From: Saul Perlmutter (S_Perlmutter@lbl.gov)
Date: Sat Jan 01 2005 - 16:22:21 PST
Hi Gabrele,
I *was* able to go through the 99ac paper, and make fixes (primarly
to the language). I think in general the paper looks quite good now!
A few things left for you to do: Do a "diff" on this draft to
compare it with the previous one to make sure that I didn't
misunderstand any of your sentences when I corrected them. Then look
for the places where I put questions/comments in triple
brackets/asterixes, "[[[" and/or "***", and let me know what the
answers are for them.
I hope you have a great 2005! Regards, --Saul
Gabriele Garavini wrote:
>Dear Saul,
>
>I attache the latex file of the paper on 99ac. As we said during the
>phone-conference we need to review the language. Right after the phone
>conference I remembered that the paper has been extensively reviewed by
>Rollin, and he has already changed and corrected the English of most of
>the paper. So I would tend to think the sections that really need some
>rewording are those I've added lately. These are section 5.1 and 6.
>
>Since you are the internal reviewer I thought I would make sense for you
>to have a look first. I think we could start with those two sections. It
>would be great if this could be done in the coming two weeks or so. In
>principle after those sections are corrected the paper could go out to the
>collaboration.
>
>This would really put the paper on a fast-track for publication. I'll be
>in Berkeley from the 19th to the 27th of January, in those days we could
>wrap up everything and get the papers (also the evolution one) as close to
>be ready for submission as possible. I would hope to submit them at the
>beginning of February or so.
>
>Thank you very much
>Merry Xmas to all of you and your families
>
>Cheers
>Gabriele
>
%-
%- File : sn99ac.tex
%- --------------------
%- Changed : Thu Dec 2 17:26:52 CET 2004
%- Authors : Gabriele Garavini (GG) - garavini@in2p3.fr
%- Rollin C. Thomas (RCT) - rcthomas@lbl.gov
%-
\documentclass[manuscript]{aastex}
% \documentclass[preprint2]{aastex}
% \documentclass[12pt,preprint]{aastex}
% \documentclass[letter]{aa}
\usepackage{natbib}
\bibliographystyle{aa}
\citestyle{aa}
\usepackage{epsfig,latexsym,graphicx,epsf,subfigure,verbatim}
\usepackage{threeparttable,hhline,longtable}
\usepackage[T1]{fontenc}
%- \usepackage{amsmath}
\usepackage{txfonts} %- RCT commented out on 6/2 -- doesn't have the pkg.
%- \linespread{1.6}
\newcommand{\rem}[1]{{\bf [#1]}} %- RCT added.
\newcommand{\wl}{$\lambda$} %- RCT added.
\newcommand{\kmps}{km s$^{-1}$} %- RCT added.
\newcommand{\halpha}{H$\alpha$} %- RCT added.
\newcommand{\rsi}{$\mathcal{R}($\ion{Si}{2}$)$} %- RCT added.
\newcommand{\vten}{$v_{10}($\ion{Si}{2}$)$} %- RCT added.
\newcommand{\dmft}{$\Delta m_{15}$} %- RCT added.
\newcommand{\vdot}{$\dot{v}$}
\def\RCS$#1: #2 ${\expandafter\def\csname RCS#1\endcsname{#2}}
\RCS$Revision: 1.21 $
\RCS$Date: 2004/09/22 19:48:35 $
\shorttitle{Spectroscopy of SN 1999ac}
\shortauthors{Garavini et al.}
\begin{document}
\title{The Unusual Type Ia SN~1999ac}
\author{
G.~Garavini\altaffilmark{1},
G.~Aldering\altaffilmark{2,3},
R.~Amanullah\altaffilmark{1},
P.~Astier\altaffilmark{4},
C.~Balland\altaffilmark{4,5},
G.~Blanc\altaffilmark{2},
M.~S.~Burns\altaffilmark{6},
A.~Conley\altaffilmark{2,7},
T.~Dahl\'en\altaffilmark{8},
S.~E.~Deustua\altaffilmark{2,9},
R.~Ellis\altaffilmark{10},
S.~Fabbro\altaffilmark{11},
X.~Fan\altaffilmark{12},
G.~Folatelli\altaffilmark{1},
B.~Frye\altaffilmark{2},
E.~L.~Gates\altaffilmark{13},
R.~Gibbons\altaffilmark{2},
G.~Goldhaber\altaffilmark{2,7},
B.~Goldman\altaffilmark{14},
A.~Goobar\altaffilmark{1},
D.~E.~Groom\altaffilmark{2},
D.~Hardin\altaffilmark{4},
I.~Hook\altaffilmark{15},
D.~A.~Howell\altaffilmark{2},
D.~Kasen\altaffilmark{2},
S.~Kent\altaffilmark{16},
A.~G.~Kim\altaffilmark{2},
R.~A.~Knop\altaffilmark{17},
B.~C.~Lee\altaffilmark{2},
C.~Lidman\altaffilmark{18},
J.~Mendez\altaffilmark{19,20},
G.~Miller\altaffilmark{},
M.~Moniez\altaffilmark{21},
M.~Mouchet\altaffilmark{22},
A.~Mour\~ao\altaffilmark{11},
H.~Newberg\altaffilmark{23},
S.~Nobili\altaffilmark{1},
P.~E.~Nugent\altaffilmark{2},
R.~Pain\altaffilmark{4},
N.~Panagia\altaffilmark{24},
O.~Perdereau\altaffilmark{21},
S.~Perlmutter\altaffilmark{2},
V.~Prasad\altaffilmark{2},
R.~Quimby\altaffilmark{2},
J.~Raux\altaffilmark{4},
N.~Regnault\altaffilmark{2},
J.~Rich\altaffilmark{25},
G.~T.~Richards \altaffilmark{26},
P.~Ruiz-Lapuente\altaffilmark{20},
G.~Sainton\altaffilmark{4},
B.~Schaefer\altaffilmark{27},
K.~Schahmaneche\altaffilmark{4},
E.~Smith\altaffilmark{17},
A.~L.~Spadafora\altaffilmark{2},
V.~Stanishev\altaffilmark{1},
R.~C.~Thomas\altaffilmark{2},
N.~A.~Walton\altaffilmark{28},
L.~Wang\altaffilmark{2},
W.~M.~Wood-Vasey\altaffilmark{2,7},
N.~Yasuda\altaffilmark{29},
T.~York\altaffilmark{2},\\
(THE SUPERNOVA COSMOLOGY PROJECT).
}
\altaffiltext{1}{Department of Physics, Stockholm University, Albanova University Center,
S-106 91 Stockholm, Sweden}
\altaffiltext{2}{E. O. Lawrence Berkeley National Laboratory, 1
Cyclotron Rd., Berkeley, CA 94720, USA }
\altaffiltext{3}{Visiting Astronomer, Cerro Tololo Interamerican
Observatory, National Optical Astronomy Observatory, which is operated
by the Association of Universities for Research in Astronomy, Inc.
(AURA) under cooperative agreement with the National Science Foundation.}
\altaffiltext{4}{LPNHE, CNRS-IN2P3, University of Paris VI \& VII,
Paris, France }
\altaffiltext{5}{Universit\'e Paris Sud, IAS-CNRS, B\^atiment 121, 91405
Orsay Cedex, France}
\altaffiltext{6}{Colorado College ~14 East Cache La Poudre St., Colorado
Springs, CO 80903}
\altaffiltext{7}{Department of Physics, University of California
Berkeley, Berkeley, 94720-7300 CA, USA}
\altaffiltext{8}{Stockholm Observatory, Albanova University Center, S-106 91 Stockholm, Sweden}
\altaffiltext{9}{American Astronomical Society, 2000 Florida Ave, NW,
Suite 400, Washington, DC, 20009 USA.}
\altaffiltext{10}{California Institute of Technology, E. California
Blvd, Pasadena, CA 91125, USA}
\altaffiltext{11}{CENTRA-Centro M. de Astrof\'{\i}sica and Department of
Physics, IST, Lisbon, Portugal }
\altaffiltext{12}{Steward Observatory, the University of Arizona, Tucson
, AZ 85721}
\altaffiltext{13}{Lick Observatory, P.O. Box 85, Mount Hamilton, CA 95140}
\altaffiltext{14}{Department of Astronomy, New Mexico State University,
Dept. 4500, P.O. Box 30001, Las Cruces, NM 88011}
\altaffiltext{15}{Department of Physics, University of Oxford, Nuclear
\& Astrophysics Laboratory Keble Road, Oxford, OX1 3RH, UK}
\altaffiltext{16}{Fermi National Accelerator Laboratory, P.O. Box 500,
Batavia, IL 60510}
\altaffiltext{17}{Department of Physics and Astronomy, Vanderbilt
University, Nashville, TN 37240, USA}
\altaffiltext{18}{European Southern Observatory, Alonso de C ordova
3107, Vitacura, Casilla 19001, Santiago 19, Chile }
\altaffiltext{19}{Isaac Newton Group, Apartado de Correos 321, 38780
Santa Cruz de La Palma, Islas Canarias, Spain}
\altaffiltext{20}{Department of Astronomy, University of Barcelona,
Barcelona, Spain }
\altaffiltext{21}{Laboratoire de l'Acc\'elerateur Lin\'eaire,
IN2P3-CNRS, Universit\'e Paris Sud, B.P. 34, 91898 Orsay Cedex, France}
\altaffiltext{22}{LUTH Observatoire de Paris, Section de Meudon, 92195
Meudon Cedex, France}
\altaffiltext{23}{Rensselaer Polytechnic Institute, Physics Dept.,
SC1C25, Troy NY 12180, U.S.A.}
\altaffiltext{24}{Space Telescope Science Institute, 3700 San Martin
Drive, Baltimore, MD 21218, USA}
\altaffiltext{25}{DAPNIA-SPP, CEA Saclay, 91191 Gif-sur-Yvette, France}
\altaffiltext{26}{University of Chicago, Astronomy \& Astrophysics
Center, 5640 s. Ellis Ave., Chicago IL 60637}
\altaffiltext{27}{University of Texas, Department of Astronomy, C-1400,
Austin, TX,78712, U.S.A.}
\altaffiltext{28}{Institute of Astronomy, Madingley Road, Cambridge CB3
0HA, UK }
\altaffiltext{29}{National Astronomical Observatory, Mitaka, Tokyo
181-8588, Japan}
\begin{abstract}
We present and interpret optical spectra of the peculiar type Ia
supernova (SN) 1999ac. The data extend from $-15$ to $+42$ days with
respect to B-band maximum, and reveal an event that is unusual in
several respects. Prior to B-band maximum, its spectra resemble those
of SN~1999aa (a slowly declining event) but possess stronger
\ion{Si}{2} and \ion{Ca}{2} signatures (more characteristic of a
spectroscopically normal SN). Spectra after B-band maximum appear
more normal. The expansion velocities inferred from \ion{Si}{2} are
among the slowest ever observed, though the event is not particularly
dim. Iron lines appears also to be at lower velocities than other
normal SNe, but the calcium line is at higher velocity than average.
The analysis of the parameters \vten, \rsi, \vdot, and
\dmft\ further underlines the unique characteristics of SN~1999ac.
We find convincing evidence of \ion{C}{2} \wl 6580 in the day $-15$
spectrum with ejection velocity $v > 16,000$ \kmps, but this signature
dissipates by day $-9$. This rapid evolution at early times
highlights the importance of extremely early-time spectroscopy.
\end{abstract}
%JD=1250.718 \citep{nicolas})
\maketitle
\section{Introduction}
\label{sec_intro}
The paucity of high-quality spectroscopy of Type Ia supernovae (SNe Ia)
at early phase (the first week or so after outburst) limits the
current understanding of these events. At early phase,
the density of the ejecta is obviously higher than at the phase of
maximum brightness and later. Two general implications follow from this
fact and conspire to permit observers to probe the outer layers of SN Ia
ejecta \citep{1992ApJ...387L..33R,1991A&A...245..114K}.
(1) A SN envelope at early phase is optically thick to a higher velocity
than at later time. Hence, the emergent spectrum (that being photons
collected from last scattering within the SN atmosphere) is imprinted
with information about the composition of the ejecta at high velocity.
(2) Line optical depth, which does the imprinting, is greater at early
phase than at maximum brightness and later. Neglecting effects from
ionization fronts, line opacity in a SN envelope drops as $t^{-2}$
\citep{1999ApJS..121..233H}. Hence, ions contributing only a weak
spectroscopic signature two weeks after outburst may present stronger
features one week after outburst.
Mapping the composition of SN Ia ejecta at high velocity leverages
other direct constraints placed on explosion models. Such constraints
include identifying the presence of ion signatures in SN spectra at
various velocity intervals, a nontrivial task since the corresponding
line profiles are often heavily blended. In particular, the
presence of carbon at high velocity is consistent with a
one-dimensional deflagration model
\citep{1984ApJ...286..644N,1985ApJ...294..619B}. Silicon-peak
elements at high velocity are more consistent with, for example, a
one-dimensional delayed-detonation model \citep{1991A&A...245..114K}. The
signature of unburned material at a variety of velocities discovered
across a sample of SNe Ia may support more recent three-dimensional
deflagration models \citep{2002A&A...391.1167R,2003Sci...299...77G}.
However, the task of obtaining early phase spectra presents numerous
logistical challenges. Efficient discovery of SNe very shortly after
outburst requires both high sensitivity and reliable rejection of false
positives. A spectrum for confirmation (that a candidate is a SN) and
classification must be obtained within a day. A program for intensive
followup may then be triggered, requiring coordination of limited
telescope time at possibly several different sites.
Seeking to augment the local SN Ia dataset used as calibrated candles
for cosmological distance measurement
\citep{1998AJ....116.1009R,1999ApJ...517..565P}, the Supernova
Cosmology Project (SCP) organized such a program together with
Several other teams [[identify the teams and/or put a citation here]]
which ran during the
Spring of 1999 \citep{2000AIPC..522...75A, 2000AIPC..540..263N}.
[[[***Perhaps a sentence here to explain our SN discovery plan?***]]] When
SN~1999ac was reported \citep{1999IAUC.7114....1M}, the SCP had
many telescope nights pre-scheduled for spectroscopy and photometry to
observe this bright event. As a result, SN~1999ac provides some of
the earliest observed spectra of a SN Ia.
SN~1999ac (R.A. = 16$^{h}$07$^{m}$15.0$^{s}$ Decl. =
+07$^{d}$58$\arcmin$20$\arcsec$, equinox 2000.0) was discovered and
confirmed on unfiltered observations taken on Feb. 26.5 and 27.5 UT at
23".9 east and 29".8 south of the nucleus of its ScD host galaxy, NGC
6063 \citep{1999IAUC.7114....1M}. \citet{1999IAUC.7122....2P} reports that a
confirmation spectrum taken on Feb. 28 UT is similar to that SN~1999aa
with stronger \ion{Si}{2} \wl6355 and well defined \ion{Ca}{2} H\&K.
Optical light curves of SN~1999ac have been discussed in
\citet{Phillips:2002cg} and \citet{2003PASP..115..453L}. They report a
B-band light curve similar to that of SN~2002cx raising as fast as
SN~1991T and declining similarly to SN~1994D until two weeks after
maximum. The V-band light curve of SN~1999ac closely resembles that of
SN~2002cx until 30 days after maximum light. The similarities between
SN~1999ac and SN~2002cx break down in R-band and I-band.
Generally, SN~1999ac may be considered to be spectroscopically similar
to SN~1999aa \citep{2001ApJ...546..734L,2004AJ....128..387G}.
SN~1999aa is considered by some to be an ``intermediate'' SN Ia with
properties ``between'' those of spectroscopically normal SNe Ia and
those of the spectroscopically peculiar, bright SN~1991T \citep[for a
discussion of whether or not this has implications for progenitor
channels, consult][]{2001ApJ...546..734L, 2001PASP..113..169B}. Here
we present the collected spectra (our photometry data set will
appear in a later work) and discuss in detail the two earliest spectra
taken at days $-15$ and $-9$ with respect to the date of maximum
brightness.
The peculiarities of SN~1999ac, however, are not limited to
early epochs. The study of the ejecta geometry underlines [indicates?] calcium
lines being formed in high velocity layers and iron and silicon
lines being formed in low velocity layers. Furthermore, when
SN~1999ac is plotted in the parameter space of \rsi, \vten, \vdot,
and \dmft\ it stands out as an extreme among other well studied SNe.
The organization of this article is as follows. In
\S\ref{sec_reduct}, a description of the reduction scheme is given.
\S\ref{sec_spectra} presents the spectra of SN~1999ac, and includes
some empirical analysis of the early time spectra. Fits to the early
spectra produced with the SN spectrum synthesis code SYNOW appear in
\S\ref{sec_analysis}. A study of the ejecta geometry is carried out
in \S\ref{ejecta} and a comparison of several spectral indicators with
those other objects in \S\ref{parameter} \S\ref{sec_conclusion}
concludes the article.
\section{Data \& Reduction}
\label{sec_reduct}
%A more complete
%description of the data reduction method will appear in a subsequent
%publication. Here we include the more important details.
The data set consists of 13 optical spectra extending from day $-15$ to
day $+42$ (all phases in this work are expressed with respect to B-band
maximum). In most cases, the spectra were acquired using different
instrumental settings for the blue and red parts of the spectrum to
avoid possible second-order contamination. Hence, the fully reduced 13
spectra are the combinations of both blue and red parts. The observation
log appears in Table \ref{tabdata}.
%- About the two-dimensional spectra.
The data were reduced using standard IRAF routines. The two-dimensional
images were bias-subtracted and flat-fielded. The sky background was
fitted, subtracted, and extracted for systematics checks on the
wavelength calibration.
%- About the calibration of the one-dimensional spectra:.
Wavelength and flux calibration were applied to the one-dimensional
extracted spectra using calibration observations taken with the same
instrumental setting and during the same night as science observations.
The accuracy of the wavelength calibration was checked against the
extracted sky spectra and generally found to agree to within 2~\AA.
%- Not sure what this is about:...
An atmospheric extinction correction was applied via tabulated
extinction coefficients for each telescope used.
Figure \ref{99acfinding} shows the position of SN~1999ac in its host
galaxy NGC 6063, an ScD galaxy with a recession velocity of 2848 \kmps\
as determined from narrow \halpha\ emission \citep{1998A&AS..130..333T}. All
spectra presented in this paper have been shifted to rest frame using
this recession velocity.
%- Science-based corrections to the one-dimensional spectra.
According to \citet{1998ApJ...500..525S}, Galactic reddening in the
direction of SN~1999ac is $E(B-V) = 0.046$ mag. According to
\citet{Phillips:2002cg}, SN 1999ac's photometric evolution does not
follow the Lira-Phillips relation \citep{1999AJ....118.1766P}, making
an estimate of reddening with this technique problematic.
\citet{2001BAAS...33.1207.} tentatively derive a total extinction to SN~1999ac of
$A_V = 0.51$ mag. We apply the Galactic extinction correction
assuming $R_V = 3.1$, i.e. $A_V = 0.14$ \citep{1989ApJ...345..245C}.
All spectra of SN~1999ac presented carry this correction, but do not
include any attempt at host galaxy extinction correction. This has no
impact on the analysis presented, since we restrict focus to line
profiles, which cover wavelength scales smaller than those affected by
reddening. For a given line profile, however, the effect of reddening
can be considered as a systematic flux offset.
The amount of host galaxy light contamination is characterized by
$\chi^2$-fitting to the data a host galaxy spectrum contribution
together with a SN spectrum template. It was found to be negligible at
all epochs, and hence no host galaxy light subtraction is performed.
Telluric corrections are applied to spectra used for synthetic spectrum
analysis, using calibration spectra from standard stars in the vicinity
of SN 1999ac. The wavelength regions with telluric corrections are marked
in the figures, and their effect can be seen by comparing Figures 2 and 3,
which show the spectra before telluric correction, with the later figures.
\section{Spectra}
\label{sec_spectra}
The timespan and sampling frequency of the data set permits study of the
spectroscopic evolution of SN~1999ac from very soon after outburst to
seven weeks beyond B-band maximum light. Figure \ref{ac_evo} presents
the 13 fully reduced spectra, along with phases of observation for each.
The top spectrum, taken at day $-15$, is one of the earliest ever
obtained of a SN Ia.
The two earliest spectra are relatively featureless, though they clearly
show the ion signatures of a SN Ia. The more obvious absorption
features are at 4900, 6100, and 8100 \AA; due to \ion{Fe}{3} blends,
the \ion{Si}{2} \wl 6355 blend, and the \ion{Ca}{2} infrared (IR)
triplet (respectively). A weak absorption at 4200 \AA\ is probably also
due to \ion{Fe}{3}, while \ion{Si}{3} is responsible for a weak
absorption at 4400 \AA. Two very weak notches at 5200 and 5500 \AA\
hint at the presence of the \ion{S}{2} ``W'' feature. A small
depression just redward of the \ion{Si}{2} 6100 \AA\ absorption could be
due to \ion{C}{2}. We return to this issue later in this article.
The absorptions strengthen up to about day $+8$. The
\ion{Si}{2} \wl 6355 absorption shifts noticeably to the red during this
time. After day $+8$, the \ion{Ca}{2} IR triplet profile morphology
remains roughly constant. The \ion{S}{2} ``W'' feature strengthens at
maximum light, but shortly afterward disappears as it is replaced by
\ion{Fe}{2} and \ion{Na}{1} lines.
The absorption at 6100 \AA, due to \ion{Si}{2} \wl 6355, changes shape
between days $+8$ and $+24$ as \ion{Fe}{2} lines strengthen and
obliterate it through line blending. By day $+24$, four robust minima
have replaced the 6100 \AA\ absorption. Furthermore, the ostensible
\ion{Si}{2} emission peak appears to shift redward during this interval.
The simplest explanation for this behavior is not a real shift per se,
but rather an effect of blending, as the bluest part of the emission
feature is overcome by the reddest absorption notch. It is interesting
to note that \ion{Fe}{2} lines begin contaminating the \ion{Si}{2}
feature as early as day $+11$, somewhat earlier than usual in SNe Ia.
From day $+24$ onward, however, the spectrum basically does not evolve.
The transition to iron-peak species dominance is complete just three
weeks after maximum light, as numerous iron-peak lines dominate the
spectrum from the ultraviolet to the near IR, excluding the \ion{Ca}{2} H\&K and
IR triplet features.
\subsection{Early-Time Comparisons}
%- -----------------------------------
Generally, the post-maximum spectroscopic evolution of SN~1999ac
resembles that of a normal SN Ia. The early-time spectra (days -15 and
-9) present the unusual opportunity of comparison with
the small set of supernovae with similarly early spectra.
In Figure \ref{-15_comparison}, the day $-15$ spectrum is presented
along with early-time spectra of SNe 1991T \citep{1992ApJ...384L..15F}, 1999aa
\citep{2004AJ....128..387G}, 1990N \citep{1991ApJ...371L..23L}, and 1994D
\citep{1996MNRAS.278..111P}. Compared to the spectroscopically peculiar SNe~1999aa
and 1991T, SN~1999ac exhibits stronger \ion{Ca}{2} and \ion{Si}{2} \wl
6355 absorptions. On the other hand, its \ion{Fe}{3} 4400, 4100 \AA\
and \ion{Si}{3} \wl 4560 features are weaker.
A small flux depression in the day $-15$ spectrum is clearly visible to
the red of the \ion{Si}{2} 6100 \AA\ feature, echoing one visible in the
spectrum of SN~1990N. Though made uncertain by the presence of the
telluric absorption at 6900 \AA, another feature at 7000 \AA\ could be
common to both spectra as well. The first feature has been identified
as \ion{C}{2} \wl 6580 in SN~1990N \citep{2001MNRAS.321..341M}, in SN~1994D
\citep{1999ApJ...525..881H}, in SN~1998aq \citep{2003AJ....126.1489B}, and in SN~1999aa
\citep{2004AJ....128..387G}. The second feature could be due to another line
from the same ion, \ion{C}{2} \wl 7234.
Overall, the $-15$ day spectrum resembles that of SN~1990N more than it
does SN~1994D, with its more obvious (and characteristic) \ion{S}{2} and
\ion{Fe}{2} blends. In summary, at day $-15$, SN~1999ac possess
spectroscopic characteristics common to both brighter SNe like SN~1991T
and spectroscopically normal ones like SN~1990N at a similar phase.
The $-9$ day spectrum of SN~1999ac (Figure \ref{-15_comparison}) does
not extend to the \ion{Ca}{2} H\&K absorption. The two \ion{Fe}{3}
features are still weaker than in SN~1991T and possess rounded minima.
The contribution from \ion{Fe}{2} in the region between 4000 and 5000
\AA\ appears weaker than in SN~1994D and in SN~1990N. The absorption
feature present on the red edge of this line and that around 7000 \AA\
at day $-15$ are no longer evident.
\section{Synthetic Spectra}
\label{sec_analysis}
One of the more interesting ion signatures suggested in the previous
section is that of \ion{C}{2}. The presence of carbon lines and their
ejection velocities have potential impact on hydrodynamical explosion
models. Carbon at high velocity is consistent with both
one-dimensional \citep{1984ApJ...286..644N} and three-dimensional
\citep{2002A&A...391.1167R,2003Sci...299...77G} deflagration models.
An apparently generic result from the newer models is the mixing of
``fuel'' (carbon and oxygen) and ``ashes'' (products of
nucleosynthesis) at all velocities. Hence, the detection of carbon at
low velocity would favor the three-dimensional deflagrations.
To further explore the \ion{C}{2} signature and its behavior at early
times, in this section we compare synthetic spectra to those observed
before maximum light. Clearly, the ideal approach would be to {\it
invert} a SN spectrum in some way to yield a composition model.
Unfortunately, the SN atmosphere problem is an ill-posed inverse
problem. Instead, one of two approaches must be adopted.
One approach is {\it detailed} analysis, where the goal is to include
all relevant transfer physics (nonlocal thermodynamic equilibrium rates,
relativity, time-dependence, energy from radioactive decays, etc.) and
numerically simulate the emergent spectrum of a given hydrodynamical
model. This approach is appopriate for validating hydrodynamical
models, or suggesting adjustments to such models in the future. Though
powerful, detailed analysis codes \citep[for example, the
general-purpose PHOENIX code,][]{Hauschildt1999} consume months of
computer time for a single calculation.
Another, complimentary approach is {\it direct} analysis. The goal of
direct analysis is more empirical; to constrain the presence or absence
of ions in a spectrum and the ejection velocities of their parent atoms.
The task is generally nontrivial, since SN spectra consist of many
blends of lines which cannot be treated simplistically. Still,
approximate techniques are used to make the process fast and iterative,
but the results are powerful. Constraints from direct analysis are of
use to both detailed modelers and explosion modelers. In the latter
case, the results of direct analysis can rule out many hydrodynamical
models (in principle) before the extensive simulation calculations
are performed.
SYNOW \citep{2000PhDT.........6F} is a direct analysis code that generates
spectra based a simple, conceptual model of a SN appropriate during
the first few weeks to months after explosion. This model consists of a
blackbody-emitting, sharply defined photosphere surrounded by an
extended line-forming, pure scattering atmosphere. The entire envelope
is assumed to be homologously expanding. Line transfer is treated using
the Sobolev method \citep{1960mes..book.....S,1970MNRAS.149..111C,1990sjws.conf..149J} so line
opacity is parameterized in terms of Sobolev optical depth. Which ions
are used in the calculation is determined by experience, guided by
the SN ion signatures atlas of \citet{1999ApJS..121..233H}. For each ion
introduced, Sobolev optical depth as a function of velocity for a
``reference line'' (usually a strong optical line) is specified.
Optical depths in other lines of the ion are set assuming Boltzmann
excitation of the levels at temperature $T_{exc}$.
The parameters $v_{phot}$ and $T_{bb}$ set the velocity and blackbody
continuum temperature of the photosphere, respectively. For each ion,
optical depth $\tau$ and a minimum ejection velocity $v_{min}$ is
specified. Optical depth scales exponentially with velocity
***What does this phrase mean?: [[[according
to an $e$-folding velocity]]]*** up to a maximum velocity given by $v_{max}$.
If $v_{min} > v_{phot}$ for an ion, we refer to the ion as ``detached.''
A sharply defined, blackbody-emitting photosphere clearly cannot serve
as a perfect substitute for the processes of continuum formation in a SN
atmosphere. Hence, the synthesized continuum level may systematically
differ from that observed in some wavelength regions. Generally, a good
fit to the blue continuum results in a brighter red synthetic continuum.
This effect has little to no bearing on line identifications or velocity
inferences.
\subsection*{Day $-15$}
Figure \ref{synow-15} is a comparison of a synthetic spectrum to the
observed spectrum at day $-15$. Choosing $T_{bb} = 11,200$ K reproduces
a satisfactory reproduction of the overall continuum shape. We also
find that the choice of $v_{phot} = 13,000$ \kmps\ assists in producing
reasonable line profiles when other factors are taken into account for
each individual ion.
The \ion{Si}{2} optical depth profiles are detached, with $v_{min} =
14,200$ \kmps. This detachment moves the synthetic \ion{Si}{2} \wl 6355
absorption higher in velocity space so that it matches the observed
feature. It also flattens the corresponding emission feature to improve
the agreement. We regard the presence of \ion{Si}{2} in this spectrum
as definite.
\ion{Si}{3} optical depth must be capped at $v_{max} = 17,000$ \kmps\ to
match the feature usually associated with it at 4400 \AA. Interestingly,
a concomitant synthetic feature appears to reproduce a feature in the
near-infrared. Though the fit is not perfect here, we regard the
presence of \ion{Si}{3} in this spectrum as definite.
The velocity range in which we introduce \ion{Ca}{2} is mainly
constrained by the \ion{Ca}{2} IR triplet since the H\&K component is
missing from the data. Nevertheless, the agreement with the observed
\ion{Ca}{2} IR triplet and the falling edge of the observable H\&K
signature is convincing. The presence of \ion{Ca}{2} in this spectrum
is definite.
Introducing \ion{Fe}{3} provides a match to two observed features, one
at 4200 \AA\ and another at 4900 \AA. Like \ion{Si}{2}, the
corresponding optical depth profile must be detached (in this case to
$v_{min} = 14,500$ \kmps) to reproduce the features. Adding some
\ion{Mg}{2} improves the fit to the absorption at 4200 \AA.
The presence of \ion{Fe}{3} is definite.
As previously mentioned, the characteristic \ion{S}{2} feature does not
yet appear to be present at phase $-15$ days. Including some \ion{S}{2}
optical depth produces an absorption blend around 5300 \AA. Generally,
however, fits to the \ion{S}{2} blend using SYNOW do not reproduce the
observed features in most cases. Nevertheless, we regard the presence
of \ion{S}{2} in the day $-15$ spectrum as probable.
It is interesting that including some \ion{Ni}{3} to the synthetic
spectrum helps to account for the flux deficit to the blue of the
\ion{Fe}{3} 4900 \AA\ feature and near 5300 \AA. The appearance of this
ion in a spectrum is somewhat unusual, nevertheless the improvement
cannot be discounted. Conservately, however, we regard the presence of
\ion{Ni}{3} as possible.
The weak absorption at 6300~\AA\ is well matched by a detached
\ion{C}{2} (v$_{min}=16,000$~km s$^{-1}$) that may also contribute to a
feature near 4500\AA\ and perhaps at 7000~\AA. The wavelength regions
were \ion{C}{2} makes its contribution are highlighted in Figure
\ref{CII-15} showing the effect of the presence of this ion on the
synthetic spectrum. The good matching of the absorption feature at
6300~\AA\ and the possible contribution near 4500~\AA\ makes the
identification of \ion{C}{2} definite in the day $-15$ spectrum.
We have considered \ion{C}{3} to match the small notch on the red side
of the \ion{Si}{3} feature near 4500 \AA. However, we find the evidence
for \ion{C}{3} less convincing than that for \ion{C}{2}. The small
optical depth used to generate the \ion{C}{3} feature prevents the
apparearance of other, weaker \ion{C}{3} lines in the synthetic
spectrum. Hence, the only sign of \ion{C}{3} (an unusual
identification) would be this line by itself. Without concomitant
\ion{C}{3} lines, we consider this ion to be only a remote possibility.
The feature is probably not due to H$\beta$, since the spectrum shows no
other signs of hydrogen.
\subsection*{Day $-9$}
\label{-9}
Figure \ref{synow-9} presents the synthetic spectrum produced for
comparison with the day $-9$ spectrum. The fit parameters used appear
in Table \ref{table-9}. The photospheric velocity has been lowered to
$v_{phot} = 11,800$ \kmps, and the blackbody temperature increased to
$T_{bb} = 13,800$ K. For the most part, the same ions are used for this
fit as in the previous one. Note that at this phase, SYNOW is
overestimating the continuum through the \ion{Si}{2} 6100 \AA\
absorption and redward. However, this offset has no effect on line
identifications or ejection velocity interval measurements.
Ions definitely present in the spectrum include \ion{Si}{2} (though no
longer detached), \ion{Si}{3} and \ion{Fe}{3}. Again the fit to
\ion{S}{2} is problematic, but its identification based on the two
notches redward of 5200 \AA\ is not unreasonable. \ion{Ni}{3} seems to
remove some excess flux again near 4700 \AA, but it no longer is
sufficient to account for all of the flux deficit.
Most interestingly, it appears that between day $-15$ and day $-9$, the
\ion{C}{2} signature has dissipated. A small notch near 4500 \AA\ could
be due to a small amount of \ion{C}{3} at photospheric velocities, but
adding \ion{C}{3} optical depth only enhances the excess flux to the red
of this feature. For this reason, we consider the presence of
\ion{C}{3} in this spectrum as unlikely, though no alternative has been
identified.
\section{Ejecta geometry}
\label{ejecta}
We have seen in previous sections tentative evidences of C II lines
moving at velocities above 20,000 km/sec. The peculiarity of SN~1999ac
is not limited to the high velocity ejecta. Distinctive
characteristics are noticed in many other absorption features pointing
out an overall unusual ejecta geometry. This can be investigated, in
first approximation, by looking at the line profile of spectral features.
The characteristic line profile during the first several weeks after
outburst is a P-Cygni profile. We here briefly review the relevant
features of the P-Cygni line profile before presenting measurements
based on it.
A P-Cygni profile arises from a
configuration consisting of an extended, expanding line-forming region
surrounding an optically thick core.Consider such a configuration, observed in the frame coinciding with the
rest frame of the core's center. Material in front of the core (as
observed) moves toward the observer, and scatters radiation out of the
line of sight, resulting in an absorption feature blueshifted with
respect to the line rest wavelength. Material not in front of the core
(and expanding away from it) scatters radiation into the observer line
of sight, resulting in an emission feature peaked at the line rest
wavelength. Together, the blueshifted absorption and rest-wavelength
centered emission features form the P-Cygni profile.
The morphology of a P-Cygni absorption component provides an estimate of
the velocity interval in which the originating line forms. The blue
edge of the absorption feature forms at the highest ejection velocities
where the line is optically thick. In practice, the minimum of an
absorption profile is used to derive a characteristic ejection velocity
describing where in velocity space a line forms.
The strength of a line can easily influence its shape. A weak line
(with smaller line opacity) produces a sharp, robust minimum that can be
measured with little ambiguity. Conversely, a strong line produces a
more rounded absorption feature, and as line strength increases, the
position of the absorption minimum shifts to the blue. Hence, a
velocity measured from such a minimum is not necessarily representative
of the minimum ejection velocity of a line.
Two lines from the same parent ion, or {\it concomitant lines}, that are
closely spaced in wavelength (as in a doublet) may generally be treated
as a single line for the purposes of velocity estimation. On the other
hand, features from other ions blending with a given line present a
problem, as they can easily shift the position of the line minimum to
the blue or red. Therefore, care must be exercised when making
inferences from absorption features that may or may not be suffering
blending effects.
\subsection{Ejecta geometry comparison}
In the following paragraphs we compare the velocity field of some of
the characteristic features of SN~1999ac with those of other SNe known
for having a peculiar ejecta geometry, namely SN~2002cx
\citep{2003PASP..115..453L}, SN~1999aa \citep{2004AJ....128..387G} and
SN~2002bo \citep{2004MNRAS.348..261B}, and with that of SN~1994D
\citep{1996MNRAS.278..111P} which we here take as a prototype of
normal supernova. SN~2002cx is a well studied under-luminous
supernova with normal B-band light curve decline rate, spectral
signatures similar to SN~1991T but with spectral lines with low
expansion velocities. SN~2002bo had a normal decline rate but showed
spectra with higher than average expansion velocities similarly to
SN~1984A \citep{1989A&A...220...83B}. Finally, SN~1999aa was found to
have a slow light curve decline rate and very weak \ion{Si}{2}
absorption features with expansion velocities constant in time. Also,
its spectra had similarities with SN~1991T at early epochs, but
rapidly changed toward normal looking spectra just before maximum
light.
\subsubsection{Pre-maximum spectra}
The comparison of the spectrum of SN~1999ac at day -9 with those of
the SNe mentioned above is shown in Figure \ref{comp99ac-9_new}, {\it
Panel A}. The most noticeable differences lie in the region of
\ion{Fe}{3} and \ion{Si}{2} lines. To investigate such differences,
we show, in {\it Panel B} and {\it Panel C}, the comparison in velocity
space of the absorption features in the wavelength region respectively
around 4250 \AA\ and 6150 \AA.
In early spectra the absorption visible at 4250 \AA\ is generally due
mostly to \ion{Fe}{3}~\wl4404 but a contribution of \ion{Mg}{2}~\wl4481 is
expected and can vary from object to object. Normal SNe have a
stronger \ion{Mg}{2} component whereas SN~1991T-like SNe are dominated
by \ion{Fe}{3}. Among the SNe we compare, the \ion{Mg}{2}
contamination is probably more important for SN~1994D and
SN~2002bo. SN~1999ac appears to have approximately the same velocity
distribution as SN~2002bo and as SN~1999aa. Both SN~1994D and
SN~2002cx have lower velocities than our object. The actual ejecta
geometry is difficult to disentangle from this comparison because of
the uneven contamination of \ion{Mg}{2} among the different
objects. However, based on the analysis we performed in section
\ref{-9}, this absorption in SN~1999ac appears to be the result of
both \ion{Mg}{2} and a dominant component of low velocity
\ion{Fe}{3}. We will come back later to this point for more
discussion.
In {\it Panel C} we compare the line profile in velocity space of the
absorption feature at around 6150 \AA. This line is due to
\ion{Si}{2}~\wl6355 and it is usually the one suffering the least
blending among all the supernova spectral features. It is thus
considered the simplest to use for expansion velocity studies. At this
epoch SN~1999ac appears to have the line minimum at the same velocity
as SN~1994D but lower velocity for its blue edge. This is consistent
with the optical depth of \ion{Si}{2} being weak in the outermost
layer. We will come back to this point later in the
analysis. SN~2002bo and SN~2002cx show respectively a faster and
slower \ion{Si}{2} layer.
\subsubsection{Spectra at maximum}
The ejecta geometry of the same SNe at around maximum light is
compared in Figure \ref{comp99ac0_new}, {\it Panel A }. As in the
previous epoch the most noticeable differences are in the iron and
silicon absorption features. Furthermore, \ion{Ca}{2}~H\&K also
appears to have a different line profile in each supernova.
The line profile in velocity space of \ion{Ca}{2}~HK ({\it Panel B})
of SN~1999ac is comparable with that of SN~2002bo and has a single
minimum while SN~1994D and SN1999aa have double minima. At this epoch
the velocity of the minima are approximately all comparable with
differences of the order of a thousand kilometer per second. The only
exception is SN~2002cx for which this feature is much weaker and at
lower velocity.
\ion{Fe}{2}~\wl5083.4 ({\it Panel C}) is at lower velocity in SN~1999ac
than in SN~1994D, SN~2002bo and SN~1999aa, while only SN~2002cx shows
even lower values. This is consistent with the iron layer of SN~1999ac
being deeper into the atmosphere than that of the other objects as
already mentioned for the case of \ion{Fe}{3}~\wl4404. At this epoch
\ion{Si}{2}~\wl6355 ({\it Panel D)} also shows lower values than all the
other SNe analyzed but the case of SN~2002cx for which this line is
still considerably weaker.
\subsubsection{Late time spectra}
In late time spectra, showed in {\it Panel A} of Figure
\ref{comp99ac24_new}, the absorption features are mainly formed by
iron lines. The velocity distribution and relative abundances of iron
in the SNe analyzed are understandable by looking at the different
strength of the small absorptions and peaks in the spectra. In {\it
Panel C}, vertical dotted lines mark the position of the various iron
lines minima. SN~1999ac shows lower velocity compared to all SNe with
the usual exception of SN~2002cx. \ion{Si}{2} at this epoch is too
weak and blended into iron lines to disentangle its complete line
profile and is not analyzed. \ion{Ca}{2}~H\&K ({\it Panel B}) is
instead still strong and SN~1999ac shows the fastest absorption
feature among the analyzed objects while \ion{Ca}{2}~IR, {\it Panel
D}, appears only marginally faster.
\subsection{Velocity time evolution}
\label{cavel}
The most streaking peculiarity of SN~1999ac seems to be the
concomitance of high velocity calcium with low velocity iron and
silicon. This characteristic becomes more evident with time.
Table \ref{tabledata} lists the velocities inferred from
the minimum of the \ion{Ca}{2}~H\&K feature, together with
those measured for SN~1999aa \citep{2004AJ....128..387G}. Figure
\ref{CaHK_vel} shows these velocities as a function of time
compared with those of other well
observed SNe. The statistical uncertainty of the measurements is
negligible compared with that related to the estimate of the host
galaxy recession velocity. This is taken to be 300 \kmps. Spectra
obtained before maximum light do not include this feature, so at those
phases no measurement can be reported. After maximum, the \ion{Ca}{2}
H\&K velocity decreases monotonically in line with those measured from
other SNe, although the values from SN~1999ac are systematically
greater.
The evolution of the velocity inferred from the minimum of \ion{Si}{2}
\wl 6355 in Figure \ref{siII_vel} is even more peculiar. (The measured
values are reported in Table \ref{tabledata} together with those
measured for SN~1999aa.) The \ion{Si}{2}
velocity of SN~1999ac appears to be monotonically
decreasing with time as is usually the case for dimmer supernovae, like
SN~1991bg or SN~1999by. However, SN~1999ac was not intrinsically dim
(M$^{max}_{B}$=-18.98(39) \citet{2003PASP..115..453L}). Some events,
such as SN~2000cx or SN~1999aa, maintain a rather constant
\ion{Si}{2} velocity
as a function of time. It should be noticed that after day $+11$ the
points might become less meaningful because of possible blending with
\ion{Fe}{2}. Figure \ref{siII_doppler1} illustrates this ambiguity
clearly: By day $+24$, the entire feature can no longer be referred to
as simply coming from \ion{Si}{2} \wl 6355, since the peak of the
emission feature has shifted to the red of the rest wavelength (zero
velocity). Still, it is clear that the velocity evolution of the
feature to day $+11$ is at an extreme among SNe Ia.
\section{Type Ia SN parameter space.}
\label{parameter}
Type Ia supernovae are currently believed to be a multi parameter
class of objects. The standard paradigm is, however, to describe the
intrinsic spread of properties through one parameter: the light curve
width stretch, {\it s}, or equivalently, through the light curve
decline rate \dmft. This, indeed has proved to be successful in
reducing the intrinsic spread in brightness and in allowing the use of
type Ia SNe as distance indicators. Several other parameters have been
proposed with the goal of fully describing SNe~Ia and their parameter
space. In the following sections we compare the measurements of the
parameters \rsi\ \citep{1995ApJ...455L.147N}, \vten\
\citep{1993AJ....105.2231B} and \vdot\ \citep{2004astro.ph.11059B} for
SN~1999ac with those of the dataset presented in
\citet{2004astro.ph.11059B}. For completeness we report here also the
data for SN~1999aa analyzed in \citep{2004AJ....128..387G}. The
measured values for both supernovae are reported in Table
\ref{tabledata1}. The symbols used in the following figures are chosen
to be as those in \citet{2004astro.ph.11059B} to maintain the same
distinction among the three clusters found in their analysis.
\subsection{\dmft\ versus \rsi}
Figure \ref{siratio} is a plot of \dmft\ as a function of the
quantity \rsi\ for several SNe Ia including SN~1999ac.
\citet{1995ApJ...455L.147N} defined \rsi\ as the ratio of the depth
of the \ion{Si}{2}~5800 \AA\ absorption to that of the
\ion{Si}{2}~6100 \AA\ absorption. The authors theorize that the
observed correlation between \dmft\ and \rsi\ is driven by
temperature (and hence nickel mass). Thus, hotter and brighter events
tend to be characterized by a small \rsi\ value; cooler and dimmer
events are characterized by a larger \rsi\ value. The plot shows that
SN~1999ac is above the general trend. Note however, that including the
reddening estimate derived by \citet{2001BAAS...33.1207.} would
improve the agreement. [[[***This is puzzling: Why does extinction
strongly affect either of these measurements???***]]]
We choose not to include this extinction value
in our analysis since is not clear if the technique used to derived can
be applied to this object because of its asymmetric B-band light
curve. The parameter space of this diagram is further expanded by the
inclusion of the values for SN~1999aw, SN~1999bp and SN~1999aa for
which the correlation appears to be still valid.
\subsection{\dmft\ versus \vten}
\citet{2000ApJ...543L..49H} investigated the spectroscopic diversity
of SNe Ia by plotting values of \rsi\ against \vten\ (the blueshift
measured in the \ion{Si}{2} \wl 6355 feature at ten days after maximum
light). The authors reasoned that if SNe Ia were a one-dimensional
family based on the mass of synthesized $^{56}$Ni, then the two
observables would be correlated. Unable to discern such a
correlation, they reasoned that the differences reflect differences in
the explosion mechanism itself. In Figure \ref{dm15vsv10}, we plot
\dmft\ (a proxy for \rsi) against \vten, including observed values
from SN~1999ac, SN~1999aw, SN~1999bp and SN~1999aa. SN~1999ac
possesses a low value for \vten\, but a normal value of \dmft\
\citep{2003PASP..115..453L}, similar to that of SNe 1989B, 1994D, and
1996X, making SN~1999ac unique in falling outside the apparent
groupings. While SN 1999ac is
spectroscopically similar to SNe 1990N and 1999aa (except with lower
velocities), its placement on the diagram away from those events
indicates possibly even higher dimensions of diversity in these
objects. The parameter space of this diagram is further expanded by
the inclusion of the values for SN~1999aw, SN~1999bp and SN~1999aa,
though perhaps in directions that appear more consistent with the
trends among the groupings.
\subsection{\dmft\ versus \vdot}
The parameter \vdot\ was introduced in \citet{2004astro.ph.11059B} as
an estimate of the expansion velocity time derivative computed after
B-band maximum light. The authors found a weak correlation with
\dmft; Fast declining, under-luminous supernovae show large
\vdot. Slow declining supernovae show small \vdot\ while normal
supernovae can have both large and small values. In the \dmft\
versus \vdot\ plane (Figure 15) SN~1999ac falls on the high edge of the normal
supernova with the highest \vdot\ measured to date -- similar to that of
SN~1983G which had different \vten\ and \rsi. The parameter space of
this diagram is again expanded by the inclusion of the values for
SN~1999aa and SN~2000cx which are found to have very small values of
\vdot\ \citep{2004AJ....128..387G}.
\subsection{\vdot\ versus \vten}
The plot of \vdot\ versus \vten\, shown in Figure 16,was left unexplored in
\citet{2004astro.ph.11059B} but it is interesting to note that the
three SN groups identified by the authors still populate different
regions of this plot, meeting each other at intermediate values.
On this plot
underluminous supernova tend to have small values of \vten\ and high
values of \vdot. Normal supernovae populate the central part of the
plane while supernova with fast expansion velocity (i. g. high \vten)
have also high \vdot. SN~1999ac has the highest \vdot\ and the lowest
\vten\ among the supernova measured, making it similar to the
faint supernovae but
on the extreme end. However, as noted, SN~1999ac did not appear to be dim. It
should be noticed also that SN~2000cx falls in an unpopulated region
of the plot with high \vten\ and low \vdot.
\section{Conclusions}
\label{sec_conclusion}
We have presented spectroscopic observations of SN~1999ac from $-15$ to
$+42$ days with respect to B-band maximum light. The earliest spectra
are similar to those of SN~1999aa, but share some characteristics of
spectroscopically normal SNe like SN~1990N. Notable is the early
appearance of iron features in the spectrum at $+11$ days
and the lack of any real evolution in the features after $+24$ days.
Using synthetic spectra, we have unambiguously identified the presence
of \ion{C}{2} at high velocity in the earliest spectrum at $-15$ days.
However, just six days later, all significant traces of \ion{C}{2} have
disappeared. This indicates that obtaining spectra even earlier than
$-10$ days with respect to maximum may be required to reliably probe
the outer layers of SNe Ia. Obtaining this level of efficiency from
an observing program requires careful coordination, similar to that
achieved by the European Research Training
Network\footnote{http://www.mpa-garching.mpg.de/~rtn/} search for SNe Ia,
and the Nearby Supernova Factory \citep{2002SPIE.4836...61A}.
Comparing the spectra of SN~1999ac with those of other supernovae we
find an unusual ejecta geometry. Iron and silicon lines
appear to be formed in deeper atmosphere layers than the corresponding
lines for other supernovae with the exception of the extreme case of
SN~2002cx. Calcium lines, however, are found to be formed in high
velocity layers. The same trends are confirmed when analyzing the time
evolution of velocities as derived from the minimum of \ion{Ca}{2}
H\&K and \ion{Si}{2} \wl 6355. The former shows a trend consistent
with normal SNe Ia, though the values are slightly higher than
average, while the latter shows monotonically decreasing
values that follow the trend of under-luminous SNe Ia.
We have presented measurements for spectroscopically derived
observables such as \rsi, \vten, \vdot\ and \rsi. While the values
for SN~1999ac weakly support the correlation between \rsi\ and
\dmft, the position of SN~1999ac on the \dmft\-versus-\vten\ plane
is somewhat off the main trend. In the plane \vdot\ versus \dmft\
SN~1999ac falls in a region of normal supernova but has the highest
value of \vdot, while in the plane \vdot\ versus \vten\ it appears to
be on the extreme of faint supernovae with the highest \vdot\ and
lowest \vten. However, SN~1999ac was not reported as a dim supernova.
This analysis points out that SN~1999ac is unlike any other known
supernovae.
% [[[I’m not convinced the following sentence/paragraph adds anything
% to the paper:]]]
%Our findings support that type Ia supernova are a multi-parameter
%class of objects for which an exhaustive representation has still to
%be achieved.
\acknowledgements
The authors thank David Branch, Adam Fisher, and Rollin Thomas for
providing the SYNOW code. The research presented in this article made
use of the SUSPECT\footnote{http://www.nhn.ou.edu/$\sim$suspect} Online
Supernova Spectrum archive, and the atomic line lilst of
\citet{1993KurCD...1.....K}. This work is based on observations made
with: the Nordic Optical Telescope, operated on the island of La Palma
jointly by Denmark, Finland, Iceland, Norway, and Sweden, in the Spanish
Observatorio del Roque de los Muchachos of the Instituto de Astrofisica
de Canarias; the Apache Point Observatory 3.5-meter telescope, which is
owned and operated by the Astrophysical Research Consortium; the Lick
Observatory Shane 3.0-m Telescope; the Cerro Tololo Inter-American
Observatory 4-m Blanco Telescope; the European Southern Observatory 3.6m
telescope and the Kitt Peak National Observatory Mayall 4-m Telescope.
This work was supported in part by "The Royal Swedish Academy of
Sciences". G. Garavini acknowledges support from the Physics Division,
E.O. Lawrence Berkeley National Laboratory of the U.S. Department of
Energy under Contract No. DE-AC03-76SF000098. A. Mour\~ao acknowledges
financial support from Funda\c{c}\~ao para a Ci\^encia e Tecnologia
(FCT), Portugal, through project PESO/P/PRO/15139/99; S. Fabbro thanks
the fellowship grant provided by FCT through project
POCTI/FNU/43749/2001.
\bibliography{/data/snova/bibtex/bib}
\clearpage
\begin{figure*}
\centering
\includegraphics[width=16cm]{sn99ac_2d1.eps}
\caption{SN~1999ac in its host galaxy R.A. =
16$^{h}$07$^{m}$15.0$^{s}$ Decl. = +07$^{d}$58$^{m}$20$^{s}$ (equinox
2000.0). B-band image obtained at NOT on 1999 March 15 UT with
SN~1999ac indicated. The field is 6\arcmin.5 across.}
\label{99acfinding}
\end{figure*}
\clearpage
\begin{figure*}
\centering
\includegraphics[width=16cm]{sn99ac.new1.eps}
\caption{SN~1999ac spectral time evolution. Epochs referred to B-band
maximum light. The $\oplus$ symbol marks the atmospheric absorptions.}
\label{ac_evo}
\end{figure*}
\clearpage
\begin{figure*}
\centering\includegraphics[width=16cm]{comp_-15.eps}
\caption{The $-$15 and $-$9 days spectra of SN~1999ac together with
those of SN~1999aa, SN~1991T, SN~1990N and SN~1994D respectively
from
\citet{2004AJ....128..387G,1992ApJ...384L..15F,1991ApJ...371L..23L,1996MNRAS.278..111P}. Epochs
are quoted in the labels. Line identification are taken as in
\citet{1999AJ....117.2709L,2001PASP..113.1178L,1999MNRAS.304...67F,1996MNRAS.278..111P,1995A&A...297..509M,1993ApJ...415..589K,1992ApJ...397..304J}}.
\label{-15_comparison}
\end{figure*}
\clearpage
\begin{figure*}
\centering\includegraphics[width=16cm]{sn99ac_-15_insets.eps}
\caption{Synthetic spectra compared with SN~1999ac spectrum for
$-$15 days. Dashed line: best match synthetic spectrum. Solid line:
data. SYNOW parameters used are presented in table
\ref{table-15}. Ions responsible for features in the synthetic
spectrum are marked. Where a telluric feature has been removed,
the spectrum is marked with an Earth symbol in parenthesis.}
\label{synow-15}
\end{figure*}
\clearpage
\begin{figure*}
\centering\includegraphics[width=16cm]{sn99ac_-15_noCII1.eps}
\caption{Synthetic spectra (heavy solid line)
compared with SN~1999ac spectrum (light solid line) for
$-$15 days in the 4500~\AA\ (left panel) and 6150~\AA\ (right panel)
region. First model from the top: $\tau_{\rm CII}=0$;
Second model from the top: $\tau_{\rm CII}\neq0$.
SYNOW parameters used are presented in table
\ref{table-15}. Ions responsible for features in the synthetic
spectrum are marked.}
\label{CII-15}
\end{figure*}
\clearpage
\begin{figure*}
\centering \includegraphics[width=16cm]{sn99ac_-9_insets.eps}
\caption{Synthetic spectra compared with SN~1999ac spectrum for
$-$9.days Dashed line: best match synthetic spectrum. Solid line:
data. SYNOW parameters used are presented in table
\ref{table-9}. Ions responsible for features in the synthetic
spectrum are marked. Where a telluric feature has been removed,
the spectrum is marked with an Earth symbol in parenthesis.}
\label{synow-9}
\end{figure*}
\clearpage
\begin{figure*}
\centering \includegraphics[width=12cm]{comp99ac-9_new.eps}\\
\includegraphics[width=4cm]{comp99ac-9_newFeIII.eps}
\includegraphics[width=4cm]{comp99ac-9_newSiII.eps}
\caption{ {\it Panel A}: The $-$9 days spectrum of SN~1999ac
together with those of SN~1999aa, SN~2002cx and SN~1994D. Epochs are
quoted in the labels. Line identification are taken as in
\citet{1999AJ....117.2709L,2001PASP..113.1178L,1999MNRAS.304...67F,1996MNRAS.278..111P,1995A&A...297..509M,1993ApJ...415..589K,1992ApJ...397..304J}. {\it
Panel B}: Comparison of the same SNe in the region of
\ion{Fe}{3}~$\lambda$4404 in velocity space. {\it Panel C}:
Comparison of the same SNe in the region of
\ion{Si}{2}~$\lambda$6355 in velocity space.}
\label{comp99ac-9_new}
\end{figure*}
\clearpage
\begin{figure*}
\centering\includegraphics[width=12cm]{comp99ac0_new.eps}\\
\includegraphics[width=4cm]{comp99ac0_newCaHK.eps}
\includegraphics[width=4cm]{comp99ac0_newFeII.eps}
\includegraphics[width=4cm]{comp99ac0_newSiII.eps}
\caption{{\it Panel A}: The spectrum at maximum light of SN~1999ac
together with those of SN~1999aa, SN~2002cx and SN~1994D. Epochs are
quoted in the labels. Line identification are taken as in
\citet{1999AJ....117.2709L,2001PASP..113.1178L,1999MNRAS.304...67F,1996MNRAS.278..111P,1995A&A...297..509M,1993ApJ...415..589K,1992ApJ...397..304J}
{\it Panel B}: Comparison of the same SNe in the region of
\ion{Ca}{2}~H\&K in velocity space. {\it Panel C}: Comparison of the
same SNe in the region of \ion{Fe}{2}~$\lambda$5083.4 in velocity
space. {\it Panel D}: Comparison of the same SNe in the region of
\ion{Si}{2}~$\lambda$6355 in velocity space.}
\label{comp99ac0_new}
\end{figure*}
\clearpage
\begin{figure*}
\centering
\includegraphics[width=12cm]{comp99ac24_newFeII_bk.eps}\\
\includegraphics[width=4cm]{comp99ac24_newCaHK.eps}
\includegraphics[width=4cm]{comp99ac24_newFeII.eps}
\includegraphics[width=4cm]{comp99ac24_newCaIR.eps}
\caption{{\it Panel A}: The $+$24 days spectrum of SN~1999ac together
with those of SN~1999aa, SN~2002cx and SN~1994D. Epochs are quoted
in the labels. Line identification are taken as in
\citet{1999AJ....117.2709L,2001PASP..113.1178L,1999MNRAS.304...67F,1996MNRAS.278..111P,1995A&A...297..509M,1993ApJ...415..589K,1992ApJ...397..304J}
{\it Panel B}: Comparison of the same SNe in the region of
\ion{Ca}{2}~H\&K in velocity space. {\it Panel C}: Comparison of the
same SNe in the \ion{Fe}{2} lines region in wavelength space. {\it
Panel D}: Comparison of the same SNe in the region of \ion{Ca}{2}~IR
triplet in velocity space. }
\label{comp99ac24_new}
\end{figure*}
\clearpage
\begin{figure*}
\centering\includegraphics[width=16cm]{CaII_vel_1.eps}
\caption{Expansion velocities of SN~1999ac as inferred from the
minima of Ca~{\sc ii} H\&K compared with the values of other SNe
taken from
\citet{1994AJ....108.2233W,2004ApJ...613.1120G,1993ApJ...415..589K,1996MNRAS.278..111P,1999ApJS..125...73J}
and references therein. Values for SN~1999ac are marked as filled
circles. Measured values are reported in Table \ref{tabledata}
together with those measured for SN~1999aa.}
\label{CaHK_vel}
\end{figure*}
\clearpage
\begin{figure*}
\centering\includegraphics[width=16cm]{siII635_vel.1.eps}
\caption{Expansion velocities of SN~1999ac as inferred from the
minima of Si~{\sc ii}~$\lambda$6355 compared with the values of
other SNe taken from
\citet{1999AJ....117.2709L,2001PASP..113.1178L,2004ApJ...613.1120G,2001MNRAS.321..254S}
and references therein. Values for SN~1999ac are marked as filled
circles. Measured values are reported in Table \ref{tabledata}
together with those measured for SN~1999aa.}
\label{siII_vel}
\end{figure*}
\clearpage
\begin{figure*}
\centering\includegraphics[width=8cm]{siII_doppler1.eps}
\caption{Evolution of the \ion{Si}{2} \wl 6355 feature in Doppler
space. The velocity at the minimum of the absorption feature is at
13,000 \kmps\ at $-15$ days, and moves to 8,000 \kmps\ at $+11$ days
relative to maximum. By day +16, contamination by \ion{Fe}{2} lines
has clearly begun.}
\label{siII_doppler1}
\end{figure*}
\clearpage
\begin{figure*}
\centering\includegraphics[width=16cm]{siratio_stefano.eps}
\caption{Light curve decline rate \dmft\ versus \rsi\ for the SNe Ia
in \citet{2004astro.ph.11059B}. SN~1999ac generally supports the
trend toward higher luminosity at lower \rsi\ but with a smaller
\rsi\ with respect to normal SNe. Including a reddening estimate
from \citet{2001BAAS...33.1207.} would improve the agreement.
Measured values are reported in Table \ref{tabledata1} togheter with
those mesured for SN~1999aa.}
\label{siratio}
\end{figure*}
\clearpage
\begin{figure*}
\centering\includegraphics[width=16cm]{v10_stefano.eps}
\caption{Plot of \dmft\ versus \vten. SN 1999ac falls in a
relatively unpopulated region of the plot with low value of \vten.
Measured values are reported in Table \ref{tabledata1} togheter with
those mesured for SN~1999aa.}
\label{dm15vsv10}
\end{figure*}
\clearpage
\begin{figure*}
\centering\includegraphics[width=16cm]{vdot_stefano.eps}
\caption{Plot of \dmft\ versus \vdot. SN 1999ac generally falls in
the region of the plot where other normal SNe lay but has the
highest measured \vdot. Measured values are reported in Table
\ref{tabledata1} togheter with those mesured for SN~1999aa.}
\label{dm15vsvdot}
\end{figure*}
\clearpage
\begin{figure*}
\centering\includegraphics[width=16cm]{vdotv10_stefano.eps}
\caption{Plot of \vdot\ versus \vten. SN~1999ac falls in a
relatively unpopulated region of the plot with the highset \vdot\
and lowest \vten. SN~2000cx falls in a unpopulated region of the
plot with low \vdot\ and high \vten. Measured values are reported in
Table \ref{tabledata1} togheter with those mesured for SN~1999aa.}
\label{vdotvsv10}
\end{figure*}
\clearpage
\begin{table*}
\caption{Data set specifications.\label{tabdata}}
\begin{center}
\begin{tabular}{lrllcllr}
\tableline\tableline
JD & Epoch\tablenotemark{f} & Telescope &Instrument&$\lambda$ Range\tablenotemark{f}&$\langle
\Delta\lambda\rangle$$\tablenotemark{a}$\tablenotemark{f}&$\langle S/N \rangle$$\tablenotemark{b}$& Comments\\
-2400000 & ref $B_{max}$ & & &[\AA] &[\AA] \\
\tableline
51236.89 & -15 & APO & DIS & 3703-10307 & 6.8 & 125&5696.16 $\tablenotemark{c}$ \\
51240.95 & -9 & MDM 2.4m& MARK III &3827-8860 & 5.4 & 63& $\tablenotemark{e}$\\
51251.39 & +0 & ESO 3.5m &EFOSC &3331-7495 & 4.0 & 325&$\tablenotemark{d}$ \\
51253.84 & +2 & CTIO 4m &RCSP & 3235-9263 & 2.0 & 76& $\tablenotemark{d}$\\
51253.72 & +2 & NOT & ALFOSC & 3285-9655 & 6.2 & 176&5852.06$\tablenotemark{c}$\\
51259.88 & +8 & CTIO 4m &RCSP & 3227-9254 & 2.0 & 94 &$\tablenotemark{d}$\\
51262.89 & +11 & CTIO 4m &RCSP & 3254-9278 & 2.0 & 114&$\tablenotemark{d}$ \\
51267.87 & +16 & CTIO 4m &RCSP & 3239-9241 & 2.0 & 90 & $\tablenotemark{d}$\\
51275.98 & +24 & KPNO 4m &T2KB & 3029-10401 & 5.4 & 59 &$\tablenotemark{d}$ \\
51279.85 & +28 & ESO 3.6m &EFOSC & 3341-10255 & 4.2 & 237&7440.45$\tablenotemark{c,}$$\tablenotemark{d}$ \\
51282.90 & +31 & Lick 3m & KAST&3321-10483 & 3.2 & 46&5489.40$\tablenotemark{c}$\\
51284.84 & +33 & ESO 3.6m &EFOSC & 3392-10128 & 4.2 & 195&7363.23$\tablenotemark{c,}$$\tablenotemark{d}$\\
51290.84 & +39 & ESO 3.6m &EFOSC & 3344-10194 & 4.2 & 146&7435.50$\tablenotemark{c,}$$\tablenotemark{d}$\\
51293.97 & +42 & Lick 3m & KAST&3268-8002 & 2.1 & 49&5417.33$\tablenotemark{c}$\\
\tableline
\end{tabular}
\tablenotetext{a}{Average wavelength-bin size.}
\tablenotetext{b}{Average signal-to-noise ratio per wavelength bin.}
\tablenotetext{c}{Beginning of red channel, [\AA].}
\tablenotetext{d}{Negligible 2$^{nd}$ order contamination.}
\tablenotetext{e}{Possible 2$^{nd}$ order contamination above 7500 \AA.}
\tablenotetext{f}{Rest Frame.}
\end{center}
\end{table*}
\clearpage
\begin{table*}
\caption{Synow parameters for $-$15 days. The fit is shown in Figure \ref{synow-15}. v$_{phot}$=13000~km s$^{-1}$,
$T_{bb}=11200$ K.\label{table-15}}
\begin{center}
\begin{tabular}{llllll}
\tableline\tableline
Ion & $\tau$ & $v_{min}$&$v_{max}$&$T_{exc}$ &$v_{e}$\\
&&$10^3$~km$s^{-1}$&$10^3$~km$s^{-1}$&\small$10^{3}$K &$10^3$~km$s^{-1}$\\
\tableline
C~{\sc ii}&0.038&16&40&15&5\\
C~{\sc iii}&0.2&-&14.2&15&5\\
O~{\sc i}&0.2&-&40&15&5\\
Mg~{\sc ii}&0.15&-&40&15&5\\
Si~{\sc ii}&0.65&14.2&40&15&5\\
Si~{\sc iii}&0.42&-&17&15&5\\
S~{\sc ii}&0.2&-&17&15&5\\
Ca~{\sc ii}&1.5&16&40&15&5\\
Fe~{\sc iii}&0.55&14.5&18&12&5\\
Co~{\sc ii}&0.006&-&40&15&5\\
Ni~{\sc iii}&5&-&40&12&5\\
\hline
\end{tabular}
\end{center}
\end{table*}
\clearpage
\begin{table*}
\caption{Synow parameters for $-$9 days. The fit is shown in Figure \ref{synow-15}.v$_{phot}$=11800~km s$^{-1}$,
$T_{bb}=13800$ K.\label{table-9}}
\begin{center}
\begin{tabular}{llllll}
\tableline\tableline
Ion & $\tau$ & $v_{min}$&$v_{max}$&$T_{exc}$ &$v_{e}$\\
&&$10^3$~km$s^{-1}$&$10^3$~km$s^{-1}$&\small$10^{3}$K &$10^3$~km$s^{-1}$\\
\tableline
C~{\sc ii}&0.015&16&40&12&5\\
C~{\sc iii}&0.75&-&12.8&12&5\\
O~{\sc i}&0.1&-&40&12&5\\
Mg~{\sc ii}&0.2&-&40&12&5\\
Si~{\sc ii}&1.2&-&40&12&5\\
Si~{\sc iii}&0.6&-&16&12&5\\
S~{\sc ii}&0.2&-&17&12&5\\
Ca~{\sc ii}&1.5&16&40&12&5\\
Fe~{\sc iii}&0.65&-&18&12&5\\
Ni~{\sc iii}&7.0&-&40&12&5\\
Co~{\sc ii}&0.045&-&40&12&5\\
\hline
\end{tabular}
\end{center}
\end{table*}
\clearpage
\begin{table*}
\caption{Measurements of the expansion velocity inferred from
\ion{Ca}{2}~H\&K and \ion{Si}{2}~$\lambda$6355 for SN~1999ac and
SN~1999aa \citep{2004AJ....128..387G}. The uncertainties on the velocity
measurements are assumed to be 300 \kmps, for further information see section \ref{cavel}. \label{tabledata}}
\begin{center}
\begin{tabular}{ccc|ccc}
\tableline\tableline
&SN~1999ac&&&SN~1999aa&\\
\hline
Epoch&\ion{Ca}{2}~H\&K&\ion{Si}{2}&Epoch&\ion{Ca}{2}~H\&K&\ion{Si}{2}\\
days&[\kmps]&[\kmps]&days&[\kmps]&[\kmps]\\
\hline
-15& - & 13409 & -11 & 20155 & - \\
-9 & - & 12169 & -7 & - & 10519 \\
0 & 14045 & 9801 & -3 & 16138 & - \\
2 & 13907 & 9488 & -1 & 16531 & 10060 \\
8 & 14000 & 8844 & 5 & - & 10233 \\
11 & 12999 & 8269 & 14 & 12730 & 10083 \\
16 & 13156 & 7410 & 19 & 12710 & 9928 \\
24 & 12687 & 6572 & 25 & 12114 & 10085 \\
28 & 12331 & 6361 & 28 & 11874 & 10107 \\
31 & 12167 & - & 33 & 11066 & 10922 \\
33 & 12192 & - & 40 & 10747 & - \\
42 & 11465 & - & 51 & 10058 & - \\
\hline
\end{tabular}
\end{center}
\end{table*}
\begin{table*}
\caption{Measured values of M$_{B}^{max}$, \dmft, \rsi, \vten, \vdot\
for SN~1999ac and SN~1999aa. \label{tabledata1}}
\begin{center}
\begin{tabular}{ll|l}
\tableline\tableline
&SN~1999ac&SN~1999aa\\
\hline
M$_{B}^{max}$&-18.98(0.39)$\tablenotemark{a}$&-19.14(0.78)\\
\dmft&1.30(0.09)$\tablenotemark{a}$&0.85(0.08)\\
\rsi &0.098(0.030)&0.163(0.016)\\
\vten&8473(300)&10100(300)\\
\vdot&137(14)&4(14)\\
\hline
\end{tabular}
\tablenotetext{a}{As reported in \citet{2003PASP..115..453L}.}
\end{center}
\end{table*}
\end{document}
This archive was generated by hypermail 2.1.4 : Sat Jan 01 2005 - 16:22:31 PST