Constraining the Properties of SNe~Ia Progenitors from Light Curves
B. Sadler, P. Hoeflich, E. Baron, K. Krisciunas, G. Folatelli, M. Hamuy, M. Khokhlov. M. Phillips, N. Suntzeff, L. Wang
aa r X i v : . [ a s t r o - ph . C O ] S e p Title of your IAU SymposiumProceedings IAU Symposium No. 281, 2011A.C. Editor, B.D. Editor & C.E. Editor, eds. c (cid:13) Constraining the Properties of SNe IaProgenitors from Light Curves
B. Sadler , Peter Hoeflich , E. Baron , K. Krisciunas , G. Folatelli ,M. Hamuy , M. Khokhlov , M. Phillips , N. Suntzeff , L. Wang Dept. of Physics, Florida State University,USA Dept. of Physics and Astronomy, University of Oklahoma, Dept. of Physics, Texas A & M University, Dept. de Astronomia, Universidad de Chile, Santiago, Chile, Dept. Astronomy and Astrophysics, University of Chicago, USA, Las Campanas Observatory, La Serena, Chile
Abstract.
We present an analysis of high precision V light curves (LC) for 18 local Type IaSupernovae, SNe Ia, as obtained with the same telescope and setup at the Las Campanas Obser-vatory (LCO). This homogeneity provides an intrinsic accuracy a few hundreds of a magnitudeboth with respect to individual LCs and between different objects. Based on the Single De-generate Scenario, SD, we identify patterns which have been predicted by model calculationsas signatures of the progenitor and accretion rate which change the explosion energy and theamount of electron capture, respectively. Using these templates as principle components and theoverdetermined system of SN pairs, we reconstruct the properties of progenitors and progenitorsystems. All LCO SNe Ia follow the brightness decline relation but 2001ay. After subtraction ofthe two components, the remaining scatter is reduced to ≈ .
01 to 0 . m . Type SNe Ia seemto originate from progenitors with Main Sequence masses, M MS > M ⊙ with the exception oftwo subluminous SNe Ia with M MS < M ⊙ . The component analysis indicates a wide range ofaccretion rates in the progenitor systems closing the gap to accretion induced collapses (AIC).SN1991t-like objects show differences in dm
15 but no tracers of our secondary parameters. Thismay point to a different origin such as DD-Scenario or the Pulsating Delayed Detonations.SN2001ay does not follow the decline relation. It can be understood in the framework of C-richWDs, and this group may produce an anti-Phillips relation. We suggest that this may be a resultof a common envelope phase and mixing during central He burning as in SN1987A.
1. Introduction
SNe Ia are thermonuclear explosions of WD (Hoyle&Fowler,1960) i.e. end-stages ofstellar evolution of stars between 1 and 7 M ⊙ . Most likely, they result from the explosionof a C/O-WD with a mass close to the Chandrasekhar limit ( M Ch ), which accretes matterthrough Roche-lobe overflow in a single degenerate scenario (SD) (Whelan & Iben,1973),or merging of two degenerate WDs (DD) (Webbink,1984, Iben & Tutukov,1984). We re-gard SDs as most likely for the majority of SNe Ia because e.g. of the homogeneity inLCs and spectra, though there is strong evidence of contributions of both to the SNe Iapopulation (see Hoeflich,2006, and references therein). One of the keys is the empiri-cal relation between maximum brightness and the rate of decline, dm dm
15 is well understood: LCs are powered by radioactive decay of N i (Colgate & McKee,1969). More N i increases the luminosity and causes the envelopes tobe hotter. Higher temperature means higher opacity and, thus, longer diffusion time scalesand, thus, slower decline rates after maximum light (Hoeflich et al.,1996, Nugent et al.,1997,Maeda et al.,2003, Kasen&Woosley,2009). The existence of a dm Figure 1.
Principle components based on theoretical models. We show the difference in Vbrightness ∆ m ( t ) (in magnitudes) as a function of time (in days) relative to a reference modelwith solar metallicity and a main sequence mass M MS of 7 M ⊙ and a central density of theexploding WD of ρ c of 2 × g cm − . The plots are for models with 7 M ⊙ and 5 × g cm − ,respectively. The annotation on the graphs give the main reason for the differences. virtually all scenarios as long as there is an excess amount of stored energy to be released(Hoeflich et al.,1996).Within SDs, the favorite models are M Ch explosions in which burning starts as de-flagration which, at some point, transitions to a detonation DDT (Khokhlov,1991). The N i production depends mostly on the DDT, i.e. one parameter. The expanding SN-envelopes have similar velocity and density structures because same masses, most of theWD undergoes burning, and the nuclear binding energies of the burning products arenearly the same. A dispersion of 0 . ... . m is expected (see Fig. 1). Its origin can berelated to the progenitor and accretion from the donor star: Changes in the metallicityand the Main Sequence Mass, M MS , will change the size of the explosion energies byabout 20 % because their influence of the size of C-depleted core formed during stel-lar He-core burning (Hoeflich et al.,1998, Dom´ınguez et al.,2001). Increasing accretionrates will decrease the central WD density ρ c at the time of explosion and, consequently,an increased electron capture will produce more stable iron-group elements at the ex-pense of N i . High precision LCs have been obtained at LCO for 18 local SNe Ia(Contreras et al.,2010,, Folatelli et al.,2010) and the individual, theoretical signatureshave been recovered ( Hoeflich et al.,2010). Note that a similar pattern as by ρ c can beproduced by off-center N i distributions but ∆ m increases later at ≈
60 to 100 days.
2. Diversity of Type Ia Supernovae
Secondary parameters:
Two SNe Ia may differ in both ρ c /accretion and the C-depleted core/ M MS . Therefore, we employed component analyses to study secondaryparameters. We use V because this color is hardly effected by metallicity, asphericityeffects, and k-corrections. The differences ∆ m ( t ) in LC pairs are described by∆ m ij,obs ( t ) = X k =1 , λ k ( ij ) f k ( t ) + Res ij ( t )with λ k ( ij ) being the coefficients for a pair of SN i and j , f k ( t ) the principle compo-nents, and Res ( t ) the residuals. In our sample, we include 18 SN from LCO, i.e. 153pairs. Only 36 λ ij = g k ( i ) /g k ( j )’s are independent where g k ( i ) is the eigenvalue of f k to be attributed a specific SNe Ia. By solving the overdetermined system for g i us-ing a Simplex Method (Nelde&Mead,1965), we obtain most likely values for ˜ λ k ( ij ). InFig. 2, the pairs SN2005al/SN2005am, SN2005el/sn2005ef, SN 2005ef/SN 2005na andSN 2005al/SN2005ef are given for ˜ λ ( ij ). Overall, the residuals are consistent with zerobut fits are not unique due to errors in brightness and time coverage (see Fig. 3). Remap- onstraining the Properties of SNe Ia Progenitors from Light Curves Figure 2.
Best fits to the observations of individual pairs of SNe Ia. We give the weightedcomponents, their sums along the observations (with error bars) and the residuals (crosses) forcases which are dominated by the central density/accretion rate (upper left), in the progenitors(upper right), similar ρ c and M ⊙ (lower left), and a mixed case (lower right). Note that theresiduals are small and, within the error bars, consistent with zero. Figure 3.
Probability distributions of ˜ λ ij for the pairs of SN2005al/SN2005am andSN2005el/sn2005ef based on MC solutions for the overdetermined system. Sparse time coverageproduce large uncertainties in the eigenvalues. ing the individual g k ( i ) to M MS and ρ c , shows that (a) ρ c are evenly distributed from1 × g cm − to 7 × gcm − , i.e. close densities leading to an accretion inducedcollapse (AIC) (b) SN Ia come from massive progenitors with M > M ⊙ but the twosubluminous SNe Ia indicate M MS between 1 ... 2 M ⊙ . All pairs of SN1991t-like objectsshow λ k ( ij ) = 0 in all components. Either, they are very similar despite differences in dm
15, or they lack a central region of high densities and similar C/O in the progenitorsas may be expected for mergers (DD) or pulsating delayed detonation models. For moredetails and a complete analysis of all 28 LCO SNe Ia, see Sadler et al. (in preparation).
SN2001ay shows that nature is even more diverse. SN2001ay is slower than any SNe Iaknown, combined with a fast rise of some 16 days (Krisciuanas et al. 2011). SN2001aywould be brighter by about 1 m based on dm
15 and the distance to the host galaxy. Infact, dm
15 is slower than implied by the instantaneous energy input by radioactive decays˙ E γ . We submit that, still, this SN can be understood within the physics underlying thedm15 relation, and in the framework of pulsating delayed detonation models originatingfrom a M Ch mass WD but with a core of some 80% C rather instead of the 15 to 20% usual for stellar central He burning. Higher C fraction means more nuclear energy by C ( α, γ ) O by ≈
40 % and faster expansion of the inner layers. Faster expansion meansthat a larger fraction of the energy by N i decay goes into expansion work rather than22 B. Sadler et al.
Figure 4.