Imaging Polarimetry of the Type I Superluminous Supernova 2018hti

Superluminous supernovae (SLSNe) are extremely bright explosions at the end stage of stellar evolution. The explosion peaks at absolute magnitudes of −21 ∼ −23, up to 100× brighter than classical SNe (Quimby et al. 2011; Gal-Yam 2012). At this brightness, it is unlikely that SLSNe are powered by the same mechanisms as normal SNe, and the powering source remains unclear (Gal-Yam 2018). There are currently three leading models to explain the SLSNe explosion. The first one is pair-instability (Kasen et al. 2011), in which a very massive stellar core (∼100 M_⊙) contracts rapidly following the loss of pressure due to electron–positron pair production and undergoes thermonuclear runaway that powers the superluminous supernova explosion. The second scenario is so-called magnetar (Maeda et al. 2007), where additional power is introduced on top of the normal SNe by spin-down of a magnetic rapid rotating neutron star. The third scenario is through interaction of SN shock with massive circumstellar medium (Woosley et al. 2007), boosting the kinetic energy to higher luminosities. These different models will intrinsically lead to (a) spherical explosions, providing clues toward discerning the powering source. In this regard, polarimetry can play a major role by peeking into the 3D structure of SLSNe explosion, as spheric/aspheric explosion will/will not induce a polarization signal. Polarimetry is especially useful as we cannot directly resolve the distant SLSNe into the extended source.

However, most SLSNe are associated with host galaxies at z > 0.1, and thus tend to have low signal-to-noise ratios even with imaging polarimetry, let alone spectropolarimetry. This hampers high precision polarization measurement, and is the reason why only a handful of SLSNe have been observed to date—LSQ14MO (z > 0.2, Leloudas et al. 2015), 2015bn (z > 0.1, Leloudas et al. 2017), and 2017egm (z = 0.03 Bose et al. 2018; Maund et al. 2019)—mostly with imaging polarimetry. Using the Very Large Telescope, Leloudas et al. (2015) concluded that LSQ14MO has no significant polarization signal. On the other hand, there was an increasing degree of polarization seen in 2015bn, suggesting two layers of shells during explosion (Leloudas et al. 2017). Bose et al. (2018) and Maund et al. (2019) both reported nondetection of polarization of the most nearby SLSNe 2017egm, although it can be partly attributed to contamination by the extend spiral host galaxy. Here we present imaging polarimetry observations of another nearby SLSN, 2018hti in a dwarf galaxy at z = 0.063.

This paper is organized as follows. In Section 2 we describe our Alhambra Faint Object Spectrograph and Camera (ALFOSC) polarimetry observations. In Section 3 we present results from polarimetry analysis, to derive the degree of polarization at different epochs; we also compare our results to other SLSNe polarimetry studies, followed by a summary in Section 4.

SN 2018hti was first reported by the ATLAS survey (Tonry et al. 2018), which was detected on 2018 November 2 with 18.43 ± 0.082 mag in the orange-ATLAS filter, but was not seen on the previous night with limiting magnitude of 19.46 mag. Spectroscopic follow-up on 2018 November 6 by the Global Supernova Project (PI: Howell; Arcavi et al. 2018) showed that SN 2018hti is similar to the SLSN-I PTF09cnd at 16 days before peak. As SN 2018hti was relatively bright and well before the peak, we realized that we could easily obtain high precision imaging polarimetry throughout the evolution of its light curve. We thus requested fast-track service observations at the Nordic Optical Telescope using its ALFOSC in the linear imaging polarimetry mode to monitor the change in polarization during the course of brightening and fading. The polarimetry mode of ALFOSC split the source light into two orthogonally polarized beams, i.e., ordinary and extraordinary components, on the same image separated by ∼15'' (as shown in Figure 1). Observations were conducted in the V-band at four different half-wave plate angles (0°, 225, 45°, and 675). Given the bright nature of SN 2018hti, a 300 s integration time per half-wave plate angle was able to deliver polarization precision of ∼0.2%, enabling us to detect polarization from an aspherical explosion, or put tight constraints on a spherical explosion. We obtained three epochs of polarimetry observations, on November 15th, and December 1st and 7th, respectively.

Zoom In Zoom Out Reset image size

The data reduction was carried out in a standard manner using the Image Reduction and Analysis Facility.¹ The polarimetry images are first bias-subtracted and flat-fielded. After detrending, we then proceed to obtain the flux of the ordinary and extraordinary components. Please note that after the half-wave plate, the polarized images of SN 2018hti are not circular, hence making point-spread function (PSF) photometry difficult. Due to this problem, we used aperture photometry instead of PSF fitting, to better extract the photometry. We used a fixed aperture size of two times the size of the FWHM, and extract the photometry using SExtractor (Bertin & Arnouts 1996). With the flux of ordinary and extraordinary beams in hand, we can proceed to derive the linear polarization as follows. Patat & Romaniello (2006) introduced the normalized flux differences F_i as

$$\begin{eqnarray}&&{F}_{i}=\displaystyle \frac{{f}_{O,i}-{f}_{E,i}}{{f}_{O,i}+{f}_{E,i}},\end{eqnarray} \tag{ 1 }$$

where ${f}_{O,i}$ and ${f}_{E,i}$ are flux from ordinary and extraordinary beams from the ith half-wave plate angle. With the normalized flux differences, we can derive the normalized Stokes parameters $\bar{Q}$ and $\bar{U}$:

$$\begin{eqnarray}&&\bar{Q}=\displaystyle \frac{Q}{I}=\displaystyle \frac{1}{2}\displaystyle \sum _{i=0}^{3}{F}_{i}\cos \left(\displaystyle \frac{\pi }{2}i\right),\end{eqnarray} \tag{ 2 }$$

$$\begin{eqnarray}&&\bar{U}=\displaystyle \frac{U}{I}=\displaystyle \frac{1}{2}\displaystyle \sum _{i=0}^{3}{F}_{i}\sin \left(\displaystyle \frac{\pi }{2}i\right),\end{eqnarray} \tag{ 3 }$$

where I is the Stokes parameter for polarization intensity, Q and U describe the linear polarization (see, e.g., Chandrasekhar 1950). The degree of linear polarization P is then

$$\begin{eqnarray}&&P=\displaystyle \frac{\sqrt{{Q}^{2}+{U}^{2}}}{I}=\sqrt{{\bar{Q}}^{2}+{\bar{U}}^{2}}.\end{eqnarray} \tag{ 4 }$$

With these formulae, we derive a polarization of 1.91% ±0.12%, 1.90% ± 0.12%, and 1.89% ± 0.13% on November 15th, and December 1st and 7th, respectively, for our observations of SN 2018hti. In addition, we also obtained polarimetry of unpolarized and strongly polarized standard stars, which all show degrees of polarization consistent with literature values. The polarization measurements on the normalized Q and U planes are shown in Figure 2.

Zoom In Zoom Out Reset image size

To check if the polarization of SN 2018hti changes during the brightening and fading phases, we compare the timing of our polarization measurements with the light curves from ASAS-SN, as shown in Figure 3. It appears that the degree of polarization (P) does not differ during the brightening and fading of the light curve. We should note that taking the face value, it appears that SN 2018hti exhibits a polarization of ∼1.9%. However, we also need to take into account the interstellar polarization (ISP), which comes from the Milky Way (MWG) and the host galaxy. This is especially the case because SN 2018hti has a relatively high line-of-sight MWG extinction, E(B − V) ∼ 0.4 from Schlafly & Finkbeiner 2011 and Schlegel et al. (1998). According to empirical upper limits from Serkowski et al. (1975), this can translate into P as high as ∼4%. To have a realistic evaluation of the ISP, we thus use a field star in the ALFOSC field of view.² Although the field star is fainter than SN 2018hti, taking the average of different epochs of the observations we found ISP to be 1.76% ± 0.6%. This suggests that the polarization we have seen toward SN 2018hti could reconcile with ISP. On the other hand, the SN could also have polarization of up to 0.8%, or an major/minor axis ratio of ∼0.9 (Hoflich 1991), and be consistent with the measurements. We note that besides MWG, the host galaxy extinction can also contribute to the polarization toward the SN. However, because previous studies have shown that reddening from the SLSN host galaxy can be negligible (see, e.g., Lunnan et al. 2014; Leloudas et al. 2015; Nicholl et al. 2016; Perley et al. 2016), we make this assumption and consider that MWG plays the major role in the ISP.

Zoom In Zoom Out Reset image size

This low polarization is similar to previous studies of other SLSNe, e.g., LSQ14mo, 2015bn, and 2017egm. This suggests that the explosion of 2018hti is spherical, which is consistent with the results of previous studies of other SLSNe. We note that Leloudas et al. (2017) reported an increased degree of polarization at >20 days after the peak of 2015bn, and attributed it to the phase transition, where the photospheric emission shifted from an outer layer to a relatively aspherical inner core. However, it can be difficult to test such a scenario in the case of 2018hti, especially due to strong contamination from the interstellar polarization.

We present imaging polarimetry results of the recently discovered nearby SLSN-I 2018hti. Our studies can be summarized as follows:

• 1.  
  Using ALFOSC, we obtained time-series imaging polarimetry before and after the peak of light curve. All three epochs of polarimetry observations show polarization at the ∼1.9% level.
• 2.  
  SN 2018hti is at a relatively high extinction line of sight, with E(B − V) ∼ 0.4. Using a field star in the ALFOSC, we also measured the interstellar polarization, which is at the level of ∼1.7%. This suggests that the polarization toward SN 2018hti is consistent with ISP within 3σ level, leading to nondetection of polarization from 2018hti itself.
• 3.  
  The insignificant polarization of 2018hti is similar to previous studies of other SLSNe, suggesting that SLSNe explosions are likely spherical.

We are indebted to the anonymous referee, whose comments greatly improved the manuscript. C.H.L. would like to thank Wen-Ping Chen and BigBoomers (especially Grant Williams) for inspiring and insightful ideas and knowledge of polarimetry and supernovae. We are grateful to the staff at the Nordic Optical Telescope, especially Dr. Tapio Pursimo, for scheduling and carrying out timely and flexible observations under the fast-track service programme. This work is based on observations made with the Nordic Optical Telescope, operated by the Nordic Optical Telescope Scientific Association at the Observatorio del Roque de los Muchachos, La Palma, Spain, of the Instituto de Astrofisica de Canarias. The data presented here were obtained with ALFOSC, which is provided by the Instituto de Astrofisica de Andalucia (IAA) under a joint agreement with the University of Copenhagen and NOTSA.