In this Letter we present evidence for a spectral sequence among Type Ia supernovae (SNe Ia). The sequence is based on the systematic variation of several features seen in the near-maximum light spectrum. This sequence is analogous to the recently noted photometric sequence among SNe Ia which shows a relationship between the peak brightness of a SN Ia and the shape of its light curve. In addition to the observational evidence we present a partial theoretical explanation for the sequence. This has been achieved by producing a series of non-LTE synthetic spectra in which only the e ective temperature is varied. The synthetic sequence nicely reproduces most of the di erences seen in the observed one and presumably corresponds to the amount of 56 Ni produced in the explosion.

No abstract

In this paper we present spectroscopic and photometric observations for four core‐collapsed supernovae (SNe), namely SNe 1994N, 1999br, 1999eu and 2001dc. Together with SN 1997D, we show that they form a group of exceptionally low‐luminosity events. These SNe have narrow spectral lines (indicating low expansion velocities) and low luminosities at every phase (significantly lower than those of typical core‐collapsed supernovae). The very‐low luminosity during the 56Co radioactive decay tail indicates that the mass of 56Ni ejected during the explosion is much smaller (MNi≈ 2–8 × 10−3 M⊙) than the average (MNi≈ 6–10 × 10−2 M⊙). Two supernovae of this group (SN 1999br and SN 2001dc) were discovered very close to the explosion epoch, allowing us to determine the lengths of their plateaux (≈100 d) as well as establishing the explosion epochs of the other, less completely observed SNe. It is likely that this group of SNe represent the extreme low‐luminosity tail of a single continuous distribution of Type II plateau supernovae events. Their kinetic energy is also exceptionally low. Although an origin from low‐mass progenitors has also been proposed for low‐luminosity core‐collapsed SNe, recent work provides evidence in favour of the high‐mass progenitor scenario. The incidence of these low‐luminosity SNe could be as high as 4–5 per cent of all Type II SNe.

We present the extension of our NextGen model atmosphere grid to the regime of giant stars. The input physics of the models presented here is nearly identical to the NextGen dwarf atmosphere models, however spherical geometry is used self-consistently in the model calculations (including the radiative transfer). We re-visit the discussion of the effects of spherical geometry on the structure of the atmospheres and the emitted spectra and discuss the results of NLTE calculations for a few selected models.

We have calculated a grid of photospheric phase atmospheres of Type Ia supernovae (SNe Ia) with metallicities from ten times to one thirtieth the solar metallicity in the C+O layer of the deflagration model, W7. We have modeled the spectra using the multi-purpose NLTE model-atmosphere and spectrum-synthesis code, PHOENIX. We show models for the epochs 7, 10, 15, 20, and 35 days after explosion. When compared to observed spectra obtained at the approximately corresponding epochs these synthetic spectra fit reasonably well. The spectra show variation in the overall level of the UV continuum with lower fluxes for models with higher metallicity in the unburned C+O layer. This is consistent with the classical surface cooling and line blocking effect due to metals in the outer layers of C+O. The UV features also move consistently to the blue with higher metallicity, demonstrating that they are forming at shallower and faster layers in the atmosphere. The potentially most useful effect is the blueward movement of the Si II feature at 6150Å with increasing C+O layer metallicity. We also demonstrate the more complex effects of metallicity variations by modifying the 54 Fe content of the incomplete burning zone in W7 at maximum light. We briefly address some shortcomings of the W7 model when compared to observations. Finally, we identify that the split in the Ca H+K feature produced in W7 and observed in some SNe Ia is due to a blending effect of Ca II and Si II and does not necessarily represent a complex abundance or ionization effect in Ca II.