observations were carried out from 2021 April 3 to June 20, covering the peak phase, and a total of six spectra were obtained. The spectra of the SN and a standard star were obtained at a similar airmass for flux calibration. The log of the spectroscopic observations is given in Table 2. The spectral data were reduced using standard IRAF routines, including bias subtraction, flat correction, and cosmic-rays removal. The SN 2021hpr spectra were wavelength calibrated using observations of an FeAr lamp taken in the dome on the same night. Additionally, a spectrum of the nucleus of NGC 3147 was observed on 2022 February 1 using BFOSC with the same grism and slit in order to analyze the host galaxy of SN 2021hpr, and it will be discussed in section 4.4.

The optical light curves of SN 2021hpr in the $BVRI$ bands are shown in Figure 2, covering the phases from $-$14.37 to +46.67 days relative to the $B$-band maximum. The photometric evolution of SN 2021hpr is consistent with that of typical normal SNe Ia. A plateau is evident in the $R$-band light curve from t $\sim$ +18 to $\sim$ +25 days relative to peak luminosity, while a secondary hump is seen in the $I$-band light curve. The $B$-band magnitudes around maximum light were fitted using a second-order polynomial to estimate the $B$-band peak magnitude, which is $14.017\pm 0.017$ mag on MJD $59,321.862\pm 0.450$. The $V$-band peak magnitude is $14.091\pm 0.014$ on MJD $59,323.220\pm 0.733$, which is about 1.4 days later. The peak magnitudes and corresponding MJDs of the $BVRI$ bands are calculated and listed in Table 3. The post-maximum magnitude decline rate $\Delta m_{15}(B)$; (Phillips, 1993) is estimated by SNooPy (Burns et al., 2011) to be $0.949\pm 0.019$ mag. It is slightly lower than the typical value of 1.1 mag (Phillips et al., 1999), but still in the range for SNe Ia of 0.75$-$1.94 (Krisciunas et al., 2003).

The rise time represents the time from explosion to maximum luminosity, and is dependent on the progenitors of SNe Ia (Leibundgut & Pinto, 1992). Assuming SN 2021hpr exploded like an expanding fireball, with a constant temperature and velocity, its luminosity or flux $f$ should be proportional to the second power of time (Riess et al., 1999). The rise time $t_{rise}$ can be obtained from the relation:

where $t$ represents the phase relative to the $B$-band maximum and $\alpha$ describes the velocity of the rising flux. This relationship is consistent with the SNe light-curve patterns of some sky surveys (i.e., Riess et al., 1999; Goldhaber et al., 2001; Conley et al., 2006; Garg et al., 2007; Hayden et al., 2010; Ganeshalingam et al., 2011; Firth et al., 2015) and of some well-sampled individual SNe Ia (i.e., SN 2011fe, Nugent et al., 2011).

Based on Equation (1), the rise time of SN 2021hpr can be derived as $16.424\pm 0.078$ days by fitting the data from the first observation day to $\sim-$8 days. The data taken on 2021 April 3 were not included for data fitting due to the large errors.

The spectral sequence of SN 2021hpr is shown in Figure 3. A total of six optical low-resolution spectra cover the phases from $-$14.37 days to +63.68 days relative to the $B$-band maximum. The redshift of the host galaxy was corrected for in all spectra. The first spectrum of SN 2021hpr obtained at t $\sim-$14.37 days is the earliest one among all publicly available spectra.

In the first spectrum of SN 2021hpr, the absorption lines are mainly those of intermediate-mass elements, such as Mg ii, Si ii, S ii, and Ca ii. The most distinct features characterized by Si ii $\lambda$6355 and the Ca ii NIR triplet (8498, 8542, and 8662 $\text{\AA}$) are broad and strong with a very large velocity, and they can be clearly identified near 5900 $\text{\AA}$ and 7800 $\text{\AA}$, respectively. The blended features of Fe ii $\lambda$4404 and Mg ii $\lambda$4481 can be found near 4200 $\text{\AA}$, and the broad absorption feature near 4700 $\text{\AA}$ could be caused by Fe ii $\lambda$5018, Si ii $\lambda$5051 and Fe iii $\lambda$5129. The weak feature near 7400 $\text{\AA}$ should be O i $\lambda$7774. No apparent Na I D doublet lines could be identified in the spectrum, which indicates that the interstellar extinction can be ignored.

At $t\sim-4.31$ days, the characteristic of Si ii $\lambda$4130 feature appeared in the spectrum, and the absorption features of Fe ii $\lambda$4404, Mg ii $\lambda$4481, Si iii $\lambda$4560, Fe ii $\lambda$5018, Si ii $\lambda$5051 and Fe iii $\lambda$5129 became more obvious and more recognizable. The ”W” shape consisting of S ii $\lambda$5468 and S ii $\lambda\lambda$5612,5654 emerged. Si ii $\lambda$5972 should be responsible for the absorption line near 5700 $\text{\AA}$. The expansion velocity of Si ii $\lambda$6355 decreased rapidly. The HV component can be distinguished from the NV component of the Ca ii NIR triplet, of which the velocity became smaller.

Two weeks after the maximum luminosity, at $t\sim 13.6$ days, the absorption feature of Si iii $\lambda$4560 became weaker, and the ”W” shape almost disappeared. Near 5700 $\text{\AA}$, a strong absorption feature could be identified as Na I D $\lambda$5893, while the absorption feature of Si ii $\lambda$5972 was barely visible. The expansion velocity of the Ca ii NIR triplet decreased further.

From the third week to the second month since peak luminosity, according to the spectra observed at t $\sim$ 22.6, $\sim$ 46.7, and $\sim$ 63.7 days, the iron-peak element features became stronger and gradually dominated the spectrum. The absorption feature of Si ii $\lambda$6355 became weaker, while the Fe ii lines gradually appeared.

In Figure 4, the $BVRI$-band light curves of SN 2021hpr are plotted with those of other SNe Ia, which are ASASSN-14bd (Brown et al., 2014), SN 2017cbv (Wang et al., 2020), SN 2003du (Anupama et al., 2005; Stanishev et al., 2007; Hicken et al., 2009; Ganeshalingam et al., 2010; Silverman et al., 2012), SN 2005cf (Pastorello et al., 2007; Hicken et al., 2009; Ganeshalingam et al., 2010; Silverman et al., 2012; Brown et al., 2014), SN 2009ig (Brown et al., 2012; Hicken et al., 2012; Silverman et al., 2012; Brown et al., 2014; Stahl et al., 2019), SN 2011fe (Richmond & Smith, 2012; Silverman et al., 2012; Tsvetkov et al., 2013; Munari et al., 2013; Brown et al., 2014; Stahl et al., 2019), and SN 2018oh (Li et al., 2019). The first two SNe Ia have a flux excess in their early phase, while the others are normal SNe Ia. The light curves of all samples are normalized to their peaks. It can be seen that the $B$-band light curve of SN 2021hpr has a similar shape to that of other normal SNe Ia in the phase before the maximum luminosity. The decline rate $\Delta m_{15}(B)$ of SN 2021hpr (see Section 3.1) is close to the $0.96\pm 0.02$ mag of SN 2008oh (Li et al., 2019), $0.89\pm 0.02$ mag of SN 2009ig (Foley et al., 2012), and $0.990\pm 0.013$ mag of SN 2017cbv (Wang et al., 2020). The decline rate is slower than that of SN 2011fe, which has a typical value for the decline rate of $1.103\pm 0.035$ mag, while its luminosity is brighter. After $t\sim$ +30 days, the $B$-band luminosity is relatively brighter than those of other SNe Ia during the same phase. The $VR$-band light curves of SN 2021hpr follow the evolution of other normal SNe Ia. The second maximum in the $I$ band of SN 2021hpr is slightly stronger than those of other SNe Ia. Furthermore, ASASSN-14bd (in $BV$-band) and SN 2017cbv are brighter than the others before reaching the maximum.

In Figure 5, $B-V$, $V-R$, and $V-I$ color curves of SN 2021hpr are compared with those of the other SNe Ia in Section 4.1, except for ASASSN-14bd because it lacks data in the $RI$-band. All the color curves have been corrected for the Galactic and host galaxy reddening. As for SN 2021hpr, the Galactic extinction is estimated to be $A_{B}=0.088$ mag, $A_{V}=0.067$ mag, $A_{R}=0.053$ mag and $A_{I}=0.037$ mag on the basis of the dust map of the Milky Way (Schlafly & Finkbeiner, 2011). The host galaxy extinction of SN 2021hpr is almost negligible, which will be discussed in Section 4.4.

From $t\sim-15$ to +20 days, the $B-V$ color evolution of SN 2021hpr shows a resemblance to most of the selected SNe, while after $t\sim$ +20 days the color is bluer than most of the other supernovae. SN 2021hpr reaches the bluest $B-V$ color of $-$0.06 mag at $t\sim-5$ days, then gradually turns into its reddest color of $\sim 1.0$ mag at $t\sim+30$ days. SN 2017cbv is an SN Ia with early excess. Its $B-V$ color before the $B$-band maximum is bluer than that of most of the normal SNe, including SN 2021hpr, which suggests that SN 2021hpr may not have a bump feature in early phase. Before reaching maximum luminosity, the $V-R$ color curve of SN 2021hpr has a plateau phase at $\sim 0.25$ mag, which is similar to that of SN 2017cbv. However, the latter starts to decline after t $\sim-7$ days. The $V-R$ color of SN 2021hpr reaches the minimum value of approximately $-0.2$ mag, and undergoes rapid reddening until it reaches the reddest color of $\sim 0.33$ mag at $t\sim+30$ days. The $V-I$ color of SN 2021hpr is similar to that of SN 2011fe. SN 2021hpr reaches its bluest $V-I$ ($\sim-0.5$ mag) at $t\sim+10$ days and its reddest $V-I$ ($\sim 0.8$ mag) at $t\sim+30$ days. In addition, the $V-I$ color of SN 2021hpr is redder than that of most of the selected SNe after $t\sim 20$ days.

In Figure 6, we compare the spectra of SN 2021hpr with that of 2011fe at four different epochs (t $\approx$ $-$14, $-$4, +13 and +70 days relative to the $B$-band maximum). All the spectra have been corrected for redshift. In addition, the spectra of SN 2021hpr before or near peak luminosity downloaded from the Transient Name Server (TNS) website are also added into comparison, and are shown in Figure 6a and Figure 6b. Furthermore, we obtained the velocity of Si ii $\lambda$6355 of SN 2021hpr by Gaussian fitting.

At t $\sim-14$ days, the spectra of SN 2021hpr obtained by 2.16 m telescope are similar to the spectra from TNS, but shifted nearly half a day earlier (see Figure 6a). The expansion velocity of Si ii $\lambda$6355 of the former reaches $\sim$ 20,440 km s^-1, which is slightly higher than that in the latter spectrum from TNS. SN 2021hpr shows a very high velocity for the Ca ii NIR triplet, which is nearly 28,000 km s^-1. The velocities of these two features are much higher than those of SN 2011fe at this phase. A weak absorption feature of C ii $\lambda$6580 is seen on the redder side of Si ii $\lambda$6355 in the spectrum of SN 2011fe, while no obvious carbon feature can be recognized in the spectrum of SN 2021hpr. At least 30$\%$ of SNe would show lines of carbon in their spectra before the luminosity reaches maximum (Thomas et al., 2011).

About four days before maximum (Figure 6b), the Si ii $\lambda$6355 velocity of SN 2021hpr decreased more rapidly to $\sim$ 12,420 km s^-1. The velocity gradient from $t\sim-14.37$ to $-4.31$ days is about 800 km s^-1 day^-1, which confirms that SN 2021hpr belongs to the HVG SNe Ia group. Even so, this velocity is still higher than that of SN 2011fe and other NV SNe Ia, whose velocity is about 11,800 km s^-1 (Wang et al., 2009). This indicates that SN 2021hpr should be a transitional object between the HV and Normal groups, which is also confirmed by the slower $B$-band decline rate (see 3.1). An HV component of the Ca ii NIR triplet absorption profile can be found in the spectrum of SN 2021hpr, which is similar to that of SN 2011fe. The velocity range of this absorption feature is between 14,000 and 20,000 km s^-1, which is higher than that of SN 2011fe. The blended feature composed of Fe ii $\lambda$4404 and Mg ii $\lambda$4481 seems stronger in the spectrum of SN 201hpr than that in that of SN 2011fe.

Figure 6c displays the spectra two weeks after maximum luminosity. The Si ii and Ca ii absorption features still remain, while the ”W” shape of S ii has almost disappeared. The Si ii $\lambda$6355 velocity of SN 2021hpr is also slightly higher than that of SN 2011fe. Figure 6d shows the spectra two months after maximum luminosity. SN 2021hpr and SN 2011fe show comparable absorption features in their spectra, and the Fe features became dominant. The Ca ii NIR triplet absorption feature is still very strong in both spectra.

NGC 3147 is a face-on spiral galaxy and classified as SA(rs)bc (Blanc et al., 2013) with a redshift of 0.00935 and a size of about $3^{\prime}.9\times 3^{\prime}.4$. Its apparent magnitude in the $V$ band is 10.61 mag (Gil de Paz et al., 2007). Panessa & Bassani (2002) classified it as a type II Seyfert galaxy. However, a recent study discovered very broad emission lines in its nucleus (Bianchi et al., 2019), suggesting that it is a low accretion rate type I active galactic nucleus with a massive black hole.

We observed the nucleus of NGC 3147 with BFOSC on 2022 February 1, as shown in Figure 7. The H$\alpha$ region of the spectrum is shown in Figure 8. The ratio of H$\alpha$ and [N ii] $\lambda$6584 suggests that NGC 3147 could be classified as a low-ionization nuclear emission-line region (LINER).

NGC 3147 is also an SNe factory in which six SNe have been observed over the past 50 yr (1972-2022): SN 1972H (Patat et al., 1997), SN 1997 (Jha et al., 2006), SN 2006gi (Duszanowicz, 2006), SN 2008fv (Tsvetkov & Elenin, 2010), SN 2021do, and SN 2021hpr. Four of them are SNe Ia, one is an SN Ib, and one is an SN Ic. In Table 4, we summarize the SN type and equatorial coordinates (J2000) of these SNe, whose locations are shown in Figure 9. Interestingly, the four SNe Ia are all located in the galactic disk, while the Type Ic SN 2021do is closest to the galactic nucleus, and the Type Ib SN 2006gi is the farthest from the galactic center, in the outskirt of the galactic disk.

According to the statistics of Thöne et al. (2009), there are more than 16 galaxies in which SNe have been detected more than four times. The highest record is still held by NGC 6946, where 10 SNe have been observed in recent decades. It is followed by NGC 4303 with eight SNe, NGC 3690 with seven SNe, and then NGC 3147 with six SNe. The SN explosion rate in NGC 3147 is high if compared to the average of about three per galaxy per century. Using radio data, Thöne et al. (2009) estimates the SN explosion rate of NGC 3147 as 0.18 SNe per year, with 0.06 per year for Type Ia, 0.09 per year for Type Ibc, and 0.03 per year for Type II. On average, three SNe Ia, 4.5 SNe Ibc, and 1.5 SNe II will explode in NGC 3147 every 50 yr. Therefore, the observed explosion rate is slightly higher than that expected for SNe Ia in NGC 3147, but much lower for SNe Ibc and SNe II.

Furthermore, we examined the metallicity of the surroundings of SN 2021hpr. Figure 10 shows the spectrum combined with spectra of three star-formation regions within the distance of SN 2021hpr of 30^′′. The flux ratio of decomposed H$\alpha$ and H$\beta$ is estimated to be $\sim 2.74$. Therefore, the host galaxy extinction of SN 2021hpr is negligible. The spectrum near 6600 $\text{\AA}$ displayed in Figure 10 is shown in Figure 11, where it exhibits features of [N ii] $\lambda$6584 and H$\alpha$ $\lambda$6563. Since the two emission lines are blended, we use double Gaussian fitting to derive the flux ratio as [N ii] $\lambda$6584/H$\alpha$ $\lambda$6563 $\approx$ 0.33, corresponding to the metallicity of 12+log(O/H) $\approx$ 8.648, comparable to the solar metallicity of 8.69 (Asplund et al., 2021). This is consistent with the suggestion by Pan (2020) that HV SNe might be present in massive galaxies with metal-rich environments.

Based on the redshift of NGC 3147, we can obtain the the distance modulus of $m-M=33.31~{}mag$.²²2Given by the NASA/IPAC Extragalactic Database (NASA/IPAC Extragalactic Database (2019), NED), assuming $H_{0}$ = 67.8 km s ^-1 Mpc^-1 (Virgo Infall only) and $\Omega_{\text{matter}}=0.308$ However, in the local universe, measurements need a better “standard candle,” such as SNe Ia. The distance modulus of NGC 3147 can be calculated if we know the absolute magnitude $M_{B,max}$ of SN 2021hpr. Based on the decline rate $\Delta m_{15}(B)=0.949\pm 0.019$ mag (see Section 3.1) and the Phillips relation (Phillips et al. 1999; e.g., Wu et al. 1995), we can derive $M_{B,max}=-19.531\pm 0.210$ mag. Then, the distance modulus should be $m-M$ = $33.46\pm 0.21$ mag, which is the same as the result of Jha et al. (2007) ($m-M$ = $33.46\pm 0.11$ mag), who measured the distance modulus for SN 1997bq using the multicolor light-curve shape method (MLCS2K2). However, the distance moduli vary from 32.41 mag to 33.46 mag even when measured from the same SNe (e.g. SN 1997bq; Reindl et al., 2005; Wang et al., 2006; Prieto et al., 2006; Jha et al., 2007; Kowalski et al., 2008; Takanashi et al., 2008; Amanullah et al., 2010; Tully et al., 2013). While differences in the final result could be due to different physics, distance modulus measurements can be also affected by many other factors, such as photometric accuracy, internal reddening of the host galaxy, and template subtraction.

To search for the progenitor of SN 2021hpr, we inspected the Hubble Space Telescope (HST) images before and after (see Figure 12) SN 2021hpr exploded, and we failed to detect it. However, we can obtain an upper limit (3$\sigma$) for it of 28.50 mag, 28.35 mag and 27.87 mag (AB) in the WFC3/F350LP, WFC3/F555W and WFC3/F814W bands, respectively. Using the distance modulus obtained above, we calculate that the absolute magnitudes of the progenitor of SN 2021hpr should be fainter than -4.96 mag , -5.11 mag, and -5.59 mag in the above three bands.