arXiv:astro-ph/0406263v1 10 Jun 2004
Frontier in Astroparticle Physics and Cosmology
Circumstellar Interaction of Type Ia Supernova SN 2002ic
Ken’ichi NOMOTO, Tomoharu SUZUKI, Jinsong DENG, Tatsuhiro UENISHI
Department of Astronomy & Research Center for the Early Universe, School of
Science, University of Tokyo, Tokyo 113-0033, Japan
Institute of Earth Science and Astronomy, University of Tokyo, Meguro–ku,
Osservatorio Astronomico di Trieste, via G.B.& Tiepolo 11, I-34131 Trieste, Italy
SN 2002ic is a unique supernova which shows the typical spectral features
of Type Ia supernovae (SNe Ia) near maximum light, but also apparent hydrogen
features that have been absent in SNe Ia. We have calculated hydrodynamical
models for the interaction between the SN Ia ejecta and the H-rich circumstellar
medium (CSM) to reproduce the observed features of SN 2002ic. Based on our
modeling, we suggest that CSM is aspherical (or highly clumpy) and contains ∼
4-5 M⊙. Possible progenitor systems of SN 2002ic are discussed.
1. Type Ia Supernovae and Circumstellar Medium
Type Ia supernovae (SNe Ia) are characterized by the lack of hydrogen and
the prominent Si line in their spectra near maximum light and widely believed
to be thermonuclear explosions of mass-accreting white dwarfs in binary systems.
SNe Ia have been used as a “standard candle” to determine cosmological param-
eters thanks to their relatively uniform light curves and spectral evolution. SNe
Ia are also the major sources of Fe in the galactic and cosmic chemical evolution.
Despite such importance, the immediate progenitor binary systems have not been
clearly identified yet (e.g., [10, 13]).
For a model of SN Ia progenitors, Hachisu et al. [5, 6] proposed a single
degenerate model in which the white dwarf blows a massive and fast wind (up
to 10−4M⊙yr−1and 2000 km s−1) and avoids a formation of common envelope
when the mass transfer rate from the normal companion exceeds a critical rate of
∼ 1 × 10−6M⊙yr−1. Such an evolutionary phase is dubbed “accretion wind
evolution” instead of “common envelope evolution.” Such a binary can keep its
c ?2004 by Universal Academy Press, Inc.
separation almost unchanged. The white dwarf can steadily accrete a part of the
transferred matter and eventually reach the Chandrasekhar mass.
In the strong wind model, the white winds form a circumstellar envelope
around the binary systems prior to the explosion. When the ejecta collide with
the circumstellar envelope, X-rays, radio, and Hα lines are expected to be emitted
by shock heating. Attempts have been made to detect such emissions, but so far
no signature of circumstellar matter has been detected.
The upper limit set by X-ray observations of SN 1992A is˙M/v10= (2−3)
× 10−6M⊙yr−1. Radio observations of SN 1986G have provided the most
stringent upper limit to the circumstellar density as ˙M/v10= 1 × 10−7M⊙yr−1,
where v10 means v10 = v/10 km s−1. This is still 10 − 100 times higher than
the density predicted for the white dwarf winds, because the white dwarf wind
velocity is as fast as ∼ 1000 km s−1. For Hα emissions, the upper limit of˙M/v10=
6 × 10−6M⊙yr−1has been obtained for SN 1994D.
SN 2002ic was discovered on 2002 November 13 UT at magnitude 18.5 by
the Nearby Supernova Factory search . Hamuy et al.  reported strong Fe
III features and a Si II λ6355 line in the early-time spectra of SN 2002ic and
classified it as a SN Ia.
However, strong Hα emission was also observed. The emission was broad
(FWHM > 1000 km s−1) suggesting that it was intrinsic not to an H II region of
the host galaxy but to the supernova. The detection of Hα is unprecedented in a
SN Ia (e.g., [1, 10]).
Hamuy et al.  suggested that it arose from the interaction between the
SN ejecta and a dense, H-rich circumstellar medium (CSM), as in Type IIn SNe
(SNe IIn). If this interpretation is correct, SN 2002ic may be the first SN Ia
to show direct evidence of the circumstellar (CS) gas ejected by the progenitor
system, presenting us with a unique opportunity to explore the CSM around a
SN Ia and the nature of the progenitor system.
2.1.Spectroscopic Features of SN 2002ic
The late-time spectrum of SN 2002ic is strikingly similar to those of Type
IIn SNe 1997cy  and 1999E  as shown in Figure 1. Spectroscopically,
SN 2002ic evolve with time significantly, in particular in the Hα line and its
complex profile. Hamuy et al.  detected in the early-time spectra an unresolved
narrow Hα emission on top of a ∼ 2000 km/s base, which were superimposed on
dominant SN Ia line features. One year after the explosion, however, the Hα line
Spectral comparison between SNe 2002ic (red thick lines; Subaru, ∼ 222 d),
1997cy (blue thin lines) and 1999E (green dashed line).
became much more prominent and consisted of a narrow core and a ∼ 5000 km/s
component. Other strong features identified in Figure 1 include Ca and O lines
as broad as ≥ 10,000 km/s and broad permitted Fe II multiplets .
SNe 1997cy and 1999E were initially classified as Type IIn because they
showed Hα emission. SN 2002ic would also have been so classified, had it not been
discovered at an early epoch. SN 1997cy (z = 0.063) is among the most luminous
SNe discovered so far (MV < −20.1 about maximum light), and SN 1999E is also
bright (MV < −19.5). Both SNe 1997cy and 1999E have been suspected to be
spatially and temporally related to a GRB [4, 14]. However, both the classification
and the associations with a GRB must now be seen as highly questionable in view
of the fact that their replica, SN 2002ic, appears to have been a genuine SN Ia at
an earlier phase.
2.2.Observed Light Curve of SN 2002ic
The UVOIR bolometric light curve of SN 2002ic has been constructed by
Deng et al.  from the available BVRI photometry and the spectrophotom-
etry [8, 18] as shown in Figure 2. To construct the light curve of SN 2002ic,
we first integrated the Subaru spectrum. This yielded L = (5.9 ± 0.6) × 1042
ergs s−1, corresponding to Mbol∼ −18.2, assuming a distance of 307 Mpc (H0=
65 km s−1Mpc−1). The bolometric corrections thus estimated was used to con-
vert the early-time photometry in Hamuy et al.  and the late-time MAGNUM
telescope photometry into bolometric luminosities.
The light curve of SN 2002ic is brighter at maximum and declines much
more slowly than typical SNe Ia . The late time light curve of most SNe is
powered by the radioactive decay of56Co to56Fe. The decline of SN 2002ic is
much slower than the Co decay rate, which indicates the presence of another
source of energy.
In fact the overall light curve of SN 2002ic resembles SNe 1999E  and
1997cy  (see Figure 2). We use UBV RI bolometric light curves of SNe 1997cy
and 1999E for comparison [17, 14]. Assuming E(B − V ) = 0.06 for the Galactic
extinction (NED), SN 2002ic is only a factor of 1.3 dimmer than SN 1997cy, but
more than 100 times brighter at late phases than typical SNe Ia. The light curve
of SN 1997cy has been modeled in the context of circumstellar interaction ,
which is very likely the same energy source for SN 2002ic.
Comparison of the UBV RI bolometric light curves of SN 2002ic (red filled
squares) with those of SNe 1997cy (blue open circles) and 1999E (green crosses),
and the normal SN Ia 1994D (black dashed line).
3. Circumstellar Interaction Models
We calculated the interaction between the expanding ejecta and CSM (de-
tails will be seen in Suzuki et al., in preparation). For the supernova ejecta, we
used the the carbon deflagration model W7 ; its kinetic energy is E = 1.3 ×
1051erg. For CSM we assumed the power-law density distribution:
ρ = ρ0(r/R0)−ng cm−3
2002ic (red filled circles ).
Model light curve (black thick line) compared with the observation of SN
where the parameters are the radius (R0) and density (ρ0) of the point where the
ejecta and CSM start interacting, and the index (n) of the density distribution of
CSM. These parameters are constrained from comparison with the observed light
curve. The spherical Lagrange hydrodynamical code and input physics are the
same as in Suzuki & Nomoto .
When the expanding ejecta interacts with CSM, the interaction creates
the forward shock which is propagating through the CSM and the reverse shock
which is propagating through the ejecta.
Shocked matter is heated to T ∼ 107K for the reverse shock and T ∼
K for the forward shock. Both shocked regions emit thermal X-rays. For
the reverse shock, because of relatively high densities in the ejecta, cooling time
scale is shorter than shock propagation so that the shocked ejecta soon forms
a dense cool shell . This dense cool shell absorbs the X-ray and re-emits in
UV-optical. This re-emitted photons are observed. We assume that a half of the
X-rays emitted in the reverse-shocked ejecta is lost into the supernova center, and
that the other half is transferred outwardly through the cooling shell. We also
assume that a half of the X-rays emitted in the CSM is transferred inwardly to
be absorbed by the cooling shell. We take into account the change in time of the
column density of the cooling shell to evaluate the X-rays absorbed by the shell
and the optical luminosity.
Figure 3 shows the successful model for the light curve of SN 2002ic. Here
R0= 2 × 1014cm, ρ0= 4 × 10−13g cm−3, and n = 1.8 for inner CSM of 4 M⊙.
In the early phase, the model with n = 2.0 (steady mass loss) declines too fast to
-2-1 0 1 2 3 4 5
Velocity profiles in the interacting ejecta and CSM.
be compatible with the observation. This implies that CSM around the SN was
created by unsteady mass loss of the progenitor system.
After day ∼ 350, the light curve starts declining. To reproduce the declin-
ing part of the light curve, we add the outer CSM of 0.7 M⊙ where the density
declines sharply as n = 6. This implies that the total mass of CSM is ∼ 4.7 M⊙.
We note in Figure 4 that the velocity of the ejecta decelerated by the CSM-
interaction is <
spectral features (∼ 10,000 km s−1). On the other hand, in order to produce high
enough luminosity to explain the light curve, such a strong interaction between
the ejecta and CSM should occur.
To reproduce both the light curve and the observed velocity of SN 2002ic,
CSM needs to be aspherical. Suppose the CSM is aspherical consisting of a dense
region and a thin region. The expanding ejecta interacting strongly with the
dense region can produce high enough luminosity to explain the light curve. On
the other hand, the ejecta interacting with the thin region can expand still fast
enough to be consistent with the observed velocities (see also Deng et al. ).
A pre-existing clumpy disk was also suggested by Wang et al. , based on
4000 km s−1and too low for the value observed in the broad
There are two possible progenitor scenarios for SN 2002ic. One is the the
explosion of the C+O core of the massive AGB star (SN I+1/2), where the wind
from the AGB star formed the CSM. The other is the explosion of the white dwarf
in a close binary blowing wind to create the dense CSM (e.g., [9, 2]).
4.1. Type I+1/2 Supernovae in AGB Stars
Single star scenario is the explosion of the massive AGB star whose C+O
core becomes very close to the Chandrasekhar mass. Before explosion, mass loss
(super-wind) from the star creates a dense CSM. The C+O core explodes, which
is called Type I+1/2 supernova, and interacts with CSM.
To make this scenario possible, the metallicity of the system should be
low because low mass loss rate is necessary for the C+O to grow to reach the
Chandrasekhar mass before the envelope is completely lost.
metallicity, SN I+1/2 have never been observed. Therefore, we can explain the
rarity of SN 2002ic-like event assuming that only narrow mass range of AGB stars
can explode as SNe in low metal environment.
Aspherical CSM is not unexpected for stars approaching the end of the
Under the solar
4.2.White Dwarf Winds
Binary star scenario is the explosion of the accreting C+O white dwarf
(same as typical SNe Ia). However, the companion star is massive and the white
dwarf blows large amount of accreting gas as accretion wind to create the dense
CSM. In this scenario, the rarity is can be attributed to the fewness of the com-
panion star massive enough to produce the quite massive CSM.
As a progenitor of SN 2002ic, we need a CSM of ∼ 4.7 M⊙. Such a massive
CSM is possible only when the donor is as massive as 6 − 7 M⊙. For the model
consisting of a white dwarf and a main-sequence companion , the mass transfer
rate from such a massive main-sequence companion reaches ∼ 1 × 10−4M⊙yr−1.
Then the white dwarf blows a wind of ∼ 1×10−4M⊙yr−1and the mass stripping
rate becomes several times larger than the white dwarf wind mass loss rate .
For the symbiotic model consisting of a white dwarf and a red giant or AGB
star, the wind mass loss rate can also reach ∼ 1 × 10−4M⊙yr−1. In symbiotic
stars, the mass capture efficiency by the white dwarf is observationally estimated
to be as small as one or a few percent. Therefore, only when a large part of the
red giant wind or AGB super-wind is captured by the white dwarf, the white
dwarf can blow a very massive wind of ∼ 1 × 10−5M⊙yr−1or more. Then, the
mass stripping rate from the red giant or AGB star also reaches several times
Examples of the accretion wind evolution are identified as transient super-
soft X-ray sources, i.e., the LMC supersoft X-ray source RX J0513.9−6951 and
its Galactic counterpart V Sge . Especially in V Sge, a very massive wind of
∼ 1×10−5M⊙yr−1has been observationally suggested by the detection of radio.
Furthermore, the white dwarf wind collides with the companion and strips heavily
off its surface matter. This stripping rate reaches a few or several times the wind
mass loss rate of the white dwarf, i.e., ∼ 1×10−4M⊙yr−1or more . The matter
stripped off has a much lower velocity than the white dwarf wind itself and forms
an excretion disk around the binary. Thus the model predict the coexistence of a
fast white dwarf wind blowing mainly in the pole direction and a massive disk or
a torus around the binary. Deng et al.  propose a new classification, Type IIa
supernovae, for these events.
 Branch, D., Livio, M., Yungelson, L. R., Roffi, F. R., & Baron, E. 1995,
PASP 107, 717
 Chugai, N. N., & Yungelson, L. R. 2004, Astronomy Letters 30, 65
 Deng, J., Kawabata, K. S., Ohyama, Y., Nomoto, K., Mazzali, P., Wang, L.,
Jeffery, D. J. Iye, M., Tomita, H., & Yoshii, Y., 2004, ApJ 605, L37
 Germany, L. M., Reiss, D. J., Sadler, E. M., Schmidt, B. P., & Stubbs, C.
W. 2000, ApJ 533, 320
 Hachisu, I., Kato, M., & Nomoto, K. 1999a, ApJ 522, 487
 Hachisu, I., Kato, M., Nomoto, K., & Umeda, H. 1999b, ApJ 519, 314
 Hachisu, I., & Kato, M. 2004, ApJ 590, 445
 Hamuy, M., et al. 2003, Nature 424, 651
 Livio, M., & Riess, A. G. 2003, ApJ 594, L93
 Lundqvist, P., Sollerman, J., Leibundgut, B., Baron, E., Fransson, C., &
Nomoto, K. 2003, in From Twilight to Highlight: The Physics of Supernovae,
ed. W. Hillebrandt & B. Leibundgut (Berlin: Springer), 309
 Nomoto, K. 1982, ApJ 253, 798
 Nomoto, K., Thielemann, F.-K., & Yokoi, K. 1984, ApJ 268, 644
 Nomoto, K., et al. 2000, in Type Ia Supernovae: Theory and Cosmolgy, eds.
J. Niemeyer & J. Truran (Cambridge: Cambridge Univ. Press), 63
 Rigon, L., et al. 2003, MNRAS 340, 191
 Schlegel, D.J., Finkbeiner, D.P., Davis, M., 1998, ApJ 500, 525
 Suzuki, T., & Nomoto, K., 1995, ApJ 455, 658
 Turatto, M., Mazzali, P.A., Young,T.R., Nomoto, K., et al. 1998, ApJ 498,
 Wang, L., Baade, D., H¨ oflich, Wheeler, J. C., Kawabata, K., & Nomoto, K.
2004, ApJ 604, L53
 Wood-Vasey, W. M., et al. 2002, IAUC, 8019