EARLY SPECTROSCOPIC IDENTIFICATION OF SN 2008D^*

On 2008 January 9.56 UT, while observing the supernova (SN) 2007uy in the nearby spiral galaxy NGC 2770 (z = 0.007), the X-Ray Telescope onboard Swift detected a bright X-ray transient (XT), with a peak luminosity of 6 × 10⁴³ erg s⁻¹ and a duration of about 10 minutes (Soderberg et al. 2008b). Its power-law spectrum and light-curve shape were reminiscent of gamma-ray bursts (GRBs) and X-ray flashes, but the energy release was at least two orders of magnitude lower than for typical and even subluminous GRBs, also allowing for beaming (e.g., Amati 2006; Ghirlanda et al. 2007). The discovery of the XT prompted the search for, and discovery of, an optical counterpart (Deng & Zhu 2008; Valenti et al. 2008b).

We performed spectroscopy of the source as soon as possible, starting 1.75 days after the XT, using the FORS2 spectrograph on the ESO Very Large Telescope (VLT). Subsequent spectroscopic monitoring of the object was carried out at the Nordic Optical Telescope (NOT) and the William Herschel Telescope (WHT). All spectra have been reduced using standard techniques. On January 18.22 UT (8.65 days after the XT), we secured a high-resolution spectrum using the UVES instrument on the VLT. For this observation, we adopted the ESO CPL pipeline (v3.3.1), and flux calibration was performed using the master response curves. The observing log of the spectra is reported in Table 1.

Imaging observations were conducted using the NOT, the VLT, the Liverpool Telescope (LT), and the United Kingdom Infrared Telescope (UKIRT). Image reduction was carried out using standard techniques. For photometric calibration, we observed optical standard star fields on five different nights, and defined a local sequence in the NGC 2770 field. In the near-infrared, we used Two Micron All Sky Survey stars as calibrators. Magnitudes were computed using small apertures, and subtracting the background as measured in an annulus around the SN position. The contribution from the underlying host galaxy light was always negligible, as also apparent from archival Sloan Digital Sky Survey images. Our photometric results are listed in Table 2.

One of our spectra covers the nucleus of NGC 2770, allowing a precise measurement of its redshift: z = 0.0070 ± 0.0009. This is slightly larger than the value listed in the NED database (z = 0.0065). For H₀ = 71 km s⁻¹ Mpc⁻¹, the luminosity distance is 29.9 Mpc.

2.1. The Early Spectrum of SN 2008D

Our first optical spectrum of the transient source (Figure 1) exhibits Na i D absorption lines at z = 0.0070, thus establishing its extragalactic nature. Broad features are also apparent across the whole spectrum (FWHM = (1–3) × 10⁴ km s⁻¹), which led us to identify the object as a core-collapse SN (Malesani et al. 2008). Soderberg et al. (2008a) describe nearly simultaneous spectra as featureless, probably due to their smaller covered wavelength range (4500–8000 Å). Modjaz et al. (2008b) report features consistent with those in our data.

Zoom In Zoom Out Reset image size

We initially classified the SN as a very young Type Ib/c, based on the absence of conspicuous Si and H lines (Malesani et al. 2008). As the spectrum is among the earliest observed for any SN, comparable only to the very first spectrum of SN 1987A (Menzies et al. 1987), there is no obvious resemblance with known SN spectra. It is notable, however, that the earliest spectrum of the Type-Ic SN 1994I was essentially flat with broad, low-amplitude undulations (though the covered wavelength range was limited; Filippenko et al. 1995). Early spectra, also mostly featureless, are available for the H-rich Type-IIP SN 2006bp (Quimby et al. 2006). Dessart et al. (2008) interpret them in terms of high temperature and ionization.

A striking feature in the spectrum is a conspicuous W-shaped absorption with minima at 3980 and 4190 Å (rest frame). It was detected using two different instrument setups (Figure 1), and also reported by Modjaz et al. (2008b). If interpreted as due to P Cyg profiles, the inferred expansion velocity is ∼15,000 km s⁻¹, computed from the position of the bluest part compared to the peak. Its origin is unclear, although, following Quimby et al. (2007), Modjaz et al. (2008b) propose that it is due to a combination of C iii, N iii, and O iii. Interpreting the broad absorption at ∼5900 Å as Si iiλλ 6347, 6371, some ejecta reached ∼22,000 km s⁻¹. Such large velocities have been seen only in broad-lined (BL) Type-Ic SNe, at significantly later stages (Patat et al. 2001; Hjorth et al. 2003; Mazzali et al. 2006).

2.2. The Photospheric Phase

Figure 2 shows the optical and near-infrared light curves of SN 2008D. In the first days after the XT, the flux dropped faster in the bluer bands, with the color becoming progressively redder. This can be interpreted as due to the stellar envelope cooling after the shock breakout (Soderberg et al. 2008b). Our first spectra were taken during this stage, before energy deposition by radioactive nuclei became dominant; hence the physical conditions of the emitting material might be different than later. We note that the W-shaped absorption discussed in Section 2.1 was no longer visible from 3.5 days after the XT onward (Figure 3).

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

The later spectra, acquired during the radioactivity-powered phase and extending over more than two months in time, established SN 2008D as a Type-Ib SN (Modjaz et al. 2008a). From January 17 and onward unambiguous He lines are observed (Figure 3), consistent with other reports (Modjaz et al. 2008b; Valenti et al. 2008a; Soderberg et al. 2008b; Tanaka et al. 2008; Mazzali et al. 2008). In Figure 4, we plot the velocities at maximum absorption of a few transitions determined using the SYNOW code (Fisher et al. 1999). For comparison, we also plot the Si ii velocity of the BL SN 1998bw (Patat et al. 2001) and of the normal Type-Ic SN 1994I (Sauer et al. 2006), showing that the velocities of SN 2008D are lower than those of BL SNe.

Zoom In Zoom Out Reset image size

2.3. Dust Extinction

The precise explosion epoch is so far only known for a few Type-II SNe, thanks to either the detection of the neutrino signal (SN 1987A; Hirata et al. 1987; Bionta et al. 1987) or of the UV flash by Galaxy Evolution Explorer (Shawinski et al. 2008; Gezari et al. 2008), and for BL Type-Ic SNe associated with GRBs (e.g., Galama et al. 1998; Hjorth et al. 2003; Stanek et al. 2003; Campana et al. 2006). SN 2008D is the first Type-Ib SN with a precisely constrained explosion epoch, since the XT is expected to occur less than 1 hr after the stellar collapse (Li 2007; Waxman et al. 2007). The nature of the XT—shock breakout versus relativistic ejecta—is still debated (Soderberg et al. 2008b; Xu et al. 2008; Li 2008; Mazzali et al. 2008; Chevalier & Fransson 2008); thus it is unclear whether this phenomenon is common.

The light-curve evolution of SN 2008D is very similar to that of SN 1999ex (Stritzinger et al. 2002). The initial fading can be interpreted as due to the envelope cooling through expansion following the initial X-ray/UV flash (Stritzinger et al. 2002; Campana et al. 2006; Soderberg et al. 2008b). The subsequent rebrightening is due to the energy released by radioactive material in the inner layers and gradually reaching the optically thin photosphere.

From our UBVRI data, we constructed the bolometric light curve of SN 2008D (Figure 5). For comparison we also show the two other He-rich SNe caught during the early cooling phase: SN 1993J (Richmond et al. 1994) and SN 1999ex (Stritzinger et al. 2002). Strikingly, the three SNe had very similar light curves during the photospheric phase. Given its peak luminosity, SN 2008D synthesized about 0.09 M_☉ of ⁵⁶Ni based on Arnett's rule (Arnett 1982). The emission during the early cooling phase, however, varied substantially for the three SNe (seen only in U for SN 1999ex). Furthermore, significant radiation may be emitted blueward of the U band during this early phase, so that the bolometric values may be underestimated.

Zoom In Zoom Out Reset image size

SN 2008D has different properties from the SNe associated with GRBs, namely the presence of He in the ejecta, a lower peak luminosity (≈1.5 mag), lower expansion velocities (a factor of ≈2), and lower ⁵⁶Ni mass (a factor of ≈5). The SN environment is also unlike that of GRBs (Soderberg et al. 2008b; Thöne et al. 2008). Whether these two kinds of high-energy transients are separate phenomena or form a continuum is unclear. Addressing this issue will require theoretical modeling and an enlarged sample. The discovery of a short-lived XT associated with an ordinary Type-Ib SN opens the possibility of accessing the very early phases of ordinary SNe, which will provide new insights into SN physics. Future X-ray sky-scanning experiments, such as Lobster or eROSITA, may turn out, rather unexpectedly, ideally suited to examine this issue, alerting us to the onset of many core-collapse SNe.

The Dark Cosmology Centre is supported by the DNRF. J.G. is supported by the Spanish research programs AYA2004-01515 and ESP2005-07714-C03-03, and P.M.V. by the EU under a Marie Curie Intra-European Fellowship, contract MEIF-CT-2006-041363. P.J. acknowledges support by a Marie Curie European Re-integration Grant within the 7th European Community Framework Program under contract number PERG03-GA-2008-226653, and a Grant of Excellence from the Icelandic Research Fund. We thank M. Modjaz and M. Tanaka for discussion, and the observers at VLT, NOT, WHT, UKIRT, in particular A. Djupvik, S. Niemi, A. Somero, T. Stanke, J. Telting, H. Uthas, and C. Villforth.