Projets Spectroscopie

Supernovae Ia


Retour Accueil






Type Ia supernovae are noticely caracterized by the absence of Balmer lines in their spectra. Thus, they are understood to be the result of the explosion of CO white dwarf which lost its H and He enveloppe during the AGB stade. The white dwarf becomes a supernova if it is driven above the Chandrasekhar mass limit (1,4 solar mass). The accretion of hydrogen and helium from a companion in a close binary system (i.d. a cataclysmic or a symbiotic star) is the most successful model to explain both light curve and spectral evolution.

The total electromagnetic output is ~1049 erg.
The kinematic energy is ~ 1051 erg
Absolute magnitude MB ~ -19


"There's a wide consensus that type Ia supernovae are the outcome of the thermonuclear explosion of a carbon-oxygen white dwarf in a binary system. Nervertheless, the nature of this system, the process of ignition itself and the development of the explosion continue to be a mystery."
J. Isern & al., 11th Symposium on Nuclei in the Cosmos, 2010.


The white dwarf undergoes a degenerate carbon burning in the core. The most successful model,"W7", (Nomoto & al., 1984) predicts the synthesis of 0.8 solar mass of iron peak elements, including 0.58 solar mass of 56Ni, while a subsonic burning front propagates throught the outer layers and make the star explode leaving no remnant behind. The outermost layers (about 0.1 solar mass of unburned Carbon and Oxygen) expands at higher velocities (up to about 22 000 km/s).
The deflagration wave produce about 0.5 solar mass of intermediate mass elements, such as Si, Ca, S, Na, ejected at speeds between 10 and 15 000 km/s. This layer produce the main features of the spectrum. The central region (0.8 solar mass) expands up to 10 000 km/s : 56Ni decays in 56Co (6.1d half-life) and 56Co decays in 56Fe (77d half-live).
The gamma-rays produce by these decays are converted to UV and optical photons upon their interacion with the expanding enveloppe. This conversion powers the light-curve and makes the supernova optically bright.
After the explosion, the different layers undergo free expansion, retaining the velocity acquired within a few seconds after the explosion. But strict stratification is not preserved ; some mixing, especially in the intermediate and outer layers.

Photospheric phase
At early phases, spectra of SN show broad absorption lines on a hot stellar-like (B type) continuum. This epoch is called "photospheric phase". A photosphere can be defined somewhere in the envelope. It receeds with time, while the layers expands and their density decreases.

Nebular phase
About 4 weeks after maximum, the density of the ejecta becomes so low that the photosphere diappears and the spectrum become nebular, with strong emission lines from forbidden transitions and no underlying continuum.




Lines identification and spectral evolution



Lines identification is not easy in a supernovae :
- lines are very broad,
- "lines" are often blends, or even "trough"
- for distant galaxies, lines are redshifted
Computer modeling of the expanding ejecta has resulted in reliable identifications for most features, after decades of uncertainty. See for instance : Comparative direct analysis of type Ia supernovae spectra, I. SN 1994D, Branch & al., PASP, 2005.

In general, at very early times they are attributed to lines of neutral and singly ionized intermediate-mass elements (O, Mg, Si, S, Ca)
. The spectrum probes the outermost layers of the progenitor which are the fastest moving

Near maximum light the main features are :

- Ca II H & K lines - not seen in this spectrum
1. Si II
2. trough (unresolved blend of many lines) centered at ~ 4300 anstöms, with FeII and FeIII lines, but also Mg II contribution for the fainter SN, and TiII for very faint objects.
3. trough around ~ 4800 angströms, essentially Fe II and Fe III with almost no contamination of other lines
4. SII absorptions causes a typical W shape with minima observed at ~ 5250 and 5400 angströms. About 7 days after maximum SII "W" feature weakens significatively and blends with FeII and Si II. The two braod features are sometimes so broadened (in highest expension velocities SN) that they are almost indistinguishable.
6. the first Si II blend is seen at ~ 5250 angströms. The rest wavelenght is 5972 angströms (average)
7. the second Si II blend at ~ 6150 angtröms is the most caracterictic line in a SNIa spectrum. The rest wavelenghts are 6347 and 6371, for a weight average of 6355 angströms.

Observed wavelenght
~ 4300

An important breakdown is observed about 6 days before maximum light.

By t approx 2 weeks the spectrum is dominated by lines of Fe II, which is consistent with an iron-rich core, but some lines of intermediate-mass elements are still present (e.g. Si II, Ca II). The relative contribution of iron-group elements quickly increases as the photosphere recedes into the ejecta. A broad emission near 6500 Å, which begins to develop around t = 2 weeks and remains thereafter with some changes in shape, is not Halpha but rather Fe II and later [Fe II].












About 4 weeks after maximum, the density of the ejecta becomes so low that the photosphere diappears and the spectrum become nebular. Forbidden [FeII], [FeIII] and Co[III] become the dominant spectral features













Near maximum luminosity
Post maximum SiII Phase

SN 2011fe - 13-09-2011 - 4 days after maximum




Two weeks after maximum luminosity
SiII to FeII Transition Phase

SN 2011fe - 24-09-2011 - 14 days after maximum
Grey line : previous spectrum for comparison
Note the decrease of the blue pseudo-continuum, disappearance of SiII 5972 and SII features


Radial velocity measurements

Description of the method (Hachinger & al. 2006) :

"The mean expansion velocity of a given line is determied from the bluesshift of the absorption relative to the rest wavelenght. The velocity is physically meaningful only for the single lines or for close multiplets that are sufficiently isolated from other strong features.
Since most lines have a P Cygni shape and are blends, one has to think thoroughly about how to measure the 'mean' wavelenght of absorption.

We used two different methods : the first is to use the gaussian fit routine with IRAF ; the second is to estimate the centre of the absorption by eye (taking into account problems such as the sloping of the continuum). The values obtained from several such measurements were then averaged ; the respective standart deviation is a crude estimate of the error introduced by the manual measurements and was used to calculate the error in the cases mentioned above."








Vexp measurement by gaussian fitting
Si II line in SN 2011fe | 13-09-2011

A gaussian fitting is realized in Excel 2010. The result (red line) is the sum of two gaussians :

standart deviation
Raie 1
Raie 2

The wavelenght of the minimum of the gaussian fit is 6146 angströms. The blueshift in respect to the rest wavelenght (6355 anströms) is 209 angströms, i.e. 9870 km/s.
The mean expension velocity is therefore 9870 km/s for this line.


Results on SN 2011fe - Spectra from ARAS campaign :

Si II "6355" velocity

Red squares : published values
Blue squares : measurements from ARAS data base spectra

Note the slope break at ~ -7 days


Accreting white dwarf model for type I supernovae
Nomoto & al., A. J., 286 : 644-658, 1984

Exploring the spectroscopic diversity of Type Ia supernovae
S. Hachinger & al., Mon. Not. R. Astron. Soc., 370, 299-318, 2006

Optical spectra of supernovae
Filippenko, Annu. Rev. Astron. Astrophys. 1997. 35: 309-355

Type Ia Supernova Cosmology : Quantitative Spectral Analysis

Standardisation spectroscopique des supernovae de type Ia
Thèse Jérémy Le Du