1: \documentclass[preprint]{aastex}
2: \usepackage{xspace}
3:
4: \newcommand{\eg}{{\it e.g.}\xspace}
5: \newcommand{\ie}{{\it i.e.}\xspace}
6: \newcommand{\cf}{{\it cf.}\xspace}
7: \newcommand{\etal}{{\it et al.}\xspace}
8: \newcommand{\nifsx}{${}^{56}$Ni\xspace}
9: \newcommand{\cofsx}{${}^{56}$Co\xspace}
10: \newcommand{\fefsx}{${}^{56}$Fe\xspace}
11:
12: \slugcomment{To appear in Ap. J.}
13:
14: \lefthead{Ivanov \etal}
15: \righthead{SN~Ia Age-Luminosity Relation}
16:
17: \begin{document}
18:
19: \title{On the Relation Between Peak Luminosity and Parent Population
20: of Type Ia Supernovae: A New Tool for Probing the Ages of Distant
21: Galaxies}
22:
23: \author{Valentin D. Ivanov, Mario Hamuy, Philip A. Pinto}
24: \affil{Steward Observatory, The University of Arizona, Tucson, AZ 85721,\\
25: vdivanov, mhamuy, \& ppinto@as.arizona.edu}
26:
27: \begin{abstract}
28: We study the properties of Type Ia Supernovae (SNe~Ia) as functions of the
29: radial distance from their host galaxy centers. Using a sample of 62 SNe~Ia
30: with reliable luminosity, reddening, and decline rate determinations, we
31: find no significant radial gradients of SNe~Ia peak absolute magnitudes or
32: decline rates in elliptical+S0 galaxies, suggesting that the diversity of
33: SN properties is not related to the metallicity of their progenitors. We
34: do find that the range in brightness and light curve width of supernovae in
35: spiral galaxies extends to brighter, broader values. These results are
36: interpreted as support for an age, but not metallicity, related origin of
37: the diversity in SNe~Ia. If confirmed with a larger and more accurate
38: sample of data, the age-luminosity relation would offer a new and powerful
39: tool to probe the ages and age gradients of stellar populations in galaxies
40: at redshift as high as $z\sim1-2$. The absence of significant radial
41: gradients in the peak $\rm (B-V)_0$ and $\rm (V-I)_0$ colors of SNe Ia
42: supports the redding correction method of Phillips \etal (1999). We find no
43: radial gradient in residuals from the SN~Ia luminosity-width relation,
44: suggesting that the relation is not affected by properties of the
45: progenitor populations and supporting the reliability of cosmological results
46: based upon the use of SNe~Ia as distance indicators.
47: \end{abstract}
48:
49: \keywords{cosmology: observations -- distance scale --
50: galaxies: general -- supernovae: general}
51:
52:
53:
54: \section{Introduction}
55: In recent years the search for a precise extragalactic distance indicator
56: has stimulated the study of Type Ia Supernovae (hereafter SN~Ia). A number
57: of surveys (\eg \citealt{ham93,ham96a,ham96b,ham96c,rie99,per99})
58: have yielded samples with high-quality light curves and well-defined
59: selection criteria. The availability of such data provides an opportunity
60: for studying SNe~Ia at an unprecedented level of detail.
61:
62: Perhaps the most important result of these surveys is that the SNe~Ia do
63: not constitute a homogeneous family of objects -- SNe~Ia are not all the
64: same. They span a range in peak B luminosity of approximately three
65: magnitudes and exhibit a variety of light curve morphologies
66: \citep{maz94,ham95}. Their heterogeneity has also been demonstrated
67: spectroscopically \citep{phi92}.
68:
69: While the discovery of a significant dispersion in peak brightness seemed
70: to spell disaster for the use of SNe~Ia as extragalactic standard candles,
71: \citet{phi93} demonstrated a tight correlation between the peak luminosity
72: and the rate of decline from peak in the first 15 days. This relation was
73: refined by \citet{ham95}, and shown to reduce the dispersion about the
74: Hubble relation by a factor of two by \citet{ham96b}. This dispersion was
75: then reduced even further by \citet{phi99}, who
76: included a precise correction for extinction to achieve a dispersion
77: between 0.09 and 0.12 magnitudes, depending on photometric band.
78:
79: \citet{rie96}, using a different statistical technique, have also shown
80: that the light curve shape near peak is an accurate predictor of luminosity.
81: \citet{Perlmutteretal97} have demonstrated that not only is there a
82: correlation of the light curve shape with luminosity but that, near peak,
83: the light curve is the {\em same} in all supernovae up to a correlated
84: scaling in time and luminosity. All three methods are capable of achieving
85: similarly accurate determinations of a supernova's peak brightness.
86:
87: The discovery and refinement of the luminosity-width relation in SNe~Ia has
88: enabled \citet{rie99} and \citet{per99} to use observations of SNe~Ia over
89: a large range of redshifts to show that the cosmological expansion is
90: accelerating at a rate which cannot be explained without the introduction
91: of new physics such as a cosmological constant.
92:
93: It is widely accepted that SNe~Ia result from the thermonuclear
94: incineration of a carbon/oxygen white dwarf, and that the light curve is
95: powered by the radioactive decay of \nifsx produced in the explosion. The
96: systematic diversity of SNe~Ia may provide important clues to the nature of
97: their progenitors and the explosion mechanism. It has been suggested that
98: the range in peak luminosities displayed by SNe~Ia reflects a range in the
99: nucleosynthetic \nifsx yield. \citet{hof96} demonstrated that a variety of
100: light curve shapes could be generated from various explosion models, and
101: that the width of the light curve was correlated in some models with the
102: \nifsx mass. \citet{PintoE00} showed that changing only the \nifsx yield in
103: a typical explosion model can reproduce both the slope and the absolute
104: normalization of the luminosity-width relation to well within the observed
105: dispersion. Neither of these models explains the origin of SN~Ia diversity,
106: however.
107:
108: Unfortunately, much of the most important physics in current models for the
109: evolution and explosion of SNe~Ia is far too complex to be modeled in
110: anything close to a predictive fashion. For example, because we lack a
111: detailed theory of turbulent combustion, the behavior of the nuclear
112: burning rate, critical to the properties of the resulting explosion, must
113: be put in ``by hand''. \citet{hof98}, and \citet{ume99a,ume99b} have
114: explored the effects of progenitor age and metallicity on SN properties
115: using one particular set of assumptions. They find that in this case the
116: supernova progenitor's properties affect the $\rm^{56}Ni$ production and
117: the light curve width, with younger and more metal-rich progenitors leading
118: to brighter SNe Ia.
119:
120: An additional route towards understanding the origin of SN~Ia diversity is
121: the study of the supernova environments, in particular the location of the
122: supernov\ae\ in the host galaxy. Since properties of the stellar population
123: differ from galaxy to galaxy and vary with location within a galaxy, such
124: studies can show which properties of the parent population affect the
125: nature of the SN~Ia explosion.
126:
127: \citet{ham96a} found that faster-declining (fainter)
128: SNe~Ia tend to occur in earlier-type galaxies, suggesting that the age of
129: the progenitor is a significant parameter. This result was confirmed by
130: \citet{rie99}, who also found that after correcting
131: peak absolute magnitudes for extinction and applying the width-luminosity
132: relation, there was no residual difference in the corrected supernova
133: magnitudes between early- and late-type galaxies.
134:
135: The location of SNe in their host galaxies has been extensively studied
136: (\cf \citealt{van97} and references therein). \citet{sha79} noted a
137: deficiency of SNe~Ia in the innermost regions of galaxies, but attributed
138: it to a selection effect; in photographic surveys, it is difficult to
139: discern SNe against the higher background on which they are projected near
140: the centers of the galaxies. \citet{wan97} and \citet{how00} later confirmed this effect, at least in the older photographic SNe searches. \citet{ham99}
141: also showed this to be the case in a more rigorous analysis of the
142: Cal\'an-Tololo survey.
143:
144: Major problems which have plagued these and similar studies of the radial
145: gradients of SNe~Ia in their host galaxies include the difficulty of
146: determining true galactocentric distances and the intrinsic spread of
147: properties among the host galaxies themselves. The goal of this paper is
148: to use the best-observed sample of SNe~Ia available to study the radial
149: distribution of SNe~Ia in their host galaxies and the variations of their
150: properties with galactocentric distance. We improve on these previous
151: studies by: (i) deprojecting the observed angular distances between the SN
152: and the host galaxy center, removing ambiguities arising from projecting
153: the distances on the plane of the sky; (ii) normalizing such
154: distances to the sizes of the hosts; (iii) applying statistical methods to
155: estimate the significance of radial gradients in SN properties; and (iv)
156: exploring the possible applications of SNe~Ia as tools to probe the stellar
157: populations of distant galaxies. We will compare our results with current
158: theoretical understanding of the origin of the peak luminosity - decline
159: rate relation, and will assess consistency with the known radial variation
160: of age and metallicity in galaxies.
161:
162: \section{The SNe~Ia Sample: Selection and Properties}
163:
164: Our sample consists of 62 well-observed SNe~Ia (listed in
165: Table~\ref{tbl-1}) with reliable estimates of peak luminosity, reddening,
166: and decline rate $\rm\Delta m_{15}(B)$ as determined by \citet{phi99};
167: throughout this paper we use only the reddening-corrected absolute
168: magnitudes as presented in that work. Several SNe with no peak BVI
169: measurements were used exclusively in tests that require only positional
170: information. SN1992K and SN1991bg were excluded from our analysis as
171: peculiarly subluminous; there is considerable debate in the supernova
172: community as to whether these objects are a different sort of explosion or
173: merely the faintest tail of the ``normal'' supernovae distribution.
174:
175: The Hubble type and redshifts of the host galaxies were taken from the
176: NASA/IPAC Extragalactic Database (hereafter NED). The galaxy semi-major
177: axes and position angles are taken from the RC3 \citep{dev91}, the UGC
178: \citep{nil73}, and the ESO/Upsala catalogs \citep{hol74}. For galaxies with
179: no available measurements in the literature, we determined galaxy
180: parameters from the Cal\'an/Tololo B-band images. SN peak magnitudes and
181: decline rates were determined by template-fitting following the
182: prescription of \citet{ham96a}. We use Cepheid distances measured by
183: \citet{gib00} for six nearby SNe, and distance derived from observed
184: redshifts for the remainder of the sample. Throughout this paper we adopt
185: a value for the Hubble constant $\rm H_0~=~69~km~sec^{-1}~Mpc^{-1}$ for
186: consistency with the Gibson \etal distances. We also assume a cosmology
187: with $\rm \Omega_0=0.2$, $\rm \Omega_\Lambda=0.0$ to calculate the linear
188: galactocentric distances.
189:
190: \placetable{tbl-1}
191:
192: To de-project the angular distances of SNe Ia from the centers of disk
193: galaxies, we assume that all SNe~Ia populate disks which are thin compared
194: with the galaxy diameters. This is a reasonable assumption for spirals and,
195: perhaps, for S0 galaxies. To compute actual galactocentric distances, we
196: express the SN position in a radial coordinate system centered at the host
197: galaxy nucleus and aligned with the semi-major axis. We then use the axial
198: ratio to calculate the inclination of the disk. In most cases the projected
199: angular offsets (listed in Table~\ref{tbl-1}) come from the discovery
200: announcements of the supernovae, with positions verified against the
201: NED. We did not de-project the angular distances for the SNe in ellipticals
202: as the depths of these galaxies along the line of sight are comparable to
203: their apparent sizes.
204:
205: \placefigure{fig1}
206:
207: The distributions of the major properties of our sample are shown in
208: Figure~\ref{fig1}. 23 SNe are in elliptical and S0 hosts and 39 in spirals.
209: Observational selection \citep{ham99} accounts for
210: the higher redshift of the SNe in elliptical+S0 than in spirals; early-type
211: galaxies tend to dominate the statistics for larger projected
212: galactocentric distances. Reddening in spirals is seen to be higher than in
213: early-type galaxies, with the notable exception of SN1986G ($\rm
214: E(B-V)=0.50\pm0.05$ mag) whose host is the peculiar S0 Seyfert~2 galaxy
215: NGC~5128. Naturally, we find in our sample a deficiency of SNe fainter
216: than $\rm M_B = -19$ mag, with $\rm E(B-V)>0.10$ mag.
217:
218: \placetable{tbl-2}
219:
220: The brightness at maximum light of SNe~Ia in our sample is related to the
221: morphological type of the host galaxy. \citet{ham96a} pointed out that
222: faster-declining (fainter) SNe tend to occur in earlier type galaxies, a
223: result later confirmed by \citet{rie99}. We carried out a
224: Kolmogorov-Smirnov (hereafter KS) test between the peak absolute magnitude
225: distributions for the SNe in early-type and late-type galaxies. The
226: significance level P for the null hypothesis that the two sets of data are
227: drawn from the same parent distribution is given in
228: Table~\ref{tbl-2}. Small values of P suggest that the distributions are
229: different. We obtained only 2, 2 and 7\% of probability, that the
230: distributions of SNe Ia peak magnitudes, respectively for B, V, and I-band
231: observations, in elliptical+S0 galaxies and spirals were drawn from the
232: same parent distribution. The higher probability for the I-band reflects
233: the smaller amplitude of the luminosity-width relation in the near infrared.
234: The rate of decline is an even better indicator of a difference between SNe~Ia
235: in early- and in late-type galaxies, giving a significance level of only
236: 0.001\%. The de-reddened $\rm (B-V)_0$ SNe colors, on the other hand, are
237: identical to a confidence level of 90\%. The very small uncertainties in
238: this analysis reflect major differences of the intrinsic properties of the
239: host galaxies such as age, metallicity, and dust content, and strongly
240: suggest that the properties of SNe~Ia are determined by their environments.
241:
242: \section{Radial Gradients of SNe Ia Properties}
243:
244: The SN host galaxies show a spread of sizes, masses and metallicities --
245: presumably because larger and more massive galaxies can better retain the
246: heavy elements produced by stellar evolution
247: \citep{ari87}. Galaxy-to-galaxy abundance variations can be corrected, to
248: some extent, by a proper normalization of galactocentric
249: distances. Following \citet{van97}, we divided the
250: galactocentric distances by the semi-major axis. This approach has the
251: advantage that the semi-major axis can be used as a "metalometer". Since
252: semi-major axes are not always accurately known, we will carry out our
253: analyses using both normalized and absolute radial distances, for
254: comparison.
255:
256: \placefigure{fig2}
257:
258: We first carried out linear fits for the absolute peak magnitude against
259: the linear distance $\rm R_{dp}$ (in Kpc) and for the distance normalized
260: to the semi-major axis $\rm R_{dp}/a$. We then calculated the residuals
261: $\Delta$ from the luminosity-width relation for each SN, using the
262: quadratic form of the relation given by \citet{phi99}, in order to look for
263: trends of $\Delta$ with galactocentric distance. We then computed fits for
264: the decline rate $\rm \Delta m_{15}$ against both
265: distances. Table~\ref{tbl-3} lists the coefficients, their $1\sigma$ errors
266: in brackets, and the standard deviation of the fits shown in
267: Figure~\ref{fig2}. We carried out separate analyses for the entire sample,
268: for SNe in elliptical+S0 galaxies, and for SNe in spirals. We concentrate
269: on fits to B-band magnitudes and widths because the amplitude of the
270: luminosity-width rate relation is the largest at these wavelengths,
271: rendering our results both more sensitive to differences among supernovae
272: and to the de-reddening procedure employed by Phillips \etal.
273:
274: \placetable{tbl-3}
275:
276: Taking into account the entire sample, we find that SNe closer to the
277: galaxy centers are brighter, in agreement with the conclusion of
278: \citet{wan97}. However, when we divide the SNe Ia according to the host
279: galaxy type, we find no significant radial gradient of the absolute peak
280: magnitudes. The gradient in the sample as a whole can easily be understood;
281: a brighter population of SNe Ia in spirals dominates the sample at small
282: galactocentric distances $\rm (R_{dp} \leq 7.5)$ while the fainter
283: population of SNe Ia in ellipticals and S0's dominates at larger radial
284: distances $\rm (R_{dp} > 7.5)$.
285:
286: There seems to be a small but significant $(2-5\sigma)$ correlation between
287: decline rate and radial distance, but this result is due mainly to one
288: object (SN1992bo) illustrating the importance of extending the SN sample.
289: When the early- and late-type galaxies are treated separately, the decline
290: rate gradient is reduced by a factor of 2 or more. Employing normalized
291: radial distances makes the trends of $\rm \Delta m_{15}$ with distance
292: disappear at the $2-2.5\sigma$ level.
293:
294: \placefigure{fig3}
295:
296: The de-reddened SNe~Ia $\rm (B-V)_0$ color at maximum light shows no radial
297: variation (Figure~\ref{fig3}). If we use the linear distance fits, the
298: color change is only 0.03 mag over 30 Kpc for the entire sample. The change
299: is just sightly larger for spirals alone (0.06 mag over 10 Kpc). Similarly,
300: there is no radial gradient of $\rm (V-I)_0$ color at peak within the
301: uncertainties. This results lend credence to the reddening correction
302: method of Phillips \etal.
303:
304: \section{Discussion}
305:
306: The radial metallicity gradient in elliptical galaxies is an established
307: observational fact (\citealt{hen99} and references therein;
308: \citealt{kob99,tam00}). At the same time, the population in these galaxies
309: is co-eval \citep{wor94,vaz96,vaz97}. The stars near the center of spiral
310: galaxies are, however, in general both more metal rich {\em and} older when
311: compared with outer regions \citep{hen99,bel00}. This difference provides
312: an ideal opportunity to separate the effects of age and metallicity on
313: SNe~Ia properties.
314:
315: The lack of any radial gradients in SNe~Ia properties in elliptical+S0
316: galaxies suggests that neither the absolute peak magnitude nor the decline
317: rate is a strong function of the metal abundance of the progenitor. The
318: residuals from the luminosity-width relation seem also to be independent of
319: metallicity in the host population. At the same time, SNe~Ia in late-type
320: galaxies exhibit a significantly different distribution in both peak
321: magnitude and decline rate from their cousins in early-type galaxies.
322:
323: These results are consistent with the hypothesis that the {\em age} of the
324: progenitor, but not the metallicity, may be responsible for the diversity
325: of SNe Ia peak luminosities and decline rates. Current theoretical
326: understanding of the problem is fraught with uncertainties, but a possible
327: scenario was suggested by \citet{ume99a,ume99b}. They argue that, no matter
328: what the details of explosive burning may be, the larger specific energy
329: available from burning larger mass fractions of carbon in a C/O dwarf will
330: result in greater \nifsx production and, hence, brighter SNe~Ia. They then
331: argue that the oldest SN progenitor systems are those with the
332: smallest-mass companions, and that these systems can transfer less mass in
333: all to the pre-SN white dwarf. Thus, the pre-SN systems which take longest
334: to evolve to explosion must be those with the largest initial C/O white
335: dwarf mass. Since larger C/O core masses have smaller carbon abundances,
336: these older systems will lead to lower \nifsx masses, and eventually,
337: fainter supernovae.
338:
339: Such a scenario naturally explains the absence of bright SNe~Ia in
340: ellipticals, with their older populations, and is also consistent with
341: the larger dispersion of peak luminosity found near the center of spiral
342: galaxies \citep{wan97} where SNe Ia would be produced from a mixture of
343: stellar populations with various ages.
344:
345: Our results are consistent with the predictions of \citet{von97} who
346: carried out a semi-empirical modeling of the white dwarf mass function,
347: and derived the SN Ia luminosity function for a range of progenitor
348: masses and ages. Their work used the older sub-Chandrasekhar explosion
349: models of \citet{woo94}, different from the Chandrasekhar mass
350: explosions considered by \citet{ume99a,ume99b}.
351:
352: The range in accretion rate necessary to produce SNe~Ia from Chandrasekhar
353: white dwarfs in single-degenerate systems is quite narrow \citep{NK91}, too
354: narrow to lead to a SNe~Ia rate in accord with observations.
355: \citet{HKN96,HKN99} have suggested that in systems with greater mass
356: transfer rates, a wind may remove all but the necessary amount from the
357: system, leading to a much broader range in binary systems capable of
358: leading to SNe~Ia. The wind is driven by radiation pressure on metal
359: lines, and its strength thus depends upon both the metallicity of the
360: system and the luminosity from burning accreted material. This luminosity
361: is larger for larger C/O dwarf masses, and thus for a given C/O dwarf mass
362: the metallicity of the system must be greater than some minimum value.
363: Below a critical metallicity, the radiation momentum absorbed by the wind
364: is insufficient to reduce the mass transfer rate below that leading to a
365: common-envelope system. The papers by Umeda \etal suggest that this will
366: lead to a metallicity effect on the range of SN~Ia luminosities -- the
367: larger the C/O dwarf mass, the lower the critical metallicity required for
368: the wind. Thus, lower-metallicity systems will tend to have larger initial
369: C/O dwarf masses and hence fainter supernov\ae.
370:
371: This metallicity effect seems not to be supported by our results. Since
372: we do not find strong radial variation of the SNe Ia parameters separately
373: in elliptical+S0 galaxies, and in spirals, such an
374: effect seems unlikely to explain the lack of bright supernovae in the
375: ellipticals. Indeed, as we find no variation in supernova properties with
376: radius in early-type galaxies, any wind in SN~Ia progenitor systems must be
377: much less sensitive to metallicity than posited by Hachisu, Kato, \& Nomoto.
378:
379: Given a range in \nifsx yield, \citet{PintoE00} show that this is in itself
380: sufficient to lead to the observed luminosity-width relation. They suggest
381: that this relation is thus unlikely to be affected by metallicity effects
382: on the radiation transport responsible for the relation, and the lack of
383: any radial gradient in SN properties in ellipticals and S0's bears this
384: out.
385:
386: Most of the SNe Ia with Cepheid calibrations \citep{gib00} happen to lie at
387: large radial distances in their host galaxies and are therefore fainter
388: than average. One might be concerned that a selection bias against SNe~Ia
389: at small angular separations from their host galaxy centers (the ``Shaw
390: effect'') would make local and distant samples of SNe Ia incommensurate.
391: Our results imply that after correcting for extinction and luminosity-width
392: relation, SNe Ia become true standard candles, independently of their
393: galactocentric distance, and thus, metal abundance.
394:
395: Indeed, the lack of any correlation found to the {\em corrected} supernovae
396: magnitudes should be heartening to those employing SNe~Ia at large
397: redshifts to measure cosmological properties. The range in both metallicity
398: and age spanned by the galaxies in our sample exceeds the difference in
399: average values of these parameters between the local universe and at a
400: redshift $z\sim1$. To explain away the cosmological results found by
401: \citet{Riessetal98} and by \citet{per99}, the corrected peak magnitudes
402: of the lower metallicity and/or younger supernovae found at $z\sim1$
403: would have to be fainter on average by more than 0.2 mag, a result
404: clearly ruled out by Figure~(\ref{fig2}).
405:
406: At lesser distances, the lack of any gradient in residuals from the
407: luminosity-width relation argues against a bias in the determination of
408: the Hubble constant based on intrinsically fainter nearby SNe Ia.
409:
410: A possible test of the relation between the peak luminosities (and light
411: curve shapes) of SNe~Ia and the age of their progenitor populations can be
412: carried via spectral indices age diagnostics of the host galaxy population
413: \citep{wor94,vaz96,vaz97}. If confirmed, this relation might be used to
414: probe the ages and age gradients of stellar populations in galaxies. The
415: intrinsic brightness of these SNe makes it possible to carry out such an
416: analysis to redshifts beyond $z\sim1$ with current observational
417: capabilities. Such a technique is rendered even more important by the
418: advent of new telescopes and instruments in the near future. For instance,
419: \citet{dah99} estimate that NGST is expected
420: to detect about 45 SNe Ia per field in a single $10^4$ sec exposure, with
421: mean redshift of $\rm z\approx2$. From the ground, the 8.4m wide-field Dark
422: Matter Telescope proposed by \citet{hin99} would provide high-quality
423: observations of many hundreds of supernovae per year to $z\sim1.5$ (Pinto
424: 1999, private communication).
425:
426: \section{Summary}
427:
428: Our KS tests confirm the result of \citet{ham96a}
429: that SNe~Ia in elliptical+S0 and in spiral host galaxies are intrinsically
430: different, with faster-declining (fainter) SNe~Ia occurring mostly in
431: earlier-type galaxies.
432:
433: We find no significant radial variation of the SNe~Ia peak absolute
434: brightness in ellipticals+S0s and in spirals. Taken together, SNe do show a
435: correlation in peak brightness with radial distance. This is a consequence
436: of the fact that fainter SNe~Ia are found in ellipticals, and these SNe
437: dominate the sample at larger galactocentric distances.
438:
439: Our analysis shows no significant radial gradient of the
440: reddening-corrected peak magnitude $\rm (B-V)_0$ and $\rm (V-I)_0$ colors,
441: lending support to the reddening correction method of \citet{phi99}.
442:
443: The peak absolute luminosity and the decline rate of SNe Ia are
444: independent of the metallicity of the SNe progenitors, within the
445: observational uncertainties. Determinations of extragalactic
446: distances and related cosmological parameters are thus not biased by
447: metallicity- or age-related evolutionary effects.
448:
449: Our results are consistent with an age-related origin of the
450: luminosity-width relation. If confirmed, this age {\em vs.} decline rate
451: relation in SNe Ia offers an important new tool to probe the ages and age
452: gradients of stellar populations in galaxies at redshifts as high as
453: $z\sim1-2$.
454:
455:
456: \acknowledgments
457:
458: We are grateful to Dr. M.M. Phillips for the useful discussions. The
459: research has made use of the NASA/IPAC Extragalactic Database which is
460: operated by the Jet Propulsion Laboratory, California Institute of
461: Technology, under contract with NASA. This work has been supported by the
462: National Science Foundation (CAREER grant AST9501634). PAP
463: gratefully acknowledges support from the Research Corporation though a
464: Cottrell Scholarship.
465:
466: \clearpage
467:
468: \input{tab1}
469: \clearpage
470: \input{tab2}
471: \clearpage
472: \input{tab3}
473: \clearpage
474:
475: \begin{thebibliography}{}
476: \bibitem[\protect\citeauthoryear{{Arimoto} \& {Yoshii}}{{Arimoto} \& {Yoshii}}{1987}]{ari87} Arimoto, N., \& Yoshii, Y.,
477: 1987, \aap, 173, 23
478: \bibitem[\protect\citeauthoryear{{Bell} \& {de Jong}}{{Bell} \& {de Jong}}{2000}]{bel00} Bell, E.F., \& de Jong, R.S.,
479: 2000, \mnras, 312, 497
480: \bibitem[\protect\citeauthoryear{{Dahl\'en} \& {Fransson}}{{Dahl\'en} \& {Fransson}}{1999}]{dah99}
481: Dahl\'en, T., \& Fransson, C., 1999, \aap, 350, 349
482: \bibitem[\protect\citeauthoryear{{de Vaucouleurs} et~al.}{{de Vaucouleurs} et~al.}{1999}]{dev91} de Vaucouleurs, G., de
483: Vaucouleurs, A., Corwin, J.R., Buta, R.J., Paturel, G., \&
484: Fouque, P., 1991, Third Reference Catalog of Bright Galaxies,
485: (Springer-Verlag Berlin Heidelberg New York).
486: \bibitem[\protect\citeauthoryear{{Gibson} et~al.}{{Gibson} et~al.}{2000}]{gib00}
487: Gibson, B.K., Stetson, P.B.,
488: Freedman, W.L., Mould, J.R., Kennicutt Jr., R.C., Huchra, J.P.,
489: Sakai, S., Graham, J.A., Fassett, C.I., Kelson, D.D., Ferrareze,
490: L., Hughes, S.M.G., Illingworth, G.D., Macri, L.M., Madore, B.F.,
491: Sebo, K.M., \& Silbermann, N.A., 2000, ApJ, 529, 723
492: \bibitem[\protect\citeauthoryear{{Hachisu}, {Kato}, \& {Nomoto}}{{Hachisu}, {Kato}, \& {Nomoto}}{1996}]{HKN96}
493: {Hachisu}, I., {Kato}, M., \& {Nomoto}, K. 1996, \apj, 470, L97.
494: \bibitem[\protect\citeauthoryear{{Hachisu}, {Kato}, \& {Nomoto}}{{Hachisu}, {Kato}, \& {Nomoto}}{1999}]{HKN99}
495: {Hachisu}, I., {Kato}, M., \& {Nomoto}, K. 1999, submitted to \apj.
496: \bibitem[\protect\citeauthoryear{{Hamuy} et~al.}{{Hamuy} et~al.}{1993}]{ham93} Hamuy M., Maza, J, Phillips, M.M.,
497: Suntzeff, N.B., Wischnjewsky, M., Smith, R.C., Antezana, R.,
498: Wells, L., et al., 1993, \aj, 106, 2392
499: \bibitem[\protect\citeauthoryear{{Hamuy} et~al.}{{Hamuy} et~al.}{1995}]{ham95} Hamuy M., Phillips, M.M., Maza, J.,
500: Suntzeff, N.B., Schommer, R.A., \& $\rm Avil\acute{e}s$, R., 1995,
501: \aj, 109, 1
502: \bibitem[\protect\citeauthoryear{{Hamuy} et~al.}{{Hamuy} et~al.}{1996a}]{ham96a} Hamuy M., Phillips, M.M., Schommer,
503: R.A., Suntzeff, N.B., Maza, J., \& $\rm Avil\acute{e}s$, R., 1996b,
504: \aj, 112, 2391
505: \bibitem[\protect\citeauthoryear{{Hamuy} et~al.}{{Hamuy} et~al.}{1996b}]{ham96b} Hamuy M., Phillips, M.M., Maza, J.,
506: Suntzeff, N.B., Schommer, R.A., \& $\rm Avil\acute{e}s$, R., 1996a,
507: \aj, 112, 2398
508: \bibitem[\protect\citeauthoryear{{Hamuy} et~al.}{{Hamuy} et~al.}{1996c}]{ham96c} Hamuy M., Phillips, M.M., Suntzeff,
509: N.B., Schommer, R.A., Maza, J., Antezana, R., Wischnjewsky, M.,
510: et al., 1996c, \aj, 112, 2408
511: \bibitem[\protect\citeauthoryear{{Hamuy} \& {Pinto}}{{Hamuy} \& {Pinto}}{1999}]{ham99} Hamuy M., \& Pinto, P.A., 1999,
512: \aj, 117, 1185
513: \bibitem[\protect\citeauthoryear{{Henry} \& {Worthey}}{{Henry} \& {Worthey}}{1999}]{hen99} Henry, R.B.C. \& Worthey, G.,
514: 1999, \pasp, 111, 919
515: \bibitem[\protect\citeauthoryear{{Hinz},{Wolf}, \& {Angel}}{1999}]{hin99}
516: Hinz, P., Wolf, N., \& Angel, R., 1999, in "Working on the Fringe:
517: An International Conference on Optical and IR Interferometry from
518: Ground and Space", Dana Point, CA, May 24-27, 1999, ASP Conf. Ser.
519: (Eds. S. Unwin and R. Stachnik), p. 55
520: \bibitem[\protect\citeauthoryear{{Holmberg} et~al.}{{Holmberg} et~al.}{1974}]{hol74} Holmberg, E.B., Lauberts, A.,
521: Schuster, H.E., \& West, R.M., 1974, \aaps, 18, 463
522: \bibitem[\protect\citeauthoryear{{Howell}, {Wang}, \& {Wheeler}}{{Howell}, {Wang}, \& {Wheeler}}{2000}]{how00} Howell, D.A., Wang, L., \&
523: Wheeler, J.C., 2000, ApJ, 530, 166
524: \bibitem[\protect\citeauthoryear{{H\"oflich} \& {Khokhlov}}{{H\"oflich} \& {Khokhlov}}{1996}]{hof96} H\"oflich,
525: P., \& Khokhlov, A., 1996, \apj, 472, L81
526: \bibitem[\protect\citeauthoryear{{H\"oflich}, {Wheeler}, \& {Thielemann}}{{H\"oflich}, {Wheeler}, \& {Thielemann}}{1998}]{hof98}
527: H\"oflich, P., Wheeler, J.C., \& Thielemann, F.K.,
528: 1998, \apj, 495, 617
529: \bibitem[\protect\citeauthoryear{{Kobayashi} \& {Arimoto}}{{Kobayashi} \& {Arimoto}}{1999}]{kob99} Kobayashi, C., \& Arimoto, N.,
530: 1999, \apj, 527, 573
531: \bibitem[\protect\citeauthoryear{{Maza} et~al.}{{Maza} et~al.}{1994}]{maz94} Maza, J., Hamuy, M., Phillips, M.M.,
532: Suntzeff, N.B., \& $\rm Avil\acute{e}s$, R., 1994, \apj, 424, L107
533: \bibitem[\protect\citeauthoryear{{Nilson}}{{Nilson}}{1973}]{nil73} Nilson, P., 1973, Nova Acta Regiae Soc. Sci.
534: Upsaliensis, Ser. V:A, 1, 1
535: \bibitem[\protect\citeauthoryear{{Nomoto} \& {Kondo}}{{Nomoto} \& {Kondo}}{1991}]{NK91}
536: {Nomoto}, K. \& {Kondo}, Y. 1991, \apj, 367, 119.
537: \bibitem[\protect\citeauthoryear{{Perlmutter} et~al.}{{Perlmutter} et~al.}{1997}]{Perlmutteretal97}
538: {Perlmutter}, S. et~al. 1997, \apj, 483, 565
539: \bibitem[\protect\citeauthoryear{{Perlmutter} et~al.}{{Perlmutter} et~al.}{1999}]{per99} Perlmutter, S., Aldering, G.,
540: Goldhaber, G., Knop, R.A., Nugent, P., Castro, P.G., Deustus, S.,
541: Fabbro, S., et al., 1999, \apj, 517, 565
542: \bibitem[\protect\citeauthoryear{{Phillips} et~al.}{{Phillips} et~al.}{1992}]{phi92} Phillips, M.M., Wells, L.A.,
543: Suntzeff, N.B., Hamuy, M., Leibundgut, B., Kirshner, R.P., \&
544: Foltz, C.B., 1992, \aj, 103, 1632
545: \bibitem[\protect\citeauthoryear{{Phillips}}{{Phillips}}{1993}]{phi93} Phillips, M.M., 1993, \apj, 413, L105
546: \bibitem[\protect\citeauthoryear{{Phillips} et~al.}{{Phillips} et~al.}{1999}]{phi99} Phillips, M.M., Lira, P., Suntzeff,
547: N.B., Schommer, R.A., Hamuy, M., \& Maza, J., 1999, \aj, 118, 1766
548: \bibitem[\protect\citeauthoryear{{Pinto} \& {Eastman}}{{Pinto} \& {Eastman}}{2000}]{PintoE00}
549: {Pinto}, P.A., \& {Eastman}, R.G. 2000, submitted to \apj
550: \bibitem[\protect\citeauthoryear{{Riess}, {Press}, \& {Kirshner}}{{Riess}, {Press}, \& {Kirshner}}{1996}]{rie96} Riess, A.G., Press, W.H.,
551: \& Kirshner, R.P., 1996, \apj, 473, 88
552: \bibitem[\protect\citeauthoryear{{Riess} et~al.}{{Riess}
553: et~al.}{1998}]{Riessetal98}
554: {Riess}, A., et~al. 1998, \aj, 116, 1009
555: \bibitem[\protect\citeauthoryear{{Riess} et~al.}{{Riess} et~al.}{1999}]{rie99} Riess, A.G., Kirshner, R.P., Schmidt,
556: B.P., et al., 1999, \aj, 117, 707
557: \bibitem[\protect\citeauthoryear{{Shaw}}{{Shaw}}{1979}]{sha79} Shaw, R.L., 1979, \aap, 76, 188
558: \bibitem[\protect\citeauthoryear{{Tamura} et~al.}{{Tamura} et~al.}{2000}]{tam00} Tamura, N., Kobayashi, C., Arimoto,
559: N., Kodama, T., \& Ohta, K., 2000, preprint (astro-ph/0001174)
560: \bibitem[\protect\citeauthoryear{{Umeda} et~al.}{{Umeda} et~al.}{1999a}]{ume99a} Umeda, H., Nomoto, K., Yamaoka, H.,
561: \& Wanajo, S., 1999a, \aj, 513, 861
562: \bibitem[\protect\citeauthoryear{{Umeda} et~al.}{{Umeda} et~al.}{1999b}]{ume99b} Umeda, H., Nomoto, K., Kobayashi, C.
563: Hachisu, I., \& Kato, M., 1999b, \apj, 522, L43
564: \bibitem[\protect\citeauthoryear{{van den Berg}}{{van den Berg}}{1997}]{van97} van den Berg, S., 1997, \aj, 113, 1
565: \bibitem[\protect\citeauthoryear{{Vazdekis} et~al.}{{Vazdekis} et~al.}{1996}]{vaz96} Vazdekis, A., Casuso E., Peletier,
566: R.F., \& Beckman, J.E., 1996, \apjs, 106, 307
567: \bibitem[\protect\citeauthoryear{{Vazdekis} et~al.}{{Vazdekis} et~al.}{1997}]{vaz97} Vazdekis, A., Peletier, R.F.,
568: Beckman, J.E., \& Casuso E., 1997, \apjs, 111, 203
569: \bibitem[\protect\citeauthoryear{{von Hippel} et~al.}{{von Hippel} et~al.}{1997}]{von97} von Hippel, T., Bothun, G.D., \& Schommer, R.A.,
570: 1997, \aj, 114, 115
571: \bibitem[\protect\citeauthoryear{{Wang}, {H\"oflich}, \& {Wheeler}}{{Wang}, {H\"oflich}, \& {Wheeler}}{1997}]{wan97} Wang, L.,
572: H\"oflich, P., \& Wheeler, J.C., 1997, \apj, 483, L29
573: \bibitem[\protect\citeauthoryear{{Woosley} \& {Weaver}}{{Woosley} \& {Weaver}}{1994}]{woo94} Woosley, S.E., \& Weaver, T.A., 1994, \apj, 423, 371
574: \bibitem[\protect\citeauthoryear{{Worthey}}{{Worthey}}{1994}]{wor94} Worthey, G., 1994, \apjs, 95, 107
575: \end{thebibliography}
576:
577: \clearpage
578:
579: \begin{figure}[!t]
580: \epsscale{0.8}
581: \plotone{f1.eps}
582: \caption{Properties of the SNe~Ia sample. The distributions of SNe Ia in
583: elliptical and S0 galaxies are given with solid lines, and the
584: distributions of the SNe Ia in spirals are given with dashed lines.}
585: \label{fig1}
586: \end{figure}
587:
588: \begin{figure}[!t]
589: \epsscale{0.8}
590: \plotone{f2.eps}
591: \caption{SN parameters as function of the galactocentric distances (left
592: panels), and galactocentric distance normalized to the semi-major axis
593: (right panels). The absolute B-band magnitude $\rm (M_B)$ is shown at the
594: top. The residuals from the peak luminosity vs. decline rate relation
595: $(\Delta_B)$ are shown at the center. The decline rate $\rm (\Delta m_{15})$
596: is shown at the bottom. The open circles represent SNe Ia in spirals and the
597: solid dots represent SNe Ia in elliptical and S0 galaxies. The fit to the
598: entire sample and its $1\sigma$ errors are shown with solid lines, the fit
599: to the SNe Ia in elliptical and S0 galaxies is shown with a dashed line and
600: the fit to the SNe Ia in spirals is shown with a dotted line.}
601: \label{fig2}
602: \end{figure}
603:
604: \begin{figure}[!t]
605: \epsscale{0.8}
606: \plotone{f3.eps}
607: \caption{
608: The reddening-corrected SN colors as function of
609: the galactocentric distances (left panels), and galactocentric distance
610: normalized to the semi-major axis (right panels). $\rm (B-V)_0$ color is
611: shown at the top panel, and $\rm (V-I)_0$ color is shown at the bottom
612: panel. The fits are indicated as in Figure 2.
613: }
614: \label{fig3}
615: \end{figure}
616:
617: \end{document}
618:
619: