0705.4086/ms.tex
1: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2: %% Filename:      ms.tex
3: %% Author:        Wesley Ketchum <ketchum@nhn.ou.edu>
4: %% Created at:    
5: %% Modified at:   Mon May 28 13:39:22 2007
6: %% Modified by:   Eddie Baron <baron@nhn.ou.edu>
7: %% Version:       
8: %% Description:   
9: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
10: \documentclass[12pt,preprint]{aastex}
11: 
12: %\RequirePackage{xspace}
13: % \DeclareRobustCommand\xspace{\futurelet\next\@xspace}
14: %  \def\xspace{\futurelet\next\@xspace}
15: %  \def\@xspace{%
16: %   \ifx\@let@token\bgroup\else
17: %   \ifx\@let@token\egroup\else
18: %   \ifx\@let@token\/\else
19: %   \ifx\@let@token\ \else
20: %   \ifx\@let@token~\else
21: %   \ifx\@let@token.\else
22: %   \ifx\@let@token!\else
23: %   \ifx\@let@token,\else
24: %   \ifx\@let@token:\else
25: %   \ifx\@let@token;\else
26: %   \ifx\@let@token?\else
27: %   \ifx\@let@token/\else
28: %   \ifx\@let@token'\else
29: %   \ifx\@let@token)\else
30: %   \ifx\@let@token-\else
31: %   \ifx\@let@token\@xobeysp\else
32: %   \ifx\@let@token\space\else
33: %   \ifx\@let@token\@sptoken\else
34: %    \space
35: %    \fi\fi\fi\fi\fi\fi\fi\fi\fi\fi\fi\fi\fi\fi\fi\fi\fi\fi}
36: 
37: %\usepackage{lmacs}
38: % GENERAL DEFINITIONS
39: % Time-stamp: <14/08/2006 10:48 baron>
40: 
41: %\RequirePackage{xspace}
42: %\RequirePackage{amsmath}
43: %\RequirePackage{amsfonts}
44: %\RequirePackage{amssymb}
45: 
46: \def\ifundefined#1{\expandafter\ifx\csname#1\endcsname\relax}
47: 
48: % \newif\ifpdf
49: % \ifx\pdfoutput\undefined
50: %    \pdffalse      % we aren't running pdflatex
51: % \else
52: %    \pdfoutput=1   % we are running pdflatex
53: %    \pdftrue
54: % \fi
55: 
56: \usepackage{lscape}
57: 
58: \def\la{\mathrel{\hbox{\rlap{\hbox{\lower4pt\hbox{$\sim$}}}\hbox{$<$}}}}
59: \def\ga{\mathrel{\hbox{\rlap{\hbox{\lower4pt\hbox{$\sim$}}}\hbox{$>$}}}}
60: 
61: \newcommand{\be}{\begin{eqnarray}}
62: \newcommand{\ee}{\end{eqnarray}}
63: 
64: \ifundefined{ensuremath}\def\ensuremath#1{\relax\ifmmode{#1}}
65: \else${#1}$\fi\else\relax\fi
66: \ifundefined{nuc}\def\nuc#1#2{\relax\ifmmode{}^{#1}{\protect\textrm{#2}}
67: \else${}^{#1}$#2\fi}\else\relax\fi
68: %\ifundefined{nuc}\def\nuc#1#2{\relax\ensuremath{{}^{#1}}}%
69: %{\protect\textrm{#2}}\else${}^{#1}$#2\fi}\else\relax\fi
70: \def\doublespace{\setlength{\baselineskip}{23pt}}
71: \def\singlespace{\setlength{\baselineskip}{14pt}}
72: \newcommand{\etal}{et al.}
73: \newcommand{\gcm}{g~cm$^{-3}$}
74: \newcommand{\kmps}{\ensuremath{\mathrm{km}~\mathrm{s}^{-1}}}
75: \newcommand{\is}{\mathrm{s}$^{-1}$}
76: \newcommand{\iyr}{\mathrm{yr}$^{-1}$}
77: \newcommand{\msol}{\ensuremath{{\mathrm{M}_\odot}}}
78: \newcommand{\zsol}{\ensuremath{{\mathrm{Z}_\odot}}}
79: \newcommand{\foe}{\ensuremath{10^{51}}}
80: \newcommand{\nni}{\nuc{56}\mathrm{Ni}}
81: \newcommand{\xni}{\ensuremath{\mathrm{X}_{\mathrm{Ni}}}}
82: \def\ang{\ensuremath{\mathrm{\AA}}}
83: \def\Tmod{\ensuremath{T_{\mathrm{model}}}}
84: \def\Teff{\ensuremath{T_{\mathrm{model}}}}
85: \def\teff{\ensuremath{T_{\mathrm{model}}}}
86: \def\tstd{\ensuremath{\tau_{\mathrm{std}}}}
87: \def\Rzero{\ensuremath{R_0}}
88: \newcommand{\RSi}{$\Re_{Si}$}
89: \newcommand{\RCa}{$\Re_{Ca}$}
90: \newcommand{\vno}{\ensuremath{v_0}}
91: \def\alog#1{\times 10^{#1}}
92: \newcommand{\phx}{\texttt{PHOENIX}}
93: \newcommand{\snia}{SN~I\lowercase{a}}
94: \newcommand{\sneia}{SNe~I\lowercase{a}}
95: \newcommand{\gamray}{$\gamma$-ray}
96: \newcommand{\halpha}{H$\alpha$}
97: \newcommand{\hbeta}{H$\beta$}
98: %\newcommand{\articlesize}{
99: %\doublespace
100: %\setlength{\textwidth}{6.5in}
101: %\setlength{\textheight}{8.4in}
102: %\voffset=-0.75in
103: %\hoffset=-0.55in}
104: %\newcommand{\checkit}{\fbox{{\tiny$\surd$}}}
105: 
106: 
107: 
108: 
109: 
110: %Graphics
111: %\graphicspath{{./}}
112: %\DeclareGraphicsExtensions{{.eps},{.ps}}
113: 
114: %\received{} 
115: %\accepted{} 
116: %\journalid{}{} 
117: %\articleid{}{} 
118: \shortauthors{Ketchum, W. et~al.}
119: \shorttitle{Detailed Spectral Analysis of SN 1999dn}
120: 
121: 
122: \bibliographystyle{apj}
123: 
124: \begin{document}
125: 
126: 
127: \title{Detailed Spectral Analysis of the Type Ib Supernova 1999dn.
128: Paper I:  Hydrogen-free Models}
129: 
130: 
131: \author{ 
132:   Wesley Ketchum,
133:   E.~Baron\altaffilmark{1},
134:   and David Branch
135:  }
136: \affil{Homer L.~Dodge Department of Physics and Astronomy,
137:    University of Oklahoma, 440 West Brooks, Rm.~100, Norman, OK
138:    73019-2061 USA}
139: \email{ketchum,baron,branch@nhn.ou.edu}
140: 
141:  \altaffiltext{1}{Computational Research Division, Lawrence Berkeley
142:    National Laboratory, MS 
143:    50F-1650, 1 Cyclotron Rd, Berkeley, CA 94720-8139 USA}
144: 
145: \begin{abstract}
146: We present spectral fits to five epochs of the  typical Type Ib
147: supernova 1999dn using the generalized, non-LTE, stellar atmospheres
148: code \phx.  Our goal is threefold:  to determine basic physical
149: properties of the supernova ejecta, such as velocity, temperature, and
150: density gradients; to reproduce He I absorption lines by invoking
151: non-thermal excitation; and, to investigate possible spectral
152: signatures of hydrogen, especially a feature around 6200 \AA, which
153: has 
154: been attributed to high velocity \halpha.  Our models assume an
155: atmosphere with uniform composition devoid of any hydrogen.  Our
156: model spectra fit the observed spectra well, successfully reproducing most of
157: the features, including the prominent He I
158: absorptions.  The most plausible alternative to \halpha~as the source
159: of the 6200 \AA~feature is a blend of Fe~II and Si~II lines, which can
160: be made stronger to
161: fit the observed feature better by increasing the metallicity of the ejecta.
162: High-metallicity models fit well at early epochs, but not as well as
163: solar-metallicity models after maximum light.  While this blend of
164: metal lines is a reasonable explanation of the source of the 6200
165: \AA~feature, it is still important to investigate hydrogen as the
166: source; therefore, a second paper will present models that include a
167: thin shell of hydrogen around the main composition structure.
168: \end{abstract}
169: 
170: \keywords{stars:  atmospheres---supernova:  individual (SN 1999dn)} 
171: 
172: 
173: \section{Introduction} 
174: 
175: Supernovae (SNe) are divided into two different types, I and II, based
176: on the presence of hydrogen in their optical spectra.  Type II
177: supernovae spectra have conspicuous hydrogen lines while Type I
178: supernovae spectra do not.  SNe~I can further be divided into three
179: more spectral types:  Ia, Ib, and Ic.  SNe~Ia characteristically
180: contain strong absorption lines caused by singly-ionized silicon;
181: SNe~Ib contain strong neutral helium absorption lines; and, SNe~Ic
182: lack both of these characteristics \citep{filarev97}.
183: 
184: Because of their almost exclusive location in the star-forming regions
185: of spiral galaxies, and because of their spectral properties, it is
186: believed that Type II, Ib, and Ic supernovae all result from the
187: core-collapse of massive stars.  Furthermore, studies have shown that
188: SNe Ib/c are likely to result from binary star systems \citep{heger03}.
189: 
190: Until recently, the very small number of SNe~Ib spectra available near
191: the period of maximum luminosity limited the study of these events in
192: detail.  However, a great number of Ib spectra have become available
193: thanks to the published spectra of \citet{math_sn1bc} and others.
194: Recent work with these spectra has shown that there may be traces of
195: hydrogen in SNe~Ib spectra, and of both hydrogen and helium in SNe~Ic
196: \citep{branIb02,branch_hIc06,elm_1bc06,parrent05bf07}.  This work used
197: SYNOW---a highly-parameterized, supernova spectrum-synthesis code that
198: assumes a sharp photosphere that emits a blackbody continuous spectrum
199: and line formation by resonance scattering, treated in the Sobolev
200: approximation---to calculate spectral models.  Other work on modeling
201: the spectra of SNe~Ib has also used SYNOW
202: \citep{deng00,benetal91D02,anupama05bf05}.
203: 
204: The speed and relative simplicity of SYNOW makes it very useful for
205: identifying spectral lines, line velocities, and other constraints on 
206: the ejected mater \citep{branIb02}.  However, these model spectra
207: do not naturally account for non-thermal excitation, which is required to see
208: He I absorption lines; the optical depth of He I absorptions is an adjustable
209: modeling parameter.  For this purpose, as well as to gain more
210: information about the physical properties of SNe Ib, we conduct a more detailed
211: spectral analysis using the generalized, non-LTE, stellar atmospheres
212: code \phx\ \citep{hbjcam99} to complement this work.  We choose to study SN~Ib
213: 1999dn:  a typical, well-covered Ib supernova
214: \citep{deng00, math_sn1bc}.  SN 1999dn has also been modeled
215: using SYNOW by \citet{deng00} and \citet{branIb02}.  It is
216: believed by both sets of authors that an absorption feature
217: around 6200 \AA~is most likely \halpha at early times, though they
218: leave open as possibilities Ne~I, Si~II, and C~II.  \citet{deng00}
219: argue that the feature later evolves into C~II in the post-maximum
220: light spectra; \citet{branIb02} suggest that the feature may be
221: attributed to lines of Fe~II.
222: 
223: We hope to accomplish the following in a two-paper study:  (1) determine
224: basic properties of the supernova such as effective temperature and
225: velocity of the photosphere, (2) determine if He~I absorption lines
226: can be properly formed by including non-thermal excitation, and (3)
227: study the issue of the existence of spectral signatures of hydrogen,
228: particularly determine the most plausible source of the 6200
229: \AA~feature.  Here we present synthetic spectra of SN
230: 1999dn assuming that no hydrogen is present; we use a homogeneous
231: helium core as a basic model atmosphere.  \S 2 contains an
232: overview of the supernova and the observed spectra, \phx\ code
233: calculations, and the basic parameters adjusted to better our fits.
234: \S 3 features our best fits for five of the observed spectra with a brief
235: analysis of each fit.  Also, we investigate some of the alternatives
236: to \halpha~for the 6200 \AA~feature.  In \S 4 we offer a discussion of
237: our results and a look forward to Paper II where we will add hydrogen
238: to the models.
239: 
240: \section{Methods}
241: 
242: SN~1999dn was discovered on August 20 by \citet{qiu99} in NGC 7714,
243: which has a recession velocity of 2700 \kmps~and an $E(B - V) = 0.052$
244: \citep{schlegelred98}.  In total, six spectra of SN 1999dn
245: have been published, ranging from ten days before until 38 days after
246: maximum brightness in the R-band, which occurred on August 31
247: \citep{math_sn1bc}.  The five spectra we model are shown in 
248: Figure~\ref{fig:all}.  While there was some early confusion about
249: its classification, being identified early as a Ia or Ic, SN~1999dn
250: contains conspicuous He I absorption lines, and so is classified as a
251: Type Ib \citep{math_sn1bc}.  \citet{branIb02} consider SN
252: 1999dn a typical SN~Ib, and so a study of its spectra should provide
253: insight into SNe~Ib as a whole.
254: 
255: We present spectral fits of the first five epochs of SN 1999dn
256: obtained from the multi-purpose stellar atmospheres program \phx\
257: \citep{hbjcam99,bhpar298,hbapara97,phhnovetal97,phhnovfe296}.  \phx\ solves  
258: the radiative transfer equation along characteristic rays  in
259: spherical symmetry, taking into account relativistic effects.  The
260: non-LTE (NLTE) rate equations for selected ionization states
261: are solved, thereby including effects of non-thermal
262: electrons from the $\gamma$-rays produced by the radioactive $\beta$ decay
263: of Ni-56, which are essential to making He I and any Ne I absorptions.
264: 
265: Table~\ref{tab:ions} lists the 27 ions that are treated in NLTE in our
266: models.  Each model atom includes primary NLTE transitions, used to calculate
267: the level populations and opacity, and weaker LTE transitions, which
268: are included in the opacity, but also which affect the rate equations
269: because of their effect on the solution to the transport equation
270: \citep{hbjcam99}.  All LTE line opacities not treated in NLTE
271: are treated with an equivalent two-level atom source function, using
272: a thermalization parameter, $\alpha$ = 0.05.  The atmospheres are
273: iterated to energy balance in the co-moving frame; while we neglect
274: explicit effects of time dependence on the transport equation, we
275: implicitly include them by including the $\gamma$-ray deposition rate in
276: the radiative equilibrium equation and in the rate equations for the
277: NLTE populations.
278: 
279: All of the models shown have a uniform composition structure with
280: $\gamma$-ray deposition that follows the density of the ejecta to
281: provide for the non-thermal excitation necessary to produce the
282: observed He I absorption lines (and any Ne I absorption as well).  The
283: ``standard composition'' models have solar metallicities, except
284: we assume all hydrogen has been burned to helium; the
285: ``high-metallicity'' models have three times solar metallicity.  The models are
286: parameterized by a few quantities:  the time since explosion (with an
287: assumed 12-day rise time, $t$), the effective temperature (\Tmod), the
288: photospheric velocity (the velocity, \vno, where the continuum optical depth
289: in extinction at 5000 \AA~, \tstd, is unity), the metallicity (given
290: as a factor $Z$ times solar), and the density profile's power law
291: index $n$, which follows the relation $\rho \propto r^{-n}$.
292: 
293: 
294: \section{Spectral Fits}
295: 
296: We produced spectra from the calculations of over 100 models, and
297: compared them to the observed spectra.  Table~\ref{tab:models} contains the
298: adjustable parameters for the best fits for all five epochs we modeled.  The
299: goodness of the fit is determined in a ``$\chi$-by-eye'' fashion; the
300: human eye is very adept at pattern recognition.  In general, we fit
301: the basic shape of the continuum by varying \Tmod, fit the Doppler shift
302: of the spectra by varying \vno~(we most often fit the 5876 \AA~He I
303: absorption), and fit the widths of the observed absorption lines by
304: varying $n$ and \vno.
305: 
306: Figure~\ref{fig:aug31_both} shows the day of maximum light
307: (epoch +0, obtained August 31) observed spectrum
308: and two model spectra (with parameters given in
309: Table~\ref{tab:models}).  For this
310: particular epoch we calculated 44 different models varying the basic
311: fitting parameters mentioned in the paragraph above, but mostly
312: varying the abundances of elements included in the calculations.
313: Identifications for the major features are shown in the figure,
314: corresponding to (reading the labels from left to right):  Ca~II H and
315: K, a multitude of Fe~II lines (with a small contribution from Mg~II,
316: most likely 4481 \AA;  overall it is difficult to uniquely
317: identify any one of the features due to the large amount of blending
318: in the spectra), He~I 5876 \AA, the unknown 6200 \AA~feature, He~I
319: 6678 \AA, He~I 7065 \AA, O~I 7772 \AA, and the Ca~II
320: infrared triplet.  While not shown, the He~I 10830 \AA~
321: absorption feature is very strong in the model spectra.
322: 
323: Overall, both models, the standard composition and high-metallicity,
324: fit the observed spectrum well.  The Ca~II H and K absorptions match
325: the observed spectrum very well, while the emission is slightly
326: overshot.  The (mostly) Fe~II absorption region is on the whole fit well
327: by both models, though perhaps slightly better by the standard
328: composition model as it has more subdued absorptions and emissions.
329: Neither model reproduces an absorption feature near 5100 \AA~in
330: the observed spectrum, presumably due to Fe~II, very well.  Both
331: models have an absorption too deep for the He~I 5876 \AA~
332: feature, though the standard composition has a well matched emission.
333: Both models show a feature that fits the 6200 \AA~feature, with
334: the higher-metallicity model getting the strength nearly perfect.
335: Both models then produce too strong an absorption for He~I 6678
336: \AA~and He~I 7065 \AA, and have trouble reproducing the O~I 7772 \AA~
337: feature well.  Unfortunately the range of the
338: observed spectrum does not extend far enough to include the Ca~II
339: infrared triplet.
340: 
341: As we can see in Figure~\ref{fig:aug31_both}, the simple non-thermal
342: excitation provided
343: in the models was enough to reproduce the neutral helium absorptions
344: fairly well.  Of particular
345: interest in this fit, however, is the good fit for the 6200 \AA~
346: feature, especially in the high-metallicity model.  We can see the
347: source of this feature in the model spectra in
348: Figure~\ref{fig:aug31_Fe2Si2}, which
349: contains the observed spectrum and standard composition model along
350: with single-ion spectra of Fe~II and Si~II produced from the standard
351: composition model.  The single-ion spectra show 
352: the spectral features due only to Fe~II or Si~II in the standard
353: composition model spectrum.  The 6200 \AA~feature in the model spectra is
354: the result of blending of Si II $\lambda$6355 \AA, which is formed a
355: little far to the blue, and an Fe II feature 
356: that is formed a little far to the red.  Together, they make the
357: feature seen in the standard composition model spectrum; their
358: strength is amplified by the increased abundance of both ions in the
359: high-metallicity model.
360: 
361: Figure~\ref{fig:aug31_alts} shows the observed spectrum from the +0
362: epoch with the
363: standard composition models and two other revised models:  one with
364: 10 times as much silicon abundance and one with 100 times as much
365: neon abundance.  We can see that the former revised model produces a
366: feature on the blue edge of the observed 6200 \AA~feature, but it does not
367: extend far enough to the red to be a real candidate for the source of
368: the absorption.  However, the latter revised model, with 100 times as much
369: neon as the standard composition, reproduces the 6200 \AA~feature very well.
370: The Ne I (6402 \AA) absorption line, produced like neutral
371: helium lines by non-thermal excitation, fits the feature remarkably
372: well, without introducing any unwanted features in other regions of
373: the spectrum.  Yet, such a large neon abundance does not seem reasonable.  This
374: could be an indication that the density profile for non-thermal
375: excitation is not as simple as we have modeled it, or it may be
376: coincidence.  While a model with enhanced carbon is not shown,
377: C II makes no obvious effects in the
378: standard composition model at this epoch, and so we disregard it as
379: the source of the 6200 \AA~feature as well.  The most plausible
380: identification from our models remains that of Si II and Fe II
381: blended together.
382: 
383: Figure~\ref{fig:aug21_both} shows the observed spectrum obtained ten
384: days before maximum
385: light (epoch -10, obtained on August 21) with two spectral fits:  a
386: standard composition
387: model and a high-metallicity model.  The observed spectrum contains
388: similar features to those in the maximum light spectrum.  The two
389: models fit most of these features fairly well.  The Ca~II H and K
390: absorptions are slightly too strong in the models; likewise in the
391: emissions.  Both models fit the observed spectrum fairly well over the
392: region dominated by Fe~II lines, between 4100 \AA~and 5200 \AA.  The
393: He~I 5876 \AA~absorption is fit particularly
394: well by the high-metallicity model, while the emission is better
395: modeled with the standard composition.  Both models produce a feature
396: near 6200 \AA.  That feature is stronger and better fit to the observed
397: spectrum  in the high-metallicity model; however, it is hardly a
398: poor fit of the feature using the standard composition.  The He~I 6678
399: \AA~feature is particularly weak in the observed spectrum, a subtlety the
400: models fail to replicate.  Both models also appear to form the He I
401: 7065 \AA~feature a little too far to the blue, which may be a
402: result of the simple $\gamma$-ray deposition that produces the
403: non-thermal excitation responsible for the formation of the neutral
404: helium lines in the models.  Likewise, the O~I absorption and the Ca~II
405: infrared triplet are formed a little blue in the model spectra, though
406: the features are otherwise fit reasonably well.
407: 
408: Moving to post-maximum light spectra, Figure~\ref{fig:sep10_both}
409: shows the observed spectrum
410: obtained ten days after maximum light (epoch +10, obtained September
411: 10) with two spectral
412: fits:  again, a standard composition model and a high-metallicity
413: model.  The observed spectral features are largely the same as those
414: seen in the two previous epochs.  Both fits are of
415: good quality, though the standard
416: composition model fits the observed spectrum better overall,
417: particularly at shorter wavelengths.  Both
418: models have too strong an absorption for Ca~II H and K, but their
419: emissions fit well.  In the Fe~II dominated region (which is now
420: becoming blanketed by Ti II lines as well, see
421: Figure~\ref{fig:sep17_LTE}) the
422: high-metallicity model fails to fit as well as the standard
423: composition, which in turn struggles to fit some of the observed
424: features at the correct strengths.  (However, as opposed to the +0 day
425: spectrum, the feature at 5100 \AA~fits much better.)  The
426: standard composition model fits the He I 5876 \AA~feature more
427: closely, especially on the blue edge, but the high-metallicity model
428: fits the other two He~I absorptions marginally better.  The 6200
429: \AA~feature at this later epoch is not nearly as pronounced and has
430: evolved considerably from the feature at earlier times; the feature
431: actually appears to be two distinct features centered around 6200
432: \AA.  This is not unlike the proposed explanation for the
433: feature at earlier epochs, except now the two features are
434: resolvable.  In the region around 6200 \AA, the standard composition model reproduces
435: features that are more than strong enough to fit the observed.
436: Again, the He~I 7065
437: \AA, O I, and Ca II infrared triplet absorptions are produced a
438: little to the blue in the model spectra
439: 
440: In Figure~\ref{fig:sep14_n} the observed spectrum obtained 14 days
441: after maximum light (epoch +14, obtained September 14) is shown with
442: two spectral fits. Both models have standard composition, but they
443: vary between $n = 13$ and $n = 10$.  Both models fit the observed
444: spectra fairly well.  One main difference between the two are the line
445: strengths, particularly in the Fe~II dominated region.  The $n = 10$
446: model fits the strength of the absorptions better at 5300 \AA~and 4200
447: \AA.  The shapes of some of the absorption lines, like He I 5876
448: \AA~are also fit slightly better in the $n=10$ model.  Similar
449: observations can be made about later epochs as well.  Therefore, for
450: this and later epochs the models will have $n=10$, rather
451: than $n=13$.
452: 
453: The September 14 observed spectrum is shown alongside two best fits
454: with varying metallicity in Figure~\ref{fig:sep14_both}:  one with the
455: standard composition and one with increased metallicity.  While both are good
456: fits overall, the standard composition model fits the observed
457: absorption depths better in the blue end of the spectrum, and
458: marginally worse in the redder end. Both models have absorptions and
459: emissions of the Ca II H and K lines that are stronger than
460: observed.  As mentioned above, the standard composition model fits the
461: absorption strengths in the Fe~II/Ti~II dominated region better than
462: the high-metallicity model.  Both models have strong absorption of He
463: I that fail to match the observed features perfectly. In the observed
464: spectrum barely anything remains of the 6200 \AA~feature; the models
465: produce features similar to those seen in the +10 day spectrum.  The
466: standard composition model fits the region much better because of the
467: decreased feature strength.  The models produce features that are at
468: the same wavelength of the observed spectrum for the He I 7065 \AA~and
469: O I lines; the Ca II infrared triplet absorption is only slightly too
470: blue, though a touch too strong, in the models compared to the
471: observed spectrum.
472: 
473: Figure~\ref{fig:sep17_both} shows the observed spectrum of SN~1999dn
474: obtained 17 days after maximum light (epoch +17, obtained on September
475: 17) with two spectral fits:  a standard composition model spectrum and a
476: high-metallicity model spectrum.  Both models fit the spectrum well.
477: The absorption and emission of the Ca~II H and K lines are fit
478: reasonably well in both models.  Neither model fits the Fe~II region
479: perfectly, but
480: both reproduce the features seen with varying degrees of accuracy:
481: the standard composition model fits the 4900 \AA~absorption perfectly,
482: while the high-metallicity model reproduces the features at 4400
483: \AA~and 5900 \AA~slightly better.  The standard composition model's
484: He~I 5876 \AA~absorption is slightly strong and blue, but the other
485: He~I absorption features are fit well.  Note that the features around
486: 6200 \AA~are in the observed spectrum again, looking much like they
487: did in the +10 day spectrum. The shape of these features are reproduced
488: better in the standard composition model, though the absorption
489: strengths are reproduced better in the high-metallicity model.  The
490: features in the model spectra that
491: appear in the same region can be attributed to a blending of Fe~II,
492: Si~II, and C~II.  The O~I feature is fit well by both models (though
493: slightly better by the standard composition), and both models produce
494: a too strong and too blue Ca~II infrared triplet feature.
495: 
496: As the temperature in the supernova has decreased substantially since
497: maximum light (our model temperatures decrease from 6000 K to 5000
498: K), we expect to see some new contributions to spectral features at
499: later times.  Figure~\ref{fig:sep17_LTE} shows the observed spectrum
500: for the +17 day epoch along with three single-ion spectra of lines
501: that have only been considered in LTE:  Ni II, Al I, and Ti II.   Ni
502: II and Al I both have some effect on the model spectra close to the Ca
503: II H and K absorption (near 3750 \AA).  Ti II, however, has significant
504: contributions to the spectrum all the way up to a wavelength of 5500
505: \AA, in what we
506: previously called the Fe~II dominated region.  The presence
507: of Ti II lines was noted as a possibility in SYNOW analysis of this
508: spectrum by \citet{branIb02}.  The presence of Ti II in our models
509: is perhaps part of
510: the reason why the high-metallicity models fail to reproduce this
511: region of the spectrum as well as the standard composition models: 
512: they have too much Ti II, causing their absorptions to be too strong.
513: Ti II also has a very small feature in the 6200 \AA~region in our
514: models, which may affect the evolution of that
515: feature.
516: 
517: Finally, we investigate the affects of increasing carbon abundance at
518: late times in Figure~\ref{fig:sep17_10C}.  \citet{deng00} claimed that
519: at +14 days C II was unambiguously in the spectrum.  Our late-time models show
520: some weak features of C II around the 6200 \AA~feature, but the effect
521: is minimal.  Increasing the carbon abundance by ten times in our +17
522: day standard composition model makes little effect on the overall
523: spectrum.  C II may be responsible for an observed feature near 9500
524: \AA~that is seen in the enhanced carbon model.  Unlike at earlier
525: epochs, near maximum light, C II has some affect on the shape of
526: particular features in the spectrum at late times, including around 6200
527: \AA, though its effect is small.
528: 
529: \section{Discussion}
530: Our goal in this series of papers is to (1) determine basic physical
531: properties of SN~1999dn at different epochs such as its effective
532: temperature and photospheric velocity, (2) determine if He I
533: absorption lines can be properly formed in a natural way by including
534: non-thermal excitation, and (3) determine whether there are spectral signatures
535: of hydrogen in this ``normal'' SN~Ib.  We have presented
536: models to fit five epochs of SN 1999dn that did not include hydrogen;
537: our models were homogeneous stellar atmospheres that had been
538: stripped of their hydrogen envelopes.  While this model is very
539: simple, our spectral fits are on the whole remarkably good by \phx\ standards.
540: Table~\ref{tab:models} contains a list of some of
541: the most basic physical parameters we determined, and in all epochs He
542: I absorptions were successfully produced in the models by including a
543: \gamray~deposition that followed the density of the ejecta that allowed for
544: non-thermal excitation.
545: 
546: Finally, we looked carefully at the prime
547: candidate for \halpha~absorption, a feature around 6200 \AA~in the
548: observed spectra, and tested alternative ways to reproduce the feature
549: in models.  The most successful, and a perfectly plausible,
550: alternative is an increased metallicity of the ejecta.  By increasing
551: the metallicity to three times solar, we were able to reproduce the
552: feature as a blend of Fe~II and Si~II lines well at early times
553: without affecting the rest of the spectrum too much.  At later times,
554: as the 6200 \AA~becomes weaker and the spectral contribution of metals
555: like Fe~II and Ti~II becomes more pronounced, the standard
556: composition models fit better, but we cannot rule out the
557: high-metallicity case.  We considered several other alternatives to the
558: identification of the 6200 \AA~feature:  Si~II alone created a feature
559: too far to the blue; C~II made no obvious effects in the standard
560: composition model at maximum light; and, Ne~I reproduced the feature,
561: but only after enhancing the neon abundance by a factor of 100.  Given
562: a uniform composition with \gamray~deposition following the density
563: profile of the ejecta, the most plausible alternative to
564: \halpha~appears to be a blending of Fe~II and Si~II, possibly
565: strengthened by increased metallicity.
566: 
567: We have noted that the 6200 \AA~feature evolves  with time.  At
568: early epochs it appears to be a single, strong  absorption feature, though this
569: single absorption could actually be a blend of more than one line, as
570: we showed in our high-metallicity models.  However, at later times,
571: the observed feature separates into multiple, distinct, weaker
572: features, possibly from  Fe~II, Si~II, C~II, and
573: Ti~II.  This change could be an indication that the 6200 \AA~feature
574: is \halpha~at early times, and is a blend of the above ions
575: at epochs after maximum light.  However, this evolution is also seen
576: in our models, suggesting that other physical properties play a role
577: in, and are perhaps important causes of the observed change.  Decreased
578: temperatures affect the shapes and strengths of the Fe~II and Si~II
579: lines, and allow other ions, like C~II and Ti~II, to become more
580: prominent in our models.  Decreased ejecta velocities cause less line
581: blending, allowing greater resolution of nearby features.
582: Furthermore, while our models do not include them as possibilities,
583: stratified compositions and density structures that do not follow a
584: single power law may affect the feature at later times as well.  With our
585: simple homogeneous parameterization, it is difficult to draw a strong
586: conclusion.  We hope to address the true nature and cause of this
587: evolution in Paper II.
588: 
589: In Paper II of this series, we will add a thin shell of hydrogen to
590: the outside of our current \phx\ models, and attempt to fine tune its velocity
591: coordinate and density profile to see if a model can reproduce the
592: 6200 \AA~feature.
593: 
594: \acknowledgements
595: This work was supported in part by NASA grants NAG5-3505 and
596: NNG04GB36G, and NSF grant AST-0506028.  This research used resources
597: of the National Energy Research Scientific Computing Center (NERSC),
598: which is supported by the Office of Science of the U.S. Department of
599: Energy under Contract No. DE-AC03-76SF00098.  This research has
600: made use of the NASA/IPAC Extragalactic Database
601: (NED) which is operated by the Jet Propulsion Laboratory, California
602: Institute of Technology, under contract with the National Aeronautics
603: and Space Administration.
604: 
605: \clearpage
606: %\bibliography{apj-jour,refs,baron,sn1bc,sn1a,sn87a,snii,stars,rte,cosmology,gals,agn,crossrefs}
607: 
608: \begin{thebibliography}{18}
609: \expandafter\ifx\csname natexlab\endcsname\relax\def\natexlab#1{#1}\fi
610: 
611: \bibitem[{Anupama {et~al.}(2005)Anupama, Sahu, Deng, Nomoto, Tominaga, Tanaka,
612:   Mazzali, \& Prabhu}]{anupama05bf05}
613: Anupama, G., Sahu, D., Deng, J., Nomoto, K., Tominaga, N., Tanaka, M., Mazzali,
614:   P., \& Prabhu, T. 2005, ApJ, 631, L125
615: 
616: \bibitem[{Baron \& Hauschildt(1998)}]{bhpar298}
617: Baron, E. \& Hauschildt, P.~H. 1998, ApJ, 495, 370
618: 
619: \bibitem[{Benetti {et~al.}(2002)Benetti, Branch, Turatto, Cappellaro, Baron,
620:   Zampieri, Valle, \& Pastorello}]{benetal91D02}
621: Benetti, S., Branch, D., Turatto, M., Cappellaro, E., Baron, E., Zampieri, L.,
622:   Valle, M.~D., \& Pastorello, A. 2002, MNRAS, 336, 91
623: 
624: \bibitem[{{Branch} {et~al.}(2006){Branch}, {Jeffery}, {Young}, \&
625:   {Baron}}]{branch_hIc06}
626: {Branch}, D., {Jeffery}, D.~J., {Young}, T.~R., \& {Baron}, E. 2006, PASP, 118,
627:   791
628: 
629: \bibitem[{Branch {et~al.}(2002)}]{branIb02}
630: Branch, D. {et~al.} 2002, ApJ, 566, 1005
631: 
632: \bibitem[{{Cardelli} {et~al.}(1989){Cardelli}, {Clayton}, \&
633:   {Mathis}}]{cardelli89}
634: {Cardelli}, J.~A., {Clayton}, G.~C., \& {Mathis}, J.~S. 1989, \apj, 345, 245
635: 
636: \bibitem[{Deng {et~al.}(2000)Deng, Qiu, Hu, Hatano, \& Branch}]{deng00}
637: Deng, J.~S., Qiu, Y.~L., Hu, J.~Y., Hatano, K., \& Branch, D. 2000, ApJ, 540,
638:   452
639: 
640: \bibitem[{{Elmhamdi} {et~al.}(2006){Elmhamdi}, {Danziger}, {Branch},
641:   {Leibundgut}, {Baron}, \& {Kirshner}}]{elm_1bc06}
642: {Elmhamdi}, A., {Danziger}, I.~J., {Branch}, D., {Leibundgut}, B., {Baron}, E.,
643:   \& {Kirshner}, R.~P. 2006, A\&A, 450, 305
644: 
645: \bibitem[{Filippenko(1997)}]{filarev97}
646: Filippenko, A.~V. 1997, Ann. Rev. Astr. Ap., 35, 309
647: 
648: \bibitem[{Hauschildt \& Baron(1999)}]{hbjcam99}
649: Hauschildt, P.~H. \& Baron, E. 1999, J. Comp. Applied Math., 109, 41
650: 
651: \bibitem[{Hauschildt {et~al.}(1997{\natexlab{a}})Hauschildt, Baron, \&
652:   Allard}]{hbapara97}
653: Hauschildt, P.~H., Baron, E., \& Allard, F. 1997{\natexlab{a}}, ApJ, 483, 390
654: 
655: \bibitem[{Hauschildt {et~al.}(1996)Hauschildt, Baron, Starrfield, \&
656:   Allard}]{phhnovfe296}
657: Hauschildt, P.~H., Baron, E., Starrfield, S., \& Allard, F. 1996, ApJ, 462, 386
658: 
659: \bibitem[{Hauschildt {et~al.}(1997{\natexlab{b}})Hauschildt, Schwarz, Baron,
660:   Starrfield, Shore, \& Allard}]{phhnovetal97}
661: Hauschildt, P.~H., Schwarz, G., Baron, E., Starrfield, S., Shore, S., \&
662:   Allard, F. 1997{\natexlab{b}}, ApJ, 490, 803
663: 
664: \bibitem[{{Heger} {et~al.}(2003){Heger}, {Fryer}, {Woosley}, {Langer}, \&
665:   {Hartmann}}]{heger03}
666: {Heger}, A., {Fryer}, C.~L., {Woosley}, S.~E., {Langer}, N., \& {Hartmann},
667:   D.~H. 2003, ApJ, 591, 288
668: 
669: \bibitem[{Matheson {et~al.}(2001)Matheson, Filippenko, Li, Leonard, \&
670:   Shields}]{math_sn1bc}
671: Matheson, T., Filippenko, A.~V., Li, W., Leonard, D., \& Shields, J. 2001, AJ,
672:   121, 1648
673: 
674: \bibitem[{{Parrent} {et~al.}(2007){Parrent}, {Branch}, {Troxel}, {Casebeer},
675:   {Jeffery}, {Ketchum}, {Baron}, {Serduke}, \& {Filippenko}}]{parrent05bf07}
676: {Parrent}, J., {Branch}, D., {Troxel}, M.~A., {Casebeer}, D., {Jeffery}, D.~J.,
677:   {Ketchum}, W., {Baron}, E., {Serduke}, F.~J.~D., \& {Filippenko}, A.~V. 2007,
678:   PASP, 119, 135
679: 
680: \bibitem[{Qiu {et~al.}(1999)Qiu, Li, Qiao, \& Hu}]{qiu99}
681: Qiu, Y., Li, W., Qiao, Q., \& Hu, J. 1999, AJ, 117, 736
682: 
683: \bibitem[{Schlegel {et~al.}(1998)Schlegel, Finkbeiner, \&
684:   Davis}]{schlegelred98}
685: Schlegel, D., Finkbeiner, D., \& Davis, M. 1998, ApJ, 500, 525
686: 
687: \end{thebibliography}
688: 
689: \clearpage
690: \begin{deluxetable}{ccc}
691: \tablecolumns{3}
692: \tablewidth{0pt}
693: \tablecaption{\label{tab:ions}  Ions Considered in NLTE}
694: \tablehead{}
695: \startdata
696: H I  & O I   & Si I\\
697: He I & O II  & Si II\\
698: He II& O III & Si III\\
699: C I  & Ne I  & Ca I\\
700: C II & Na I  & Ca II\\
701: C III& Na II & Ca III\\
702: N I  & Mg I  & Fe I\\
703: N II & Mg II & Fe II\\
704: N III& Mg III& Fe III\\
705: \enddata
706: \end{deluxetable}
707: 
708: \clearpage
709: \begin{deluxetable}{lccccc}
710: \tablecolumns{6}
711: \tablewidth{0pt}
712: \tablecaption{\label{tab:models} SN 1999dn Synthetic Spectra Model Parameters}
713: \tablehead{
714:   \colhead{Obs.~Spectrum Date} &
715:   \colhead{Epoch} &
716:   \colhead{Metallicity (\zsol)} &
717:   \colhead{\Tmod (K)} &
718:   \colhead{\vno (\kmps)} &
719:   \colhead{$n$}
720: }
721: \startdata
722: August 21\,\tablenotemark{a}    & -10 & 1 & 6750 & 11000 & 13\\
723:                                 &     & 3 & 7250 & 11000 & 13\\
724: August 31\,\tablenotemark{a}    & +0  & 1 & 6000 & 10000 & 13\\
725:                                 &     & 3 & 6500 & 10000 & 13\\
726: September 10\,\tablenotemark{b} & +10 & 1 & 5500 & 7000  & 13\\
727:                                 &     & 3 & 6000 & 7000  & 13\\
728: September 14\,\tablenotemark{a} & +14 & 1 & 5500 & 7000  & 10\\
729:                                 &     & 3 & 6000 & 7000  & 10\\
730: September 17\,\tablenotemark{b} & +17 & 1 & 5000 & 7000  & 10\\
731:                                 &     & 3 & 5500 & 7000  & 10\\
732: \enddata
733: 
734: \tablenotetext{a}{From \citet{deng00}.}
735: \tablenotetext{b}{From \citet{math_sn1bc}.}
736: 
737: \end{deluxetable}
738: 
739: \clearpage
740: \begin{figure}
741: \leavevmode
742: \begin{center}
743: %\includegraphics[width=0.75\textwidth,angle=90]{sn99dn_all}
744: \includegraphics[width=0.75\textwidth,angle=90]{f1}
745: \end{center}
746: \caption{\label{fig:all}The five observed spectra we model, referenced
747:   by their epochs, plotted in semi-log with arbitrary shifts added to
748:   make the spectra easier to see.  The -10, +0, and +14 day spectra
749:   are from \citet{deng00}.  The +10 and +17 day spectra are from
750:   \citet{math_sn1bc}.}
751: \end{figure}
752: 
753: \clearpage
754: \begin{figure}
755: \leavevmode
756: \begin{center}
757: %\includegraphics[width=0.75\textwidth,angle=90]{sn99dn_aug31_bothlabel}
758: \includegraphics[width=0.75\textwidth,angle=90]{f2}
759: \end{center}
760: \caption{\label{fig:aug31_both}The observed spectrum of SN~1999dn obtained
761:   on August 31, the day of maximum light (epoch +0) is compared to two PHOENIX
762:   models:  the first with the ``Standard Composition,'' the second
763:   with 3 times solar metallicity.  Both models fit the observed well,
764:   reproducing the He I absorptions fairly accurately.  The
765:   high-metallicity model fits the 6200 \AA~feature, believed to be
766:   \halpha,  better than the standard composition model.  In this and
767:   subsequent plots the observed spectra have been dereddened with $E(B
768:   - V) = 0.052$ \citep{cardelli89} and deredshifted using a recession
769:   velocity of 2700~\kmps.}
770: \end{figure}
771: 
772: \clearpage
773: \begin{figure}
774: \leavevmode
775: \begin{center}
776: %\includegraphics[width=0.75\textwidth,angle=90]{sn99dn_aug31_Fe2Si2only}
777: \includegraphics[width=0.75\textwidth,angle=90]{f3}
778: \end{center}
779: \caption{\label{fig:aug31_Fe2Si2}The 6200 \AA~feature in the
780:   observed spectrum is compared to the feature in the standard
781:   composition model spectrum.  Also shown are single-ion spectra from
782:   the standard composition model for Fe~II and Si~II.  We can see that
783:   the feature in the model spectrum is a blend of these two ions.}
784: \end{figure}
785: 
786: \clearpage
787: \begin{figure}
788: \leavevmode
789: \begin{center}
790: %\includegraphics[width=0.75\textwidth,angle=90]{sn99dn_aug31_alts}
791: \includegraphics[width=0.75\textwidth,angle=90]{f4}
792: \end{center}
793: \caption{\label{fig:aug31_alts}The 6200 \AA~ feature in the
794:   observed spectrum is compared to the feature in the standard
795:   composition model spectrum and two alternative models:  the first has
796:   10 times the Si abundance as the standard composition model; the
797:   second has 100 times the Ne abundance.  The increased Si abundance
798:   cannot account for the 6200 \AA~feature itself.  The increased Ne
799:   abundance reproduces the feature remarkably well, but such a
800:   dramatic increase in neon abundance does not seem reasonable.}
801: \end{figure}
802: 
803: \clearpage
804: \begin{figure}
805: \leavevmode
806: \begin{center}
807: %\includegraphics[width=0.75\textwidth,angle=90]{sn99dn_aug21_bothlabel}
808: \includegraphics[width=0.75\textwidth,angle=90]{f5}
809: \end{center}
810: \caption{\label{fig:aug21_both}The observed spectrum of SN~1999dn
811:   10 days before maximum light (obtained on August 21) is shown with
812:   two model spectra:  the standard composition model and a
813:   high-metallicity model.  Again, both models fit well, but the
814:   standard composition model fails to produce a strong feature near
815:   6200 \AA, while the high-metallicity model can.}
816: \end{figure}
817: 
818: \clearpage
819: \begin{figure}
820: \leavevmode
821: \begin{center}
822: %\includegraphics[width=0.75\textwidth,angle=90]{sn99dn_sep10_bothlabel}
823: \includegraphics[width=0.75\textwidth,angle=90]{f6}
824: \end{center}
825: \caption{\label{fig:sep10_both}The observed spectrum of SN~1999dn
826:   10 days after maximum light (obtained on September 10) is shown with
827:   two model spectra:  the standard composition model and a
828:   high-metallicity model.  While both models fit well, the standard
829:   composition model fits marginally better in the Fe II dominated
830:   region of the spectra between 4000 \AA~and 5000 \AA.  Also, the
831:   observed 6200 \AA~feature has become much weaker, with a better fit
832:   to it given by the standard composition model.}
833: \end{figure}
834: 
835: \clearpage
836: \begin{figure}
837: \leavevmode
838: \begin{center}
839: %\includegraphics[width=0.75\textwidth,angle=90]{sn99dn_sep14_nlabel}
840: \includegraphics[width=0.75\textwidth,angle=90]{f7}
841: \end{center}
842: \caption{\label{fig:sep14_n}The observed spectrum of SN~1999dn
843:   14 days after maximum light (obtained on September 14) is shown with
844:   the standard composition model while varying $n$ from 13 to 10.  At
845:   this epoch and in the +17 day models we prefer a less steep density
846:   profile, with the value $n = 10$.}
847: \end{figure}
848: 
849: \clearpage
850: \begin{figure}
851: \leavevmode
852: \begin{center}
853: %\includegraphics[width=0.75\textwidth,angle=90]{sn99dn_sep14_bothlabel_n10}
854: \includegraphics[width=0.75\textwidth,angle=90]{f8}
855: \end{center}
856: \caption{\label{fig:sep14_both}The observed spectrum of SN~1999dn
857:   14 days after maximum light (obtained on September 14) is shown with
858:   two model spectra:  the standard composition model and a
859:   high-metallicity model, with $n = 10$ for both models.  Again, we
860:   prefer the standard composition model at this post-maximum epoch.}
861: \end{figure}
862: 
863: \clearpage
864: \begin{figure}
865: \leavevmode
866: \begin{center}
867: %\includegraphics[width=0.75\textwidth,angle=90]{sn99dn_sep17_bothlabel}
868: \includegraphics[width=0.75\textwidth,angle=90]{f9}
869: \end{center}
870: \caption{\label{fig:sep17_both}The observed spectrum of SN~1999dn
871:   17 days after maximum light (obtained on September 17) is shown with
872:   two model spectra:  the standard composition model and a
873:   high-metallicity model.  Both models fit parts of the observed
874:   spectrum better than the other; overall, we prefer the standard
875:   composition model because of the decreased strength of the features
876:   in the blue.  The standard composition model fits the 6200
877:   \AA~feature slightly better as well.}
878: \end{figure}
879: 
880: \clearpage
881: \begin{figure}
882: \leavevmode
883: \begin{center}
884: %\includegraphics[width=0.75\textwidth,angle=90]{sn99dn_sep17_LTE}
885: \includegraphics[width=0.75\textwidth,angle=90]{f10}
886: \end{center}
887: \caption{\label{fig:sep17_LTE}The observed spectrum of SN~1999dn
888:   17 days after maximum light (obtained on September 17) is shown with
889:   single ion spectra of Ni II, Al I, and Ti II.   Note that Ti II has a
890:   very large contribution to the spectrum in the blue, in the Fe II region.}
891: \end{figure}
892: 
893: \clearpage
894: \begin{figure}
895: \leavevmode
896: \begin{center}
897: %\includegraphics[width=0.75\textwidth,angle=90]{sn99dn_sep17_10C}
898: \includegraphics[width=0.75\textwidth,angle=90]{f11}
899: \end{center}
900: \caption{\label{fig:sep17_10C}The observed spectrum of SN~1999dn
901:   17 days after maximum light (obtained on September 17) is shown with
902:   our standard composition model and a model with ten times as much
903:   carbon abundance.  Note that the enhanced carbon seems to have
904:   little effect in the 6200 \AA~region, but does reproduce an observed
905:   feature near 9500 \AA.  This may be evidence that there is increased
906:   C II at later times that may be changing the shape of some features
907:   in the spectrum, including the 6200 \AA~region.}
908: \end{figure}
909: 
910: \clearpage
911: 
912: \end{document}
913: