1: %\documentclass[12pt,preprint]{aastex}

2: %\documentclass[manuscript]{aastex}

3: %\documentclass[preprint2]{aastex}

4: \documentclass{emulateapj}

5: \usepackage{natbib}

6: \usepackage{lscape}

7: 

8: %%\input epsf.sty

9: 

10: \newcommand{\etal}{et~al.}

11: \newcommand{\eg}{e.g., }

12: \newcommand{\ie}{i.e., }

13: \newcommand{\Msun}{M_{\odot}}

14: \newcommand{\kms}{km~s$^{-1}$}

15: \newcommand{\ergs}{erg~s$^{-1}$}

16: \newcommand{\Fefs}{$^{56}$Fe}

17: \newcommand{\Cofs}{$^{56}$Co}

18: \newcommand{\Nifs}{$^{56}$Ni}

19: \newcommand{\Mms}{M_{\rm MS}}

20: \newcommand{\Mej}{M_{\rm ej}}

21: \newcommand{\Mni}{M(^{56}{\rm Ni})}

22: \newcommand{\KE}{E_{\rm K}}

23: \newcommand{\vph}{v_{\rm ph}}

24: \def\ion#1#2{{\rm #1}~{\sc #2}}

25: \def\gsim{\mathrel{\rlap{\lower 4pt \hbox{\hskip 1pt $\sim$}}\raise 1pt

26: \hbox {$>$}}}

27: \def\lsim{\mathrel{\rlap{\lower 4pt \hbox{\hskip 1pt $\sim$}}\raise 1pt

28: \hbox {$<$}}}

29: 

30: 

31: \shorttitle{The evolution of the peculiar Type Ia  SN 2005hk}

32: \shortauthors{Sahu et al.} 

33: 

34: \begin{document}

35: 

36: %%%%%%%%%%%%%%%%%%%%%%%% Title Page begins %%%%%%%%%%%%%%%%%%%%%%%%%%%%%

37: 

38: \title{The evolution of the peculiar Type Ia supernova SN 2005hk over 400

39: days}

40: 

41: \author{

42: D.K. Sahu\altaffilmark{1},

43: Masaomi Tanaka\altaffilmark{2},

44: G.C. Anupama\altaffilmark{1},

45: Koji S. Kawabata\altaffilmark{3},

46: Keiichi Maeda\altaffilmark{4,5},

47: Nozomu  Tominaga\altaffilmark{2},

48: Ken'ichi Nomoto\altaffilmark{2,5,6},

49: Paolo A. Mazzali\altaffilmark{4,6,7,8}, and

50: T.P.Prabhu\altaffilmark{1} }

51: \email{dks@iiap.res.in, mtanaka@astron.s.u-tokyo.ac.jp}

52: 

53: \altaffiltext{1}{Indian Institute of Astrophysics, II Block Koramangala,

54: Bangalore 560034, India}

55: \altaffiltext{2}{Department of Astronomy, Graduate School of Science,

56: University of Tokyo, Hongo 7-3-1, Bunkyo-ku, Tokyo 113-0033, Japan}

57: \altaffiltext{3}{Hiroshima Astrophysical Science Center, Hiroshima

58: University, 1-3-1 Kagamiyama, Higashi-Hiroshima, Hiroshima 739-8526,

59: Japan}

60: \altaffiltext{4}{Max-Planck Institut f\"ur Astrophysik,

61: Karl-Schwarzschild-Strasse 1, Postfach 1317,  D-85741 Garching, Germany}

62: \altaffiltext{5}{Institute for the Physics and Mathematics of the

63: Universe, University of Tokyo, Kashiwa, Chiba 277-8582, Japan}

64: \altaffiltext{6}{Research Center for the Early Universe, Graduate School

65: of Science, University of Tokyo, Bunkyo-ku, Tokyo 113-0033, Japan}

66: \altaffiltext{7}{INAF-Osservatorio Astronomico di Trieste, via Tiepolo 11,

67: I-34131, Trieste, Italy}

68: \altaffiltext{8}{Kavli Institute for Theoretical Physics, University of California,

69: Santa Barbara, CA 93106}

70: \begin{abstract}

71: $UBVRI$ photometry and medium resolution optical spectroscopy of peculiar

72: Type Ia supernova SN 2005hk are presented and analysed, covering the

73: pre-maximum phase to around 400 days after explosion. The supernova is

74: found to be underluminous compared to "normal" Type Ia supernovae. The

75: photometric and spectroscopic evolution of SN 2005hk is remarkably similar to

76: the peculiar Type Ia event SN 2002cx. The expansion velocity of the supernova

77: ejecta is found to be lower than  normal Type Ia events. The  spectra

78: obtained $\gsim  200$ days since explosion do not show the presence of forbidden 

79: [\ion{Fe}{ii}], [\ion{Fe}{iii}] and [\ion{Co}{iii}] lines, but

80: are dominated by narrow, permitted \ion{Fe}{ii}, NIR \ion{Ca}{ii} and

81: \ion{Na}{i} lines with P-Cygni profiles.

82: Thermonuclear explosion model with Chandrasekhar

83: mass ejecta and a kinetic energy smaller ($\KE = 0.3 \times 10^{51} 

84: {\rm ergs}$) than that of canonical Type Ia supernovae is found to well

85: explain the observed bolometric light curve.

86: The mass of \Nifs\ synthesized in this explosion is $0.18 \Msun$.

87: The early spectra are successfully modeled with this less energetic model

88: with some modifications of the abundance distribution.

89: The late spectrum is explained as a combination of a photospheric

90: component and a nebular component.

91: 

92: 

93: \end{abstract}

94: 

95: \keywords{supernovae: general --- supernovae: individual (SN~2005hk)}

96: 

97: %%%%%%%%%%%%%%%%%%%%%%%% Main Text begins %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

98: 

99: \section{Introduction}

100: \label{sec:introduction}

101: 

102: An impressive homogeneity in the light curves and peak luminosities make

103: Type Ia

104: supernovae (SNe Ia) good candidates in the determination of the extragalactic

105: distance scale. Though a majority of the observed SNe Ia belong to the

106: "normal"   type (Branch et al.\ 1993), a number of studies indicate

107: significant

108: photometric as well as spectroscopic differences. For example, studies 

109:  of nearby supernovae by Li et

110: al.\ (2001) indicate that 64\% SNe Ia are "normal", while 20\% are of the

111: overluminous SN 1991T type and 16\% of the underluminous SN 1991bg type. The

112: peak absolute luminosities of SNe Ia are well correlated with their immediate

113: post-maximum decline rate, forming a photometric sequence from the

114: luminous blue

115: events with a relatively slow decline rate to the faster, red, subluminous

116: events (Hamuy et al.\ 1996a,b; Phillips et al.\ 1999).

117: 

118: However, there are a few SNe Ia that are known to deviate from this relation,

119: the most notable amongst them being SN 2002cx. This supernova was found to be

120: underluminous, but had a light curve decline rate $\Delta$m$_{15}(B)=1.29$,

121: comparable to normal SNe (Li et al.\ 2003). The early phase ($\lsim 100$ days

122: after explosion) spectra indicate line velocities lower by a factor of 2

123: compared to those of normal SNe Ia (Li et al.\ 2003; Branch et al.\ 2004).

124: Furthermore, the late phase ($\sim250$ days after explosion) spectra are also

125: quite dissimilar compared to normal SNe Ia and possibly consist of P-Cygni

126: profiles (Jha et al.\ 2006).

127: 

128: Interestingly, SN 2002cx is not a unique event. Jha et al.\ (2006) and

129: Phillips

130: et al.\ (2007) have shown that SN 2005hk shows photometric and spectroscopic

131: behaviour almost identical to that of SN 2002cx. Further, Jha et al.\

132: (2006)  have

133: shown that SNe 2003gq and 2005P also show spectra similar to SN 2002cx.

134: Spectropolarimetric observations of SN 2005hk at the early phases by Chornock

135: et al.\ (2006) indicate low polarization levels, indicating that the

136: peculiarities of SN 2002cx-like SNe do not result from an extreme

137: asphericity.

138: Based on the analyses of the early phase spectra of SN 2002cx, Branch et al.\

139: (2004) suggest that the observed lower line velocities are consistent with

140: the deflagration models of explosion. Jha et al.\ (2006) report a possible

141: detection of \ion{O}{i} lines in the late phase spectra and suggest large

142: scale mixing

143: in the central region, which is also consistent with three-dimensional 

144: (3D) deflagration models.  Phillips et al.\ (2007) find that, qualitatively the observed 

145: light curves of SN 2005hk are in reasonable agreement 

146: with  model calculations of a 3D deflagaration model that produces 

147: $\sim 0.2~\Msun$ of $^{56}$Ni.

148: 

149: Understanding the nature of this class of SNe Ia is thus quite important

150: for the study of  homogeneity and heterogeneity of SNe Ia. It may provide a

151: caution for the cosmological use of SNe Ia and a strong constraint to the

152: explosion models.

153: 

154: We present in this paper the photometric and spectroscopic development of

155: SN 2005hk over $\sim400$ days since explosion.

156: 

157: \section{Observations and data reduction}

158: \subsection{Photometry}

159: Supernova SN 2005hk was observed in UBVRI bands with the 2m Himalayan Chandra

160: Telescope (HCT) of the Indian Astronomical Observatory (IAO), Hanle, India,

161: using the Himalaya Faint Object Spectrograph Camera (HFOSC), equipped with a

162: SITe 2k x 4k CCD. The central 2k x 2k pixels were used for imaging which

163: corresponds to a field of view of $10^\prime \times 10^\prime$, with a scale

164: of 0.296 arcsec/pixel. The monitoring of the supernova started on 2005

165: November

166: 5 (JD 2,453,680), $\sim$ 4  days and $\sim$ 6  days after its discovery by Burket \&

167: Li (2005), and  Barentine, et al. (2005), and

168: continued until 2006 September 20 (JD 2,453,999). Standard fields PG0231+051,

169: PG1047+003, PG0942-029, PG0918+027 (Landolt 1992) were observed on 2005

170: November 24 and 2005 December 28 under photometric conditions, and were used

171: to calibrate a sequence of secondary standards in the supernova field.

172: 

173: 

174: \begin{figure}

175: \epsscale{1.0}

176: \plotone{f1.eps}

177: \caption{Identification chart for SN 2005hk. The stars used as local

178: standards are marked as numbers 1-11.

179: \label{fig:field} }

180: \end{figure}

181: 

182: 

183: 

184: The data reduction was done in the standard manner, using 

185: various tasks available within IRAF. The observed data were bias subtracted,

186: flat-field corrected and cosmic-ray hits removed. Aperture photometry was

187: performed on the standard stars, using an aperture radius determined using

188: the

189: aperture growth curve, and were calibrated using the average color terms and

190: photometric zero points determined on the individual nights. The $UBVRI$

191: magnitudes of the secondary standards in the supernova field, calibrated and

192: averaged over the two nights are listed in Table \ref{tab:calmag}. The

193: secondary sequence is shown marked in Figure \ref{fig:field}. The magnitudes

194: of the supernova and the local standards were estimated using the profile

195: fitting technique, with a fitting radius equal to the FWHM of the stellar

196: profile. The difference between the aperture and profile-fitting magnitude

197: was

198: obtained using  bright standards in the supernova field and this

199: correction

200: was applied to the supernova magnitude. The calibration of the supernova

201: magnitude to the standard system was done differentially with respect to the

202: local standards.

203: 

204: SN 2005hk was also observed on 2005 December 26, 2006 June 30 and

205: November 27 with the Faint Object Camera and

206: Spectrograph (FOCAS; Kashikawa et al.\  2002) mounted

207: on the 8.2 m Subaru Telescope in the $B$ and $R$ bands.

208: The local standard star 11 was calibrated as indicated in

209: Table 1 using the HCT/HFOSC photometry of the star 2

210: in Figure 1, then the $B$ and $R$ magnitudes were derived by

211: the point spread photometry in the DAOPHOT package of IRAF.

212: 

213: The observed supernova magnitudes are listed in Table \ref{tab:mag}.

214: 

215: 

216: \subsection{Spectroscopy}

217: Spectroscopic monitoring of the supernova using the HCT began on 2005

218: November

219: 4 and continued until 2006 January 12. The journal of observations is given

220: in Table \ref{tab:spec}. The HCT spectra cover the wavelength ranges

221: 3500-7000 \AA\ and 5200-9100 \AA\ at a spectral resolution of $\sim$ 7

222: \AA. The

223: data reduction was carried out in the standard manner using the tasks

224: available

225: within IRAF. The data were bias corrected, flat-fielded and the one

226: dimensional

227: spectra extracted using the optimal extraction method. Spectra of FeAr and

228: FeNe lamps were used for wavelength calibration. The instrumental response

229: curves were obtained using spectrophotometric standards observed on the same

230: night and the supernova spectra were brought to a relative flux scale. The

231: flux

232: calibrated spectra in the two regions were combined to a weighted mean to

233: give

234: the final spectrum on a relative flux scale. The spectra were brought to an

235: absolute flux scale using zero points derived by comparing the observed flux

236: in the spectra with the flux estimated using the photometric magnitudes.

237: 

238: 

239: Spectra during the late phases were  obtained with the Subaru/FOCAS on

240: 2005 December 26, 2006 June 30 and November 27. A 0.8 arcsec  width slit

241: was used

242: with the B300 grism, which gave a wavelength coverage of 4700-9000 \AA \

243: and a

244: spectral resolution of 11 \AA. The exposure times were 900 sec, 1,800 sec

245: and 3,600 sec

246: for 2005 Dec 26, 2006 Jun 30 and Nov 27, respectively. The instrumental

247: response

248: was corrected using data of a spectrophotometric standard star (GD 153 or

249: BD+28d4211). The

250: flux was then scaled to be consistent with the R-band photometry.

251: 

252: \section{$UBVRI$ Light Curves}

253: 

254: The early phase light curves (LCs) of SN 2005hk are discussed  by 

255: Phillips et al.\ (2007) and Stanishev et al.\ (2006). We present

256: here the LCs of this supernova during the early as well as the late

257: phase (ref. Table \ref{tab:mag}). Figure \ref{fig:lc} shows the LCs of

258: SN 2005hk in the $U$, $B$, $V$, $R$ and $I$ bands, respectively. The epoch of

259: maximum in each of these bands is estimated by a cubic spline fit to the

260: points around maximum. The supernova reached a maximum in the $B$ band

261: on JD\,2453685.34 with magnitude of

262: $15.91\pm0.03$. The maximum brightness in the $V, R$ and $I$ bands occured

263: $\sim$  +4, +6.5 and +8.5 days, relative to the maximum in $B$ band,

264: respectively. The decline rate parameter in the $B$ band estimated using our

265: data is $\Delta$m$_{15}(B)$ = 1.68$\pm$0.05. Our estimates of the dates of

266: maximum, apparent maximum magnitudes and the decline rate parameter are

267: listed in Table \ref{tab:lcparam}. These estimates are consistent with the

268: values reported by Phillips et al.\ (2007) and Stanishev et al.\ (2006).

269: 

270: \begin{figure}

271: \epsscale{1.0}

272: \plotone{f2.eps}

273: \caption{The $UBVRI$ \  LCs of SN 2005hk. The LCs have been shifted by the

274: amount indicated in the legend.

275: \label{fig:lc} }

276: \end{figure}

277: 

278: 

279: The LCs of SN 2005hk are compared with those of the normal SNe 1994D

280: [$\Delta m_{15}(B)=1.31\pm0.08$; Richmond et al.\ 1995] and 2003du

281: [$\Delta m_{15}{B}=1.04\pm0.04$; Anupama et al.\ 2005], the overluminous

282: SN 1991T

283: [$\Delta m_{15}(B)=0.95\pm0.05$; Lira et al.\ 1998], the subluminous SN

284: 1991bg

285: [$\Delta m_{15}(B)=1.93\pm0.08$; Filippenko et al.\ 1992; Turatto et al.\

286: 1996]

287: and the peculiar SN 2002cx [$\Delta m_{15}(B)=1.29\pm0.011$;

288: Li et al.\ 2003, $\Delta m_{15}(B)=1.7\pm0.1$; Phillips et al.\ 2007] (Figures

289: \ref{fig:compu}--\ref{fig:compb}).

290: The early LCs (Figure \ref{fig:compu}) indicate that the pre-maximum brightening of

291: SN 2005hk was faster than the luminous SN 1991T but comparable to the normal

292: SNe 1994D and 2003du. Following the maximum, during the initial 20 days, the

293: decline in $B$ was faster than in normal SNe, but similar to SN 2002cx,

294:  while the decline in $V$ was similar to the normal SN 1994D.

295: Subsequently, beyond 30 days past maximum, the decline in $B$ and $V$ was

296: slower than all the other SNe. The secondary peak seen in the $R$ and $I$

297: bands of SNe Ia was found to be absent in SN 2005hk, similar to the

298: subluminous SN 1991bg. However, beyond $\sim$ 50 days past maximum the $R$ LC

299: of SN 2005hk had a decline similar to SN 1991T, which is slower than that

300: seen in the normal and subluminous SNe (Figure \ref{fig:compb}). The evolution of 

301: the $U$ light curve

302: of SN 2005hk was found to be faster than both the normal and the luminous

303: types.

304: 

305:  The early phase evolution of SN 2005hk is very similar to that of SN 2002cx.

306: Both objects have  similar $\Delta$m$_{15}(B)$. SN 2005hk has a 

307: $\Delta$m$_{15}(B)$ of $1.68$, while SN 2002cx has a value of 1.7, as

308: indicated by the the revised photometry by Phillips et al.\ (2007). 

309: On the other hand, at

310: epochs beyond 15 days past maximum, SN 2005hk declines somewhat slower

311: than SN 2002cx. The decline rate during $\sim 20-50$ days past $B$

312: maximum, in the $B$, $V$, $R$ and $I$ bands are, respectively, 0.021, 0.027,

313: 0.031 and 0.031 mag day$^{-1}$, while the corresponding decline rates in

314: SN 2002cx during the same phase are 0.037, 0.035, 0.041 and 0.035. It may be

315: noted here that no galaxy template subtraction is done for the data

316: presented here, and the SN magnitudes may be affected by the host galaxy at

317: the later phases. However, as suggested by Phillips et

318: al.(2007), neither $K$ correction nor the galaxy background contamination is

319: sufficient to explain the discrepancies seen in the LCs of SN 2005hk and

320: SN 2002cx.

321: 

322: \begin{figure}

323: \epsscale{1.0}

324: \plotone{f3.eps}

325: \caption {Early phase ($< 50$~days) LC of SN 2005hk together with those of 

326: SNe 2002cx, 1991bg, 1991T,  2003du and  1994D. The ordinate in each panel is the

327: magnitude below the respective maximum, and the abscissae represent the days

328: since respective $B$ maximum. For clarity, the LCs have been shifted by the amount

329: indicated in the legend.

330: \label{fig:compu} }

331: \end{figure}

332: 

333: 

334: \begin{figure}

335: \epsscale{1.0}

336: \plotone{f4.eps}

337: \caption{Late phase ($>50$~days) LC of SN 2005hk together with those of 

338: SNe 2002cx, 1991bg, 1991T, 2003du and  1994D. The ordinate in each panel is the

339: magnitude below the respective maximum, and the abscissae represent the days

340: since respective $B$ maximum. For clarity, the LCs have been shifted by the amount

341: indicated in the legend.

342: \label{fig:compb}}

343: \end{figure}

344: 

345: 

346: \section{Reddening and Color curves}

347: 

348: The reddening within our Galaxy in the direction of SN 2005hk is

349: $E(B-V )_{\rm{Gal}}=0.022$ (Schlegel et al.\ 1998). The reddening within the

350: host galaxy may be estimated from the methods suggested by Phillips

351: et al.\ (1999), Altavilla et al.\ (2004) and Lira\ (1996). The Phillips

352: et al.\ (1999) and the Altavilla et al.\ (2004) methods give

353: $E(B-V)_{\rm{Host}}$ = 0.20

354: and 0.09, respectively, while the Lira\  (1996) method indicates a higher

355: value.

356: Spectropolarimetric observations of SN 2005hk (Chornock et al.\  2006)

357: indicate an interstellar polarization of 0.27\%

358: produced by the host galaxy which, for

359: the standard dust polarization efficiency, corresponds to

360: $E(B-V)_{\rm{Host}} = 0.09$. Owing to the peculiar nature of SN 2005hk, the

361: methods based on normal SNe Ia may not be applicable. Hence, adopting the

362: estimate for the host galaxy extinction as indicated by the

363: spectropolarimetric observations, we estimate the total

364: reddening towards SN 2005hk to be $E(B-V)_{\rm{tot}} = 0.11$.

365: 

366: The $(U-B)$, $(B-V)$, $(V-R)$ and $(V-I)$ color curves of SN 2005hk are

367: shown

368: in Figure \ref{fig:color}, compared with other SNe. The color curves of

369: all SNe have been

370: corrected for reddening using the Cardelli et al.\ (1989) extinction law and

371: the $E(B-V)$ values of  $E(B-V)$ = 0.13 for

372: SN 1991T (Phillips et al.\ 1992), $E(B-V)$ = 0.034 for SN 2002cx (Li et

373: al.\ 2003),

374: $E(B-V)$ = 0.04  for SN 1994D (Richmond et al.\ 1995), $E(B-V)$ = 0.05 for

375: SN 1991bg  (Turatto et al.\ 1996) and $E(B-V)$ = 0.01 for SN 2003du (Anupama

376: et al.\ 2005).

377: 

378: \begin{figure}

379: \epsscale{1.0}

380: \plotone{f5.eps}

381: \caption{The $U-B$, $B-V$, $V-R$ and $V-I$ color curves of SN 2005hk. Also

382: shown for comparison are the color curves of SNe 2002cx, 1991T, 1991bg, 2003du and

383: 1994D. The abscissae correspond to days since respective $B$ maximum.

384: \label{fig:color}}

385: \end{figure}

386: 

387: The color curves of SN 2005hk are very similar to SN 2002cx. The $(U-B)$

388: color

389: evolution is very similar to SN 1991T. The $(B-V)$ and $(V-R)$ color

390: evolution

391: of SN 2005hk in the pre-maximum and maximum phase are very similar to the

392: other

393: SNe Ia except SN 1991bg, which had a considerably red color. Beyond $\sim 5$

394: days, until $\sim 20$ days after $B$ maximum, both $(B-V)$ and $(V-R)$

395: colors get increasingly redder compared to other SNe, but are still bluer

396: compared to SN 1991bg. The $(V-I)$ color on the other hand gets redder than 

397: other SNe $\sim$ 5 days before maximum, and also does not show the dip seen in 

398: the color curves of other SNe Ia at $\sim$ 10 days post-maximum. Beyond 

399: $\sim$20 days after maximum, the $(V-R)$ and $(V-I)$ colors of SN 2005hk 

400: follow the trend of other supernovae and become marginally blue, while 

401: $(B-V)$ continues to be redder than other SNe, including SN 1991bg.

402: 

403: Interestingly, despite a very similar color evolution during the pre-maximum

404: phase to $\sim 20$ days after $B$ maximum, SN 2005hk appears to be marginally

405: bluer than SN 2002cx at later phases.

406: 

407: \section{Bolometric light curve}

408: 

409: \subsection{Behavior of bolometric light curve}

410: 

411: The bolometric LC of SN 2005hk is estimated using the $UBVRI$

412: photometry  presented here along with the NIR $YJH$ photometry reported by

413: Phillips et al.\ (2007). The magnitudes were reddening corrected using the

414: value estimated in the previous section and the Cardelli et al.\ (1989) 

415: extinction law. The corrected magnitudes were then converted to monochromatic 

416: fluxes using the zero points from Bessell et al.\ (1998). The bolometric 

417: fluxes were derived by fitting a spline curve to the $U, B, V, R, I, Y, J$ and 

418: $H$ fluxes and integrating over the wavelength range 3100\AA \ to 

419: $1.63\micron$. In the later phases when only $B, V, R$ or $V, R$

420: magnitudes were available, the bolometric flux were estimated by applying

421: a bolometric correction to the available magnitudes. Since SN 2005hk belongs

422: to a rare class of SNe Ia, the bolometric correction is derived from the

423: light curve of SN 2005hk itself rather than assuming the corrections based on

424: normal SNe Ia (Suntzeff\ 1996; Contrado et al.\ 2000). 

425: The bolometric 

426: corrections were estimated based on the last three points in the LC for which 

427: the $BVRIJH$ bolometric flux is available. A simple average of the bolometric

428: corrections for these days give values of $B.C = -0.926\pm0.089$ in $B$,

429: $B.C=0.294\pm0.045$ in $V$ and $B.C=0.677\pm0.027$ in $R$ band. 

430: 

431: The bolometric LC is plotted in Figure \ref{fig:modelLC} (red circles). 

432: Also plotted in the figure are the bolometric LCs of 

433: the normal (but slightly underluminous) Ia SN 1992A (black squares), 

434: the subluminous Ia SN 1991bg (blue triangles) and SN 2002cx (green triangles).

435: The peak luminosity ($M_{bol} = -17.7$) is fainter than the canonical 

436: value for normal SNe Ia.

437: It suggests that a smaller amount of \Nifs\ is

438: synthesized during the explosion. 

439: 

440:  The decline of the bolometric LC

441: after the maximum is slower than that of SN 1992A

442: although they have a similar maximum luminosity, 

443: as also shown by Phillips et al. (2007).

444: Our observations extend the bolometric LC to $\sim 400$ days since explosion. 

445: At such late epochs, the bolometric luminosity of SN 2005hk is

446: still brighter than that of SN 1992A.

447: In addition, the difference in the luminosity between the two SNe

448: becomes large, reaching $\sim 1$ magnitude at $> 250$ days after

449: the explosion. The bright late phase luminosity indicates a more efficient

450: trapping of the $\gamma$-rays from decaying \Cofs\ in SN 2005hk compared to

451: SN 1992A. 

452: 

453: \begin{figure}

454: \epsscale{1.0}

455: \plotone{f6.eps}

456: \caption{

457:  Bolometric LC of SN 2005hk (red circles) compared with a synthetic LC

458: with the E03 model (red line).

459: Black squares and blue triangles are bolometric LCs of

460: SNe 1992A and 1991bg (Suntzeff 1996), respectively.

461: Model LCs for these two SNe are shown in black and blue lines, respectively

462: (based on W7 but different mass of \Nifs, see Table \ref{tab:param}).

463: Green triangles are the bolometric LC of SN 2002cx.

464: In order to fix the rise time of the LCs of SNe 1992A and 1991bg,

465: $B$ band rise time is assumed as $t_{rise} = 19.5 \times s$ days

466: (Conley et al. 2006),

467: where $s$ is stretch factor of $B$ band LC (Goldhaber et al. 2001).

468: The $B$ band rise times of SNe 1992A ($s = 0.80$) and 1991bg ($s = 0.68$)

469: are estimated as 15.6 and 13.2 days, respectively.

470: Since the LC of SN 2005hk is deviated from the normal sequence of SNe Ia,

471: we simply assume  $t_{rise} = 17$ days as in Phillips et al. (2007).

472:  The magenta line shows the synthetic LC with the E008 model.

473: \label{fig:modelLC}}

474: \end{figure}

475: 

476: 

477: \subsection{Light curve modeling}

478: 

479: 

480: Phillips et al. (2007) suggest that qualitatively the observed properties of SN 2005hk 

481: are consistent  with 3D deflagration model 

482: (\eg R\"opke et al. 2006, Blinnikov et al.\ 2006), showing the bolometric 

483: and multi-color LC model until $\sim 80$ days after the explosion.

484: Branch et al. (2004) and Jha et al. (2006) also suggest

485: a 3D deflagration model for SN 2002cx. Here we give the first investigation of 

486: the bolometric LC of this class of SNe until $\sim 400$ days after the 

487: explosion.

488: 

489: 

490: We use an LTE radiation transfer code (Iwamoto et al.\ 2000)

491: for the computation of the bolometric LCs.

492: For the $\gamma$-ray transport, a gray atmosphere is assumed. 

493: For optical radiation transport, electron-scattering and line opacity 

494: are taken into account.  The electron-scattering opacity is evaluated 

495: by solving the Saha equation 

496: while the line opacity is crudely assumed to be constant 

497: ($0.1\ {\rm cm^2 \ g^{-1}}$; Iwamoto 1997) for simplicity.

498: We use the W7 deflagration model (Nomoto et al.\ 1984) as a 

499: standard density structure and abundance distributions.

500: 

501: First, the mass of \Nifs\ [$M$(\Nifs)] is varied from the W7 value

502: ($M$(\Nifs) $= 0.59 \Msun$) to obtain a good agreement with the maximum 

503: luminosity of SNe Ia shown in Figure \ref{fig:modelLC}.

504: In this procedure, the total mass of iron group elements is kept 

505: constant to conserve the explosion energy. The black and blue lines in Figure 

506: \ref{fig:modelLC} are the synthetic LCs with $\Mni = 0.17$ and $0.088 \Msun$, 

507: respectively. 

508: Both at the early phase (ref.\ inset in Fig. \ref{fig:modelLC}) 

509: and the late phase, the LCs are in good agreement with SNe 1992A and  1991bg 

510: (Suntzeff\ 1996), respectively

511: \footnote{ This shows  that our crude assumption of constant line opacity

512: ($0.1\ {\rm cm^2 \ g^{-1}}$) is reasonable.}. 

513: However, the model sequence with various $\Mni$ never 

514: reproduces the brightness of SN 2005hk at $\gsim 50$ days after 

515: the explosion.

516: 

517:  As also suggested by Phillips et al. (2007),

518: we now try to explain the LC of SN 2005hk with less energetic models

519: to fit the LC around maximum.

520: We construct a less energetic model simply 

521: by scaling the structure of W7 homologously. 

522: The mass of the burned material should be 

523: adjusted accordingly by taking nuclear energy production into account.

524: In our model, the mass fraction of each burned element is reduced 

525: by the same fraction

526: [$(\KE + E_{\rm B} )/(E_{\rm W7} + E_{\rm B})$] in every radial shell. 

527: Here $\KE$, $E_{\rm B}$ and $E_{\rm W7}$ are 

528: the kinetic energy of the SN ejecta, the binding energy of a progenitor WD 

529: ($\sim 0.5 \times 10^{51}$ ergs), and the kinetic energy of W7 model 

530: ($\sim 1.3 \times 10^{51}$ ergs), respectively. 

531: The mass of the burned elements reduced by the procedure above

532: is compensated by unburned C and O, with the mass fraction of

533: X(C)=0.5 and X(O)=0.5.

534: %The total mass fraction reduced by the procedure above is compensated 

535: %by C (50 \%) and O (50 \%). 

536: 

537: The red line in Figure \ref{fig:modelLC} is the synthetic LC with 

538: $\Mni = 0.18 \Msun$ and $\KE = 0.3 \times 10^{51} {\rm ergs}$. 

539: This kinetic energy is similar to that 

540: of the 3D deflagration model used in Phillips et al.\ (2007).

541:  The density structure of this model is given in Figure \ref{fig:dens}

542: together with that of the original W7.

543: This model gives a reasonable fit 

544: to SN 2005hk at late phase as well as at early phase.

545:  In the figure, we assume $B$ maximum of SN 2005hk occurs 

546: 17 days after the explosion, which gives the best fit of the LC

547: (see also Phillips et al. 2007 

548: \footnote{The rise time of $15 \pm 1$ days is estimated by 

549: Phillips et al. (2007), using $3\sigma$ upper limit 

550: at 15 days before B maximum.  

551: Our model LC assuming the rise time of 17 days is consistent 

552: with the upper limit.}).

553: The mass of \Nifs\ derived from the model is somewhat smaller  than

554: $0.24 \Msun$ derived by Phillips et al. (2007).

555: It may be caused by a different distribution of \Nifs\ in the ejecta.

556: Since \Nifs\ in the model in Phillips et al. (2007) is extended

557: to the outer layers, a part of \Nifs\ does not contribute to the optical light.

558: In this paper, we call our scaled, less energetic model E03.

559: 

560: \begin{figure}

561: \epsscale{1.0}

562: \plotone{f7.eps}

563: \caption{

564: The density structure of model E03 compared with that of W7

565: (1 day after the explosion).

566: The kinetic energy of model E03 is $\KE = 0.3 \times 10^{51} {\rm ergs}$.

567: The density structure is scaled from that of W7 homologously.

568: This makes the density at the inner part higher than the original.

569: \label{fig:dens}}

570: \end{figure}

571: 

572: 

573:  The behavior of the model E03 LC is understood by simple analysis.

574: If the opacity is assumed to be constant, 

575: the time scale of the bolometric LC  ($\tau_{LC}$) is scaled as

576: $\tau_{\rm LC} \propto \Mej^{3/4} \KE^{-1/4}$ (Arnett 1982), 

577: where $\Mej$ is the mass of the ejecta.

578: Therefore, if $\Mej$ is fixed (Chandrasekhar mass), 

579: the lower energy gives a slow decline after the maximum.

580: The late phase evolution of LC is mainly determined by the 

581: optical depth to $\gamma$-rays.

582: Since the optical depth is higher for lower energetic ejecta

583: ($\tau \propto \Mej^2 \KE^{-1}$, \eg Maeda et al. 2003),

584: the LC of model E03 declines slowly at late phase.

585: 

586: The characteristic parameters of the LC models are summarized 

587: in Table \ref{tab:param}. 

588: In the lower energetic model, unburned C and O should be 

589: abundant, being $\sim 0.8 \Msun$ in total. 

590:  The presence of large amount of unburned C+O

591: is also seen in 3D deflagration model (R\"opke et al. 2006), 

592: which is used in Phillips et al. (2007).

593: Although our model is very simple, it mimics the properties

594: of the  3D deflagration model.

595: Our model is further verified by spectral modeling in \S 6.3.

596: 

597: 

598: 

599: \section{The spectrum}

600: \subsection{Evolution during early phases}

601: 

602: The spectral evolution of SN 2005hk during the early phases are shown in

603: Figure \ref{fig:spec1} -- Figure \ref{fig:spec3}. The pre-maximum spectra 

604: (ref. Fig.\ref{fig:spec1}) show a blue continuum, dominated by \ion{Fe}{iii}

605: lines, with weak \ion{Si}{ii} and \ion{Ca}{ii} H\&K absorption (see \S 6.3.1).

606: Comparing the pre-maximum spectrum with other normal Type Ia as well as 

607: peculiar Type Ia (Fig. \ref{fig:comp_photo_03duu1}), it is seen that the 

608: spectrum of SN 2005hk is very similar to that of SN 1991T. However, the minima 

609: of the \ion{Ca}{ii} H\& K, \ion{Fe}{ii}, \ion{Fe}{iii} and \ion{Si}{ii} lines,

610: indicate that SN 2005hk has much lower expansion velocities that decrease from 

611: $\sim 6900$ km s$^{-1}$ on day $-6$ to $\sim 6200$ km s$^{-1}$ on day $-4$. The

612: similarity of the pre-maximum spectrum of SN 2005hk with SN 1991T-like events

613: has also been noted by Phillips et al.\ (2007) \& Stanishev et al.\ (2006).

614: 

615: \begin{figure}

616: \epsscale{1.0}

617: \plotone{f8.eps}

618: \caption{Pre-maximum spectral evolution  of SN 2005hk. The phases marked

619: are relative to date of $B$ maximum. The spectra have  been corrected for

620: the redshift of the host galaxy, but not corrected for the reddening. For

621: clarity the spectra have been shifted vertically. The telluric lines have 

622: not been

623: removed from the spectra and are marked with the symbol {\small $\oplus$}.

624: \label{fig:spec1}}

625: \end{figure}

626: 

627: \begin{figure}

628: \epsscale{1.0}

629: \plotone{f9.eps}

630: \caption{Pre-maximum spectrum of SN 2005hk compared with those of 

631: Type Ia SNe 1991T and  2003du at similar epoch.  The phases marked are

632: relative to date of $B$ maximum. The spectra have  been corrected for

633: the redshift of the host galaxy, but not corrected for the reddening.

634: For clarity the spectra have been shifted vertically. The telluric lines 

635: have not been removed from the spectra of SNe 2005hk and  1991T, they are 

636: marked with  the symbol  {\small $\oplus$}.

637: \label{fig:comp_photo_03duu1}}

638: \end{figure}

639: 

640: 

641: 

642: The post-maximum evolution of the spectrum of SN 2005hk is presented in

643: Figures \ref{fig:spec2} and \ref{fig:spec3}. The spectrum and its evolution 

644: are quite different from normal SNe Ia, as can be seen from Figure

645: \ref{fig:comp_photo_03duu2}. The spectrum and its evolution closely matches

646: that of SN 2002cx (Li et al.\ 2003, Branch et al.\ 2004, Jha et al.\ 2006).

647: The spectrum is dominated by \ion{Fe}{ii} and lines due to \ion{Co}{ii},

648: \ion{Na}{i} and \ion{Ca}{ii} are also clearly present (see \S 6.3.1).

649: Some \ion{Cr}{ii} lines may contribute to the absorption around 4800 \AA\

650: (Branch et al.\ 2004). The photospheric velocity decreases to 

651: $\sim 4000$ km s$^{-1}$ by day $+10$. The flux in the blue continuum is found 

652: to decline steadily, and by day $+14$, the continuum is quite weak below 

653: 4500\AA. This could also be an effect of line blanketing due to \ion{Fe}{ii} 

654: lines. The spectrum remains almost unchanged from day $+10$ to day $+24$.

655: 

656: \begin{figure}

657: \epsscale{1.0}

658: \plotone{f10.eps}

659: \caption{Spectral evolution  of SN 2005hk from phase $+3$ days to

660: $+14$ days relative to date of $B$ maximum. The spectra have  been

661: corrected for the redshift of the host galaxy, but not corrected for

662: the reddening. For clarity the spectra have been shifted vertically.

663: The telluric lines 

664: have not been removed from the spectra  and are marked with 

665: the symbol {\small $\oplus$}. Line identification is based on

666: Branch et al. (2004)

667: \label{fig:spec2}}

668: \end{figure}

669: 

670: 

671: \begin{figure}

672: \epsscale{1.0}

673: \plotone{f11.eps}

674: \caption{Post-maximum spectrum   of SN 2005hk compared with those

675: of Type Ia SNe 1991T and  2003du at similar epochs.  The phases

676: marked are relative to date of $B$ maximum. The spectra have  been

677: corrected for the redshift of the host galaxy, but not corrected for

678: the reddening. For clarity the spectra have been shifted vertically.

679: The telluric lines  have not been removed from these spectra 

680: and are marked with the symbol {\small $\oplus$}.

681: \label{fig:comp_photo_03duu2}}

682: \end{figure}

683: 

684: 

685: \begin{figure}

686: \epsscale{1.0}

687: \plotone{f12.eps}

688: \caption{Spectral evolution  of SN 2005hk from phase $+21$ days to $+62$ days

689: relative to date of $B$ maximum. The spectra have been corrected for the

690: redshift of the host galaxy, but not corrected for the reddening. For clarity

691: the spectra have been shifted vertically.  The telluric lines 

692: have not been removed from the spectra of SN 2005hk taken with HCT and are 

693: marked with the symbol {\small $\oplus$}. The line identification is based on

694: Branch et al. (2004).

695: \label{fig:spec3}}

696: \end{figure}

697: 

698: The spectrum of day $+62$ is very similar to that of day $+45$, except for the

699: narrowing of lines (ref. Fig.\ref{fig:spec3}). 

700: The lines in the 6500-6800 \AA, identified with \ion{Co}{ii} 

701: (Branch et al. 2004), which by day $+32$ started weakening steadily with time

702: (but they could be \ion{C}{ii}, see \S 6.3.1). 

703: 

704: \subsection{The late phases}

705: 

706: The spectra of SN 2005hk were obtained during the late phase on days 

707: $+228$ and $+377$ (Fig. \ref{fig:spec4}). As in the case of the early phase spectra, the  

708: spectrum of SN 2005hk during these phases is very different from the spectra 

709: of normal SNe Ia at similar phases.

710: The spectrum of SN 2005hk is compared with those of other normal and peculiar

711: Type Ia supernovae in Figure \ref{fig:comp_nebular}. The difference between the

712: spectrum of SN 2005hk and spectra of other Type Ia supernovae is obvious.

713: While the other SNe Ia show blends of strong forbidden lines due to iron

714: [\ion{Fe}{ii}] (5159\AA), [\ion{Fe}{iii}] (4500-4800 \AA) and forbidden lines 

715: due to cobalt [\ion{Co}{iii}] (5890\AA) (Kuchner et al.\ 1994) 

716:  in their late phase spectrum, no 

717: signature of these lines is seen in the spectrum of SN 2005hk.

718: 

719: \begin{figure}

720: \epsscale{1.0}

721: \plotone{f13.eps}

722: \caption{Late phase spectrum of SN 2005hk at $+228$ days and $+377$ days

723: relative to date of $B$ maximum. The spectra have  been corrected for

724: the redshift of the host galaxy, but not corrected for the reddening.

725: For clarity the spectra have been shifted vertically. The telluric 

726: lines have been removed from the spectrum.

727: The line identification is based on Jha et al. (2006).

728: \label{fig:spec4}}

729: \end{figure}

730: 

731: 

732: 

733: The late time spectrum of SN 2005hk is dominated by \ion{Fe}{ii} lines. The 

734: \ion{Ca}{ii} NIR lines are  strong and \ion{Na}{i} is also present. The 

735: spectra of days $+228$ and $+377$ (ref. Fig.\ref{fig:spec4}) also show 

736: forbidden lines due to [\ion{Ca}{ii}] 7291, 7234 \AA\ and  

737: [\ion{Fe}{ii}] at 7155 \AA\ and 7453 \AA. The ratio of the fluxes of 

738: \ion{Ca}{ii} NIR and the [\ion{Ca}{ii}] lines on day $+228$ indicate densities 

739: of $\sim 10^9$ cm$^{-3}$ in the line emitting region. By day $+377$, the 

740: [\ion{Ca}{ii}] lines are stronger than the \ion{Ca}{ii} NIR lines, and the 

741: line ratios imply temperatures $\la 4500$~K and density  

742: $\sim 10^8$ cm$^{-3}$ (Fransson \& Chevalier\ 1989). The mass ejected during 

743: the explosion can also be estimated using the number density arrived at with 

744: the Ca line ratios. Assuming that the ejecta is moving with a constant velocity

745: of 1000 km s$^{-1}$ over a period of $\sim$ 380 days after the explosion and

746: a mean atomic weight $A = 40$, the estimated mass of the ejecta is

747: $\sim 0.25 M_\odot$. As pointed out by Jha et al.\ (2006) the mass of the

748: ejecta arrived at is indicative as it does not take into account factors such

749: as clumping or more complicated structures in the line emitting regions.

750: It may however be noted that a high concentration of the mass in the central

751: part is also suggested by the spectral modeling (\S 6.3). 

752: 

753: 

754: \begin{figure}

755: \epsscale{1.0}

756: \plotone{f14.eps}

757: \caption{Late phase spectrum of SN 2005hk at $+228$ days and $+377$ days

758: relative to date of $B$ maximum compared with those of other Type Ia SNe 2003du, 2002er, 1998bu, 1991T and 1991bg.

759: The spectra have  been corrected for the redshift of the host galaxy, but

760: not corrected for the reddening. For clarity the spectra have been shifted

761: vertically.

762: \label{fig:comp_nebular}}

763: \end{figure}

764: 

765: 

766: The shift in the central wavelength of the [\ion{Ca}{ii}]  forbidden emission

767: lines indicates an

768: expansion velocity of $\sim 500$ km s$^{-1}$ on day $+228$ and

769: $\sim 300$ km s$^{-1}$ on day $+377$. The full width half maximum (FWHM) of

770: the emission lines indicate a velocity dispersion of $\sim 1000$ km s$^{-1}$

771: on day $+228$ and $\sim 500$ km s$^{-1}$ on day $+377$. Similarly, the

772:  emission component of \ion{Ca}{ii} NIR 8542 \AA\ line indicates an expansion

773: velocity of $\sim

774: 235$ km s$^{-1}$ on day $+228$ and $\sim 210$ km s$^{-1}$ on day $+377$, with 

775: velocity dispersions $\sim 2100$ km s$^{-1}$ and $\sim 1200$ km s$^{-1}$,

776: respectively. Low velocity \ion{O}{i} at 7002 \AA\ and 7773 \AA \ absorption 

777: features are clearly identified in the late phase spectrum. These features were 

778: identified, for the first time in any SNe Ia, by Jha et al.\ (2006) in the 

779: ''nebular'' spectrum of SN 2002cx. The velocity of these lines, measured at the 

780: minimum is $\sim$ 800 km s$^{-1}$. Though there is an overall similarity 

781: between the spectra of SN 2002cx and SN 2005hk at late phases, there are some 

782: noteworthy differences also. Comparing the spectrum of SN 2005hk on day $+228$ 

783: with the $+227$ day spectrum of SN 2002cx (Jha et al.\ 2006), it is found that 

784: the expansion velocities and the velocity dispersions are higher in SN 2005hk. 

785: The velocity dispersion based on the [\ion{Ca}{ii}] and the \ion{Ca}{ii} lines 

786: are in fact almost twice that estimated in SN 2002cx.

787: 

788: \subsection{Spectral modeling}

789: \subsubsection{Early Phases}

790: \label{sec:spearly}

791: 

792: The observed spectra of SN 2005hk are further investigated 

793: by spectral modeling.

794: For this purpose, we compute synthetic spectra using the Monte Carlo spectrum 

795: synthesis code developed by Mazzali \& Lucy\ (1993), Lucy\ (1999) and Mazzali\ 

796: (2000). We create model spectra at 5 epochs and compare them with the observed 

797: spectra ($-6$, $+3$, $+14$, $+24$ and $+45$ days from $B$ maximum).

798:  The rise time is assumed to be 17 days, as in the LC modeling (\S 5.2).

799: 

800: The Monte Carlo spectrum synthesis code assumes a spherical, sharply defined 

801: photosphere in the ejecta as an inner boundary. By tracing rays of a number of 

802: photon packets emitted from the photosphere, the temperature structure in the 

803: SN atmosphere is computed through the flux at each radial point (see eq. (1) 

804: and (2) in Mazzali \& Lucy 1993).

805: Radiative equilibrium is assumed for the 

806: temperature determination. For the interaction between photons and the SN 

807: ejecta, electron scattering and line scattering are taken into account. 

808: Sobolev approximation is used for line scattering. 

809: Sobolev optical depth is evaluated 

810: by computing ionization fractions and electron populations of each ion for a 

811: given temperature structure. 

812: Details of the physical assumptions adopted in the 

813: code are summarized in Mazzali \& Lucy\ (1993), Lucy\ (1999) and Mazzali\ 

814: (2000).

815: 

816: The input parameters of the code are the position 

817: (\ie velocity, thanks to the homologous expansion) 

818: of the photosphere ($v_{\rm ph}$) and the bolometric luminosity ($L$). 

819: The parameter $L$ is the emergent luminosity.

820: Since some photons emitted from the photosphere 

821: are back scattered into the photosphere, 

822: the luminosity at the photosphere 

823: (total photon luminosity emitted from the photosphere), 

824: is higher than the input bolometric luminosity $L$

825: (Mazzali \& Lucy 1993).

826: As a density structure we use the less energetic model that 

827: gives the best agreement with the observed LC 

828: (model E03, see \S 5 and Table 5).

829: 

830: \begin{figure}

831: \epsscale{1.0}

832: \plotone{f15.eps}

833: \figcaption{

834: Comparison between the observed early phase spectra of SN 2005hk at

835: ($-6$, $+3$, $+14$, $+24$ and $+45$ days from $B$ maximum,

836: from top to bottom) and the synthetic spectra.

837: Lines that make a major contribution to the absorptions are marked.

838: \label{fig:early}}

839: \end{figure}

840: 

841: 

842:  The mass fractions of elements are also required for the computation

843: of the synthetic spectra.

844: We treat the mass fractions as parameters

845: rather than using the abundance distribution in the original LC model.

846: The mass fractions of elements are assumed to be homogeneous

847: above the photosphere.

848: We determine the mass fraction of each element 

849: to give a good fit of the observed line depths.

850: As time goes by, the photosphere recedes

851: and the absorption lines are formed in the inner part of the ejecta. 

852: Nevertheless, we always assume a homogeneous abundance distribution

853: at each epoch, although we change the mass fractions of the elements.

854: The mass fractions derived by the modeling of each spectrum are 

855: thought to be a representative abundance ratio near the photosphere

856: because the absorption lines are formed near the photosphere

857: most efficiently.

858: 

859: The parameters in the best fitting cases are summarized 

860: in Table \ref{tab:spparam}.

861: Figure \ref{fig:early} shows the comparison between observed spectra and 

862: synthetic spectra. Synthetic spectra are reddened with 

863: $E(B-V)_{\rm Gal} = 0.022$ and $E(B-V)_{\rm Host} = 0.09$ (\S 4.). The 

864: distance modulus $\mu = 33.46$ is assumed. In the figure, the flux and models

865:  at epoch are scaled and shifted for clarity.

866: 

867: The spectrum of SN 2005hk at day $-6$ is in good agreement with a synthetic 

868: spectrum computed with $\log L$ (\ergs) $=42.56$ 

869: and $v_{\rm ph}$ = $6500$ \kms.

870: The strong absorption features around 4300\AA\ and 5000\AA\ are mainly due to 

871: \ion{Fe}{iii} (Mazzali et al. 1995). The photospheric temperature is computed 

872: as $T = 17000$ K by taking into account the back scattering effect (Mazzali \& 

873: Lucy\ 1993). This high temperature makes the \ion{Fe}{iii} lines strong and 

874: also reproduces the color of the observed spectrum.

875: 

876: The temperature at the photosphere is relatively higher than that of 

877: normal Type Ia SNe at similar epoch, and it is comparable to

878: the temperature at very early phase ($\sim -10$ days; Tanaka et al. 2007).

879:  The weakness of \ion{Si}{ii} suggests 

880: that the outermost ejecta of SN 2005hk do not contain much Si 

881: (X(Si) $\sim$ 0.02) as in normal SNe Ia (X(Si) $\gsim$ 0.5). The \ion{S}{i} 

882: lines are similarly weak. The mass fraction of S 

883: is estimated as X(S) $\sim 0.02$.

884: The \ion{Ca}{ii} lines are also very weak and there is no clear evidence of 

885: high velocity absorptions that is usually seen at pre-maximum epochs

886: (Mazzali et al. 2005; Tanaka et al. 2006). No C lines are seen, which gives an 

887: upper limit of the C mass fraction X(C) $\lsim 0.001$. The rest of the ejecta 

888: consist of oxygen, whose mass fraction is estimated as X(O) $\sim 0.86$.

889: 

890: The day $+ 3$ spectrum matches a synthetic spectrum with 

891: $\log (L)$ \ergs $= 42.66$ and $v_{\rm ph} =6000$ \kms. The temperature at the 

892: photosphere decreases ($T \sim 9,000$K) from pre-maximum epochs. The low 

893: temperature allows some singly ionized ions such as \ion{Si}{ii}, \ion{Ca}{ii} 

894: and \ion{Fe}{ii} to appear in the spectrum.

895: 

896: The feature at 6300 \AA\ mainly consists of the \ion{Si}{ii} $\lambda$6355,

897: and it becomes stronger from pre-maximum epochs (Figs. \ref{fig:spec1} and 

898: \ref{fig:spec2}). After the maximum, Fe lines begin to contribute to this 

899: feature (Figs.\ref{fig:spec2} and \ref{fig:spec3}). No C lines are visible, 

900: giving an upper limit of C mass fraction X(C) $\lsim 0.001$. The mass fractions of 

901: Si, Ca, Fe and Ni can be estimated by the visible lines (Table 

902: \ref{tab:spparam}). The mass fraction of Si, stable Fe and \Nifs\ required for 

903: the fitting of this spectrum are larger than those for pre-maximum spectra.

904: Note again that we assume spatially homogeneous abundance distribution

905: for each epoch.

906: 

907: After maximum brightness, the evolution of the spectral shape is not

908: significant (Fig. \ref{fig:spec2}). Almost all the absorption features are 

909: identified and they are mostly \ion{O}{i}, \ion{Ca}{ii}, \ion{Fe}{ii} and 

910: \ion{Co}{ii} lines. The spectra are reproduced by almost the same element 

911: fractions (Table \ref{tab:spparam}). Estimating the mass fraction of stable 

912: Fe separately with \Nifs\ becomes difficult because more than $25 \%$ of \Nifs, 

913: which is one of the dominant elements, has decayed into \Fefs\ at $\gsim 40$ 

914: days after the explosion.

915: 

916: An interesting feature is a weak absorption at 6400 \AA\ in the spectra at day

917: +14 (and possibly in day +24). Branch et al. (2004) suggested 

918: that this is a \ion{Co}{ii} line.

919: However, if we reproduce this feature by Co, the line

920: blocking at $\lsim$ 4000 \AA\ is too strong, making the flux there too low. 

921: This line could be reproduced by the \ion{C}{ii} $\lambda$6578 line. This

922: identification is, however, quite marginal. If we assume this feature is the

923: \ion{C}{ii}, it requires X(C) $\sim$ 0.01. This is larger than the upper

924: limit derived in the preceding spectra, which traces the abundances in the 

925: outer layer.

926: 

927: The \ion{O}{i} $\lambda$7774 line is always visible in the spectra

928: from day +14 to + 45.

929: The \ion{O}{i} line is very weak in the earlier spectra,

930: as a result of a severe suppression of \ion{O}{i} fraction

931: due to the high temperature.

932: In fact, the modeling suggests a high O mass fraction

933: in the outermost layers (X(O) $\gsim$ 0.75).

934: These facts clearly indicate that unburned O exists in every place

935: of the ejecta at $v > 1000$ \kms.

936: The mass of O derived from spectral modeling is discussed

937: in \S \ref{sec:splate}.

938: 

939: 

940: \subsubsection{Late phase}

941: \label{sec:splate}

942: 

943:  The late phase spectra of SN 2005hk 

944: consist of a combination of P-Cygni profiles of permitted lines and

945: emission features of forbidden lines (Fig.\ref{fig:spec4}).

946: We note that no strong [\ion{O}{I}] line is seen in the observed spectra.

947: The presence of P-Cygni features suggests that

948: the ejecta are not completely transparent even at such late epochs.

949: 

950: 

951: We study the late phase spectra using the model E03, 

952: which has similar properties to those of 3D deflagration model.

953: Considering the observed spectra,

954: there seems to be two major problems in this model.

955: 

956: 

957: 

958: (1) The ejecta of the E03 model become optically thin at the late phases.

959: The photospheric velocity becomes zero at 250 days after the explosion

960: if a constant line opacity ($0.1$ ${\rm cm^2 g^{-1}}$) is assumed.

961: 

962: (2) In the E03 model, the mass of oxygen is as much as 0.44 $\Msun$.

963: Such a model would give conspicuous emission 

964: lines of [\ion{O}{i}] (\eg Kozma et al.\ 2005).

965: 

966:  First, we create synthetic spectra with the Monte Carlo code

967: used in the modeling of early phase spectra.

968: Here we assume the presence of the region emitting continuum light

969: at the velocity of $v_{\rm ph}  = 250$ \kms.

970: This could be realized if the central part of the ejecta

971: is denser than that of the E03 model. 

972: A synthetic spectrum computed with $\log$ ($L$) (\ergs) $= 40.86$

973: and the element mass fraction same as at day +45

974: (blue line in Fig. \ref{fig:nebmodel})

975: is in reasonable agreement with the observed spectrum at day +228

976: except for emission features at 7300 \AA\ ([\ion{Ca}{ii}]) and

977: 8600 \AA\ (\ion{Ca}{ii}).

978: The temperature at the photosphere is as low as 4400 K.

979: Most of the absorption lines are \ion{Fe}{ii} but

980: some lines of neutral atoms such as \ion{Na}{i} and \ion{Fe}{i}

981: are also found.

982: 

983: An absorption of the \ion{O}{i} $\lambda$7774 line is seen 

984: as discussed in \S 6.2 (also reported in SN 2002cx, Jha et al.\ 2006).

985: In the synthetic spectrum, we could get the \ion{O}{i} line 

986: with the mass fraction X(O) $\sim 0.5$.

987: Therefore, we conclude that the unburned O is present

988: in whole ejecta, down to very low velocities, 

989: suggesting a nearly completely mixed abundance distribution.

990:  This property is consistent with 3D deflagration models.

991: 

992: In modeling the early and late phase spectra, the mass fractions of elements are

993: optimized assuming homogeneous abundance distribution above the

994: photosphere at each epoch. 

995: If we assume these mass fractions are characteristic values at each

996: velocity range of the ejecta, the mass of each element can be computed by

997: integrating the density profile of the E03 model. In Table

998: \ref{tab:param}, the

999: masses of C, O, Si and Fe group elements derived from spectral modeling are

1000: shown. Compared with the masses of the E03 model, the smaller amount of C and

1001: Si, and larger amount of O are preferred by the spectra. The nuclear energy

1002: release corresponding to the element abundances obtained from the spectral

1003: fitting is $\sim 0.8 \times 10^{51}$ erg. 

1004:  Since the binding energy of the WD is $\sim 0.5 \times 10^{51}$erg, 

1005: this leads to a kinetic energy of

1006: $\sim 0.3 \times 10^{51}$ erg, which is consistent with  the E03

1007: model.

1008: 

1009:  In the computation above, the cooling via lines is not considered,

1010: \ie no emission lines would appear.

1011: In order to investigate the emission lines,

1012: we compute the late phase spectra using a nebular synthesis code

1013: and consider the combination with the synthetic photospheric spectrum

1014: (as in Mazzali et al. 2004).

1015: 

1016: \begin{figure}

1017: \plotone{f16.eps}

1018: \caption{Spectrum of SN 2005hk at +228 days with respect

1019: to date of  $B$ maximum (black line, $3 \times 10^{-17}$

1020: erg cm$^{-2}$ s$^{-1}$ \AA$^{-1}$ added to the flux for

1021: presentation), as compared to

1022: the synthetic "combined" spectrum (red line, the same amount added to

1023: the flux). The synthetic spectrum is a combination of

1024: the photospheric (blue line) and the nebular (green line) components.

1025: \label{fig:nebmodel}}

1026: \end{figure}

1027: 

1028: The nebular code computes the deposition from $\gamma$-rays with a gray

1029: opacity (0.025 cm$^{2}$ g$^{-1}$) by Monte Carlo method. 

1030:  Positrons are assumed to be trapped in situ.

1031: To take into account the different channels in which

1032: the deposited energy is consumed, we assume

1033: $L_{\rm neb} = f_{\rm neb} L_{{\gamma}, {e^+}}$

1034: where $L_{{\gamma}, {e^+}}$ is the luminosity deposited by

1035: the $\gamma$-rays and positrons from $0.18 \Msun$ of $^{56}$Co

1036: (originally $^{56}$Ni).

1037: The fraction, $f_{\rm neb}$, of $L_{{\gamma}, {e^+}}$ is

1038: assumed to be balanced by cooling via nebular emission lines,

1039: while the other fraction is by photospheric, allowed transitions

1040: as modeled by the early phase spectrum synthesis code.

1041: In other word, the luminosity (1-$f_{\rm neb}$) $L_{{\gamma}, {e^{+}}}$

1042: is deposited below the photosphere, while the remaining luminosity

1043: ($f_{\rm neb}) L_{{\gamma}, {e^{+}}}$ is deposited into the

1044: optically thin layer above the photosphere.

1045: With this deposition luminosity (with varying $f_{\rm neb}$),

1046: non-LTE rate equations

1047: are solved assuming the ionization-recombination equilibrium

1048: (Mazzali et al.\ 2001).

1049: Note that the code assumes that the ejecta are optically thin.

1050: For the calculation, we assume homogeneous distribution of

1051: elements and $^{56}$Co. The masses of these elements are

1052: set to be consistent with the E03 model, i.e.,

1053: $M$($^{56}$Co) $= 0.18 \Msun$ and $M$(O) $= 0.4 \Msun$.

1054: 

1055: Figure \ref{fig:nebmodel} shows the combination of the photospheric

1056: (as described in the first part of this section, with the luminosity

1057: corresponding to $M$($^{56}$Co) $= 0.27\Msun$

1058: \footnote{This luminosity is higher than that of the E03 model

1059: because the late spectrum taken with Subaru is calibrated against

1060: the photometry that is brighter than the model LC (Fig. \ref{fig:modelLC})

1061: at 250 days after the explosion.})

1062: and nebular components with $f_{\rm neb} = 0.5$.

1063: We find that the [\ion{O}{i}]$\lambda\lambda$6300, 6363 doublet is not

1064: strong in the model spectra. Indeed,

1065: the resulting [\ion{O}{i}] is so weak

1066: that it does not contradict with the non-detection of the [\ion{O}{i}],

1067: if $f_{\rm neb} \lsim 0.5$, as is consistent with

1068: the observed, allowed line-dominated spectrum.

1069: 

1070: This is an outcome of (a) the high density of the ejecta and

1071: of (b) the assumption that the elements, especially O and Fe, are

1072: homogeneously, thus microscopically, mixed together with one another.

1073: With these two conditions, the deposited luminosity is

1074: effectively consumed by copious [\ion{Fe}{ii}] lines

1075: rather than [\ion{O}{i}] \footnote{Indeed, [\ion{Fe}{ii}] around 5200\AA\

1076: is too strong in the model spectrum,

1077: thus smaller value of $f_{\rm neb}$ is favored.}.

1078: The high density is a direct consequence of the

1079: low energy SN Ia model.

1080: However, the complete mixing is a rather extreme assumption,

1081: so that we need more detailed modeling of the late phase spectra of SN 2005hk.

1082: 

1083: With the nebular component,

1084: we obtain better  fits compared to purely  photospheric component,

1085: in the following aspects:

1086: (i) Strong emission features below $\sim 5500$\AA,

1087: (ii) [\ion{Fe}{ii}]$\lambda$7155 and [\ion{Fe}{ii}]$\lambda$7172,

1088: (iii) [\ion{Ca}{ii}]$\lambda\lambda$7291, 7324, and

1089: (iv) \ion{Ca}{ii} IR and [\ion{C}{i}]$\lambda$8727.

1090: 

1091: The synthetic nebular spectrum for

1092: the E03 model results in broader lines than the observed nebular emissions.

1093: Typical example is [\ion{Ca}{ii}]$\lambda\lambda$7291,7324.

1094: The two components are separated in the observations, but

1095: blended in the model. This indicates that the 

1096: nebular line emitting region is indeed more centrally 

1097: concentrated than the model.

1098: This may again suggest the central part being

1099: denser than the E03 model.

1100: 

1101: 

1102: In summary, the major problems (1) and (2) above 

1103: could be solved if the ejecta are denser than 

1104: in the  original model and the elements are microscopically mixed.

1105: Then, the ejecta at late phases consist of three parts: 

1106: the optically thick innermost part 

1107: ($v \lsim 250$ \kms), 

1108: the optically thin region emitting forbidden lines 

1109: ($v \lsim 700$ \kms, \ie the width of the [\ion{Ca}{ii}] lines), 

1110: and the  outer layer which is thin even to $\gamma$-rays. 

1111: 

1112: 

1113: \section{Discussion}

1114: 

1115: We have studied the bolometric LC and early and late phase spectra 

1116: by computing synthetic LC and spectra.

1117: We construct a simple, less energetic model by scaling the kinetic energy 

1118: of W7 down to $0.3 \times 10^{51}$ erg.

1119: This model (E03) explains the LC of SN 2005hk from early to late phase.

1120: The ejecta of the model contain as much as $\sim 0.8 \Msun$

1121: of unburned C+O material.

1122: In this sense, our less energetic model is similar to 

1123: the 3D deflagration model.

1124: 

1125: Branch et al. (2004), Jha et al. (2006) and Phillips et al. (2007)

1126: suggested that the properties of SN 2002cx-like objects are

1127: similar to those of the 3D deflagration model.

1128: We confirm this suggestion by modeling the LC until late phase

1129: for the first time.

1130: We also find that the early phase spectra of SN 2005hk 

1131: are also explained by lower energetic model.

1132: The \ion{O}{i} absorption is always seen until late phase,

1133: suggesting that the elements in the ejecta are mixed almost homogeneously.

1134: This is also consistent with the 3D deflagration model.

1135: 

1136: The late phase spectra of SN 2005hk do not show strong [\ion{O}{i}] line

1137: and do show P-cygni profiles of permitted lines.

1138: They seem inconsistent with what 3D deflagration model predicts, 

1139:  at a first look.

1140: To study this in detail, we modeled the late phase spectra of SN 2005hk 

1141: by a combination of the photospheric and nebular components.

1142: We find that the innermost region of the ejecta should be 

1143: denser and more centrally concentrated 

1144: than that of our model in order to make an optically thick region

1145: and a smaller emitting region of nebular lines.

1146: The absence of the [\ion{O}{i}] line is explained if 

1147: the elements are microscopically mixed.

1148: 

1149: In summary, our less energetic explosion model or 3D deflagration

1150: model seems to be compatible with the late phase spectrum of SN 2005hk 

1151: (and of SN 2002cx-like objects)

1152: (1) if the density of the central region is higher than that of 

1153: our scaled model

1154: and (2) if the elements are microscopically mixed,

1155: although it is a question how such extreme conditions are realized.

1156: 

1157: The nature of this class of objects can be studied further

1158: if observations become available at  phases later than presented here.

1159: SN 2005hk is still not completely transparent at $\sim 400$ days

1160: after the explosion.

1161: Only when the ejecta become completely thin,

1162: useful information on the structure of the innermost ejecta can be obtained.

1163: 

1164: If there is a sub-class of SNe Ia with a lower kinetic energy than in normal SNe Ia,

1165: it could be harmful for cosmological use of SNe Ia.

1166: To show the effect, we construct a model less energetic than 

1167: model E03 by scaling the energy of W7 model to 

1168: $0.08 \times 10^{51}$ erg ($\sim 1/4$ of the kinetic energy of the E03 model). 

1169: If the deflagration is very weak and slow, such a situation will be realized.

1170: Since the nuclear energy release is only slightly higher than 

1171: the binding energy of the WD, such an explosion could be followed after

1172: a pulsation (pulsating deflagration model, \eg Nomoto, Sugimoto \& Neo 1976).

1173: Since the mass of \Nifs\ synthesized by such an explosion 

1174: is thought to be smaller than that of the E03 model,

1175: we set $\Mni = 0.09 \Msun$ ($1/2$ of E03).

1176: We call this model E008, and the parameters are summarized in Table 

1177: \ref{tab:param}.

1178: 

1179: In Figure \ref{fig:modelLC}, the synthetic LC with the E008 model 

1180: is shown as a  magenta line.

1181: The LC has a fainter peak magnitude than that of the E03 model 

1182: because of the smaller amount of \Nifs.

1183: However, the decline rate of the LC after maximum is slower 

1184: than in the E03 model because of the lower kinetic energy of the ejecta.

1185: This trend is opposite to the well-known LC-width relation in SNe Ia,

1186: \ie brighter SNe Ia declines more slowly.

1187: 

1188: Such explosions can be distinguished by the late phase LC.

1189: Since less energy leads to an efficient trapping of $\gamma$-rays

1190: at late phases,

1191: the decline  rate of the LC at  $\gsim 100$ days after the explosion

1192: is slower.

1193: This is clearly seen in the synthetic LC (Figure \ref{fig:modelLC}).

1194: Although the peak magnitude of the E008 model 

1195: is fainter than that of the E03 by $\sim 1$ mag, 

1196: the late phase luminosities of these two models are comparable.

1197: 

1198: 

1199: A recently discovered supernova, SN 2007qd may belong to this sub-class

1200: (SN 2002cx-like; Bassett et al. 2007, Goobar et al. 2007).

1201: This supernova was reported to have reached a fainter maximum than SN 2002cx 

1202: and SN 2005hk and lower  line velocity,

1203: being roughly half that of 2002cx or 2005hk.

1204: Such properties are realized if the kinetic energy of the ejecta

1205: is about quarter  that of SN 2002cx (similar to the E008 model),

1206: because of the relation of $\KE \propto v^2$.

1207: If this is true, the LC of SN 2007qd will evolve very slowly 

1208: as shown in the magenta line in Figure \ref{fig:modelLC}.

1209: 

1210: 

1211: 

1212: \section{Conclusions}

1213: 

1214: Photometric and spectrosopic data on the peculiar SN 2005hk

1215: are presented. The $B$ band light curve of SN 2005hk shows comparable

1216: pre-maximum brightening, 

1217: faster decline in the initial $\sim$20 days past maximum and

1218: slower decline beyond $\sim$50 days past maximum, as compared to normal

1219: Type Ia supernovae. 

1220: The fainter peak bolometric luminosity indicates synthesis of small

1221: amount of \Nifs \ in the explosion. Low expansion velocity of the ejecta

1222: together with fainter peak luminosity is explained by an explosion with lower

1223: kinetic energy. A reasonable fit to the bolometric light curve of SN

1224: 2005hk is

1225: achieved with a less energetic ($0.30 \times 10^{51}$erg) model which

1226: synthesized  $0.18 \Msun$ of \Nifs. 

1227: The light curve evolution is similar to SN 2002cx.

1228: 

1229: The pre-maximum spectrum of SN 2005hk is similar to that of SN 1991T, with

1230: much lower expansion velocity. 

1231: The spectral evolution of SN 2005hk is very similar to SN 2002cx. 

1232: The spectrum of SN 2005hk at late phases ($> 200$ days) is similar to 

1233: SN 2002cx, except

1234: for higher expansion velocities and higher velocity dispersions.

1235: The presence of P-Cygni profiles in the late phase spectrum of SN 2005hk

1236: indicates

1237: that the ejecta have  not become optically thin till our last observation.

1238: Modeling of the pre-maximum spectra of SN 2005hk indicates a relatively

1239: higher temperature which makes \ion{Fe}{iii} lines strong.  

1240: The presence of weak \ion{O}{i} line at $\lambda$7774 

1241: at almost all epochs is modeled as a

1242: consequence of high abundance of completely mixed unburned oxygen in the

1243: ejecta.

1244: The late phase spectra of SN 2005hk are modeled as a combination of the

1245: photospheric and nebular components, with the nebular line emitting region

1246: being more centrally concentrated than is expected in the case of a lower

1247: energetic model.

1248: 

1249: 

1250: \acknowledgements

1251: 

1252: We are thankful to the anonymous referee for valuable comments. This work has 

1253: been carried out under the INSA (Indian National Science Academy) -

1254: JSPS (Japan Society for Promotion of Science) exchange programme.

1255: This research was supported in part by the National Science Foundation

1256: under Grant No. PHY05-51164 and grant-in-Aid for  Scientific Research 

1257: (18104003, 18540231) and the 21st  Century COE Program (QUEST) from 

1258: the JSPS and MEXT of Japan. M.T. and N.T. 

1259: are supported through the JSPS Research Fellowships for Young Scientists.

1260: K.M. is supported through the JSPS Postdoctoral Fellowships 

1261: for Research Abroad.

1262: This work has used data collected at 

1263: Subaru Telescope, which is operated by the National Astronomical 

1264: Observatory of Japan. We thank J. Deng for useful discussions 

1265: and S. Srividya for help in observations and data reduction.

1266:  We thank all the observers of the

1267: 2-m HCT (operated by the Indian Institute of Astrophysics), who kindly

1268: provided part of their observing time for the supernova observations. 

1269: This work has made use of the NASA Astrophysics Data System and the NASA/IPAC

1270: Extragalactic Database (NED) which is operated by Jet Propulsion Laboratory,

1271: California Institute of Technology, under contract with the National

1272: Aeronautics and Space Administration.

1273: 

1274: 

1275: 

1276: 

1277: \begin{thebibliography}{}

1278: 

1279: \bibitem[]{857} Altavilla, G. et al., 2004, MNRAS, 349, 1344

1280: 

1281: \bibitem[]{859} Anupama, G. C., Sahu, D. K., Jose, J.  2005, A\&A, 429, 667

1282: 

1283: \bibitem[]{861} Arnett, W.D., 1982, ApJ, 253, 785

1284: 

1285: \bibitem[]{ber05} Barentine et al., 2005, CBET 268.

1286: 

1287: \bibitem[]{bes07} Bassett, B. et al. 2007, CBET 1137.

1288: 

1289: \bibitem[]{863} Bessell, M.S.,  Castelli, F. \& Plez, B., 1998, A\&A, 333, 231

1290: 

1291: \bibitem[]{bli}  Blinnikov, S.I., et al. 2006, A\&A, 453, 229

1292: 

1293: \bibitem[]{865} Branch, D., Baron, E., Thomas, R. C., Kasen, D., Li, W.,

1294: Filippenko, A. V.  2004, PASP, 116, 903

1295: 

1296: \bibitem[]{868} Branch, D.,  Fisher, A. \& Nugent, P., AJ, 1993, 106, 2383

1297: 

1298: \bibitem[]{870} Burket \& Li 2005, IAUC, 8625

1299: 

1300: \bibitem[]{872} Cardelli, J.A., Clayton, G.C., Mathis, J.S., 1989, ApJ, 345, 245

1301: 

1302: 

1303: \bibitem[]{877} Chornock, R., Filippenko, A. V., Branch, D., Foley, R. J., Jha,

1304: S., Li, W.  2006, PASP, 118, 722

1305: 

1306: \bibitem[]{880}Conley, A., et al. 2006, AJ, 132, 1707

1307: 

1308: \bibitem[]{882} Contardo, G., Leibundgut, B., Vacca, W.D., 2000, A\&A, 359, 876

1309: 

1310: \bibitem[]{884} Filippenko, A. V.,  Richmond, M. W., Branch, D., Gaskell, M.,

1311: Herbst, W., Ford, C. H., Treffers, R.R., Matheson, T.,  Ho, L. C. et al.

1312: 1992, AJ, 104, 1543

1313: 

1314: \bibitem[]{888} Fransson, C. \& Chevalier, R.A.,  1989, ApJ, 343, 323

1315: 

1316: \bibitem[]{890} Goldhaber, G., et al. 2001, ApJ, 558, 359

1317: 

1318: \bibitem[]{891} Goobar, et al., 2007, CBET, 1137, 2 

1319: 

1320: \bibitem[]{892} Hamuy, M., Phillips, M.M., Suntzeff, N.B., Schommer, R.A.,

1321: Maza, J., Avils, R.  1996a, AJ, 112, 2391

1322: 

1323: \bibitem[]{895} Hamuy, M., Phillips, M.M., Suntzeff, N.B., Schommer, R.A.,

1324: Maza, J., Smith, R.C., Lira, P., Avils, R.  1996b, AJ, 112, 2438

1325: 

1326: 

1327: \bibitem[]{iwa00} Iwamoto, K. 1997, PhD thsis, University of Tokyo

1328: 

1329: \bibitem[]{901} Iwamoto, K., et al. 2000, ApJ, 534, 660

1330: 

1331: \bibitem[]{903} Jha, S.,  Branch, D.,  Chornock, R., Foley, R. J., Li, W.,

1332: Swift, B. J.,  Casebeer, D.,  Filippenko, A. V.,  AJ, 2006, 132, 189

1333: 

1334: \bibitem[]{906} Kashikawa, N., Aoki, K.,  Asai, R., Ebizuka, N.,  Inata, M.,

1335: Iye, M.,  Kawabata, K. S., Kosugi, G., Ohyama, Y.,  Okita, K.,  2002,

1336: PASJ, 54, 819

1337: 

1338: \bibitem[]{910} Kozma, C., et al. 2005, A\&A, 437, 983

1339: 

1340: \bibitem[]{911}Kuchner, M.J., Kirshner, R.P., Pinto, P.A., Leibundgut, B.,  1994, 426, L89

1341: 

1342: \bibitem[]{912} Landolt, A.U.,  1992, AJ, 104, 340

1343: 

1344: \bibitem[]{916} Li, W., Filippenko, A. V.; Treffers, R. R.; Riess, A. G.; Hu,

1345: J.; Qiu, Y. , ApJ, 2001, 546, 734

1346: 

1347: \bibitem[]{919} Li, W., Filippenko, A. V., Chornock, R., Berger, E.,  Berlind,

1348: P.,  Calkins, M. L., Challis, P., Fassnacht, C. Jha, S.,  Kirshner, R.P.

1349: PASP, 2003, 115, 453

1350: 

1351: \bibitem[]{923} Lira, P. 1996, Master's thesis, Univ. Chile

1352: 

1353: \bibitem[]{925} Lira, P., Hamuy, M., Wells, L.A., et al.\ 1998, AJ, 115, 234

1354: 

1355: \bibitem[]{927} Lucy, L.B. 1999, A\&A, 345, 211

1356: 

1357: 

1358: \bibitem[]{931} Maeda, K., Mazzali, P. A., Deng, J., Nomoto, K., Yoshii, Y.,

1359: Tomita, H., \& Kobayashi, Y. 2003, ApJ, 593, 931

1360: 

1361: \bibitem[]{934} Mazzali, P.A. 2000, A\&A, 363, 705

1362: 

1363: \bibitem[]{936} Mazzali, P.A. \& Lucy, L.B. 1993, A\&A, 279, 447

1364: 

1365: \bibitem[]{938} Mazzali, P. A., Danziger, I. J., Turatto, M. 1995, A\&A, 297, 509

1366: 

1367: \bibitem[]{940} Mazzali, P.A., Nomoto, K., Maeda, K., \& Patat, F, 2001, ApJ,

1368: 559, 1047

1369: 

1370: \bibitem[]{943} Mazzali, P. A., Deng, J., Maeda, K., Nomoto, K., Filippenko, A.

1371: V., \& Matheson, T. 2004, ApJ, 614, 858

1372: 

1373: \bibitem[]{946} Mazzali, P.A., et al. 2005, ApJ, 623L, 37

1374: 

1375: \bibitem[]{nom76} Nomoto, K., Sugimoto, D., \& Neo, S. 1976, Ap\&SS, 39, L37

1376: 

1377: \bibitem[]{948} Nomoto, K., Thielemann, F.-K., \& Yokoi, K. 1984, ApJ, 286, 644

1378: 

1379: \bibitem[]{952} Phillips, M.M, Wells, L.A., Suntzeff, N.B., et al.\ 1992, AJ,

1380: 103, 1632

1381: 

1382: \bibitem[]{955} Phillips, M.M., Lira, P., Suntzeff, N.B., Schommer, R.A.,

1383: Hamuy, M., Jose, M. 1999, AJ, 118, 1766

1384: 

1385: \bibitem[]{958} Phillips, M. M., Li, W.,  Frieman, J. A., Blinnikov, S. I.,

1386: DePoy, D., Prieto, J. L.,  Milne, P., Contreras, C., Folatelli, G.;

1387: Morrell, N.  et al. 2007, PASP, 119, 360

1388: 

1389: \bibitem[]{962} Richmond, M.W., Treffers, R.R., Filippenko, A.V. et al, 1995,

1390: AJ, 109, 2121

1391: 

1392: \bibitem[]{Roe06} R\"opke, F.K., Gieseler, M., Reinecke, M., Travaglio, C., \&

1393: Hillebrandt., W. 2006, A\&A, 453, 203

1394: 

1395: \bibitem[]{965} Schelgel, D.J., Finkbeiner, D.P., Davis, M., 1998, ApJ, 500, 525

1396: 

1397: 

1398: \bibitem[]{969} Stanishev, V., Taubenberger, S., Blanc, G., Anupama, G. C.,

1399: Benetti, S., Cappellaro, E., Elias-Rosa, N., Feron, C., Goobar, A.,

1400: Krisciunas, K.  et al. 2006, astroph/0611354

1401: 

1402: \bibitem[]{976} Suntzeff, N.B.: 1996, in R. McCray and Z. Wang (eds.),

1403: Supernovae and Supernova Remnants, p. 41, Cambridge Univ. Press,Cambridge

1404: 

1405: \bibitem[]{979} Tanaka, M., Mazzali, P. A., Maeda, K., \& Nomoto, K. 2006, ApJ,

1406: 645, 470

1407: 

1408: \bibitem[]{982} Tanaka, M., et al. 2007, ApJ, in press (arxiv:0712.2823)

1409: 

1410: \bibitem[]{984} Turatto, M., Benetti, S., Cappellaro, E., Danziger, I. J.,

1411: della Valle, M., Gouiffes, C., Mazzali, P. A., Patat, F. 1996, MNRAS, 283,

1412: 1.

1413: 

1414: \end{thebibliography}

1415: 

1416: %\clearpage

1417: 

1418: 

1419: \begin{deluxetable}{lccccc}

1420: \tablewidth{0pt}

1421: \tablecaption{Magnitudes for the sequence of secondary standard stars in

1422: the field of SN 2005hk.\tablenotemark{*}}

1423: \tablehead{

1424: ID & U  & B & V &  R & I

1425: }

1426: \startdata

1427: 1 &  $15.896\pm0.020$ &  $15.425\pm0.006$ & $14.643\pm0.004$ &

1428: $14.174\pm0.012$ & $13.777\pm0.005$\\

1429: 2 &  $17.257\pm0.025$ &  $16.859\pm0.007$ & $16.067\pm0.005$ &

1430: $15.579\pm0.010$ & $15.149\pm0.010$\\

1431: 3 &  $19.175\pm0.035$ &  $17.696\pm0.002$ & $16.375\pm0.009$ &

1432: $15.564\pm0.007$ & $14.886\pm0.010$\\

1433: 4 &  $16.824\pm0.025$ &  $16.954\pm0.010$ & $16.449\pm0.016$ &

1434: $16.097\pm0.016$ & $15.750\pm0.009$\\

1435: 5 &  $17.842\pm0.027$ &  $17.645\pm0.010$ & $16.902\pm0.009$ &

1436: $16.440\pm0.018$ & $16.003\pm0.008$\\

1437: 6 &  $17.094\pm0.022$ &  $16.870\pm0.016$ & $16.150\pm0.010$ &

1438: $15.706\pm0.017$ & $15.275\pm0.005$\\

1439: 7 &  $17.834\pm0.032$ &  $17.213\pm0.011$ & $16.210\pm0.012$ &

1440: $15.646\pm0.022$ & $15.108\pm0.004$\\

1441: 8 &  $19.134\pm0.040$ &  $18.194\pm0.020$ & $17.175\pm0.014$ &

1442: $16.571\pm0.013$ & $16.055\pm0.004$\\

1443: 9 &  $16.366\pm0.024$ &  $16.470\pm0.007$ & $16.115\pm0.007$ &

1444: $15.871\pm0.012$ & $15.623\pm0.002$\\

1445: 10&  $19.516\pm0.042$ &  $18.162\pm0.020$ & $16.944\pm0.010$ &

1446: $16.201\pm0.021$ & $15.553\pm0.001$\\

1447: 11&                   &  $18.059\pm0.008$ &                  &

1448: $16.782\pm0.016$ &\\

1449: \enddata

1450: \tablenotetext{*}{The stars are identified in Figure. \ref{fig:field}}

1451: \label{tab:calmag}

1452: \end{deluxetable}

1453: 

1454: %\clearpage

1455: 

1456: \begin{deluxetable}{lccccccc}

1457: %\rotate

1458: \tablewidth{0pt}

1459: \tablecaption{Photometric observations of SN 2005hk}

1460: \tablehead{

1461:  Date & J.D. & Phase\tablenotemark{*} & U & B & V & R & I\\

1462:      & 2453000+ & (days)    &   &   &   &   &

1463: }

1464: \startdata

1465: 05/11/2005& 680.06&  -5.22&  $              $ &$16.126\pm0.014$

1466: &$16.167\pm0.018$& $ 16.023\pm 0.014$&  $16.060\pm 0.032$\\

1467: 06/11/2005& 681.04&  -4.25&  $15.482\pm0.046$ &$16.066\pm0.012$

1468: &$16.047\pm0.011$& $ 15.934\pm 0.021$&  $15.979\pm 0.016$\\

1469: 07/11/2005& 682.03&  -3.27&  $15.181\pm0.086$ &$16.004\pm0.046$

1470: &$16.007\pm0.027$& $ 15.862\pm 0.016$&  $15.866\pm 0.039$\\

1471: 08/11/2005& 683.06&  -2.25&  $15.515\pm0.060$ &$15.966\pm0.018$

1472: &$15.900\pm0.018$& $ 15.784\pm 0.013$&  $15.798\pm 0.024$\\

1473: 13/11/2005& 688.04&   2.66&  $15.741\pm0.119$ &$16.022\pm0.069$

1474: &$15.739\pm0.050$& $ 15.631\pm 0.034$&  $15.577\pm 0.051$\\

1475: 15/11/2005& 690.13&   4.73&  $15.956\pm0.066$ &$16.208\pm0.027$

1476: &$15.710\pm0.024$& $ 15.564\pm 0.017$&  $15.523\pm 0.016$\\

1477: 16/11/2005& 691.20&   5.79&  $16.030\pm0.058$ &$16.260\pm0.021$

1478: &$15.763\pm0.023$& $ 15.563\pm 0.016$&  $15.416\pm 0.018$\\

1479: 17/11/2005& 692.14&   6.72&  $16.222\pm0.062$ &$16.413\pm0.020$

1480: &$15.779\pm0.020$& $ 15.568\pm 0.019$&  $15.466\pm 0.027$\\

1481: 19/11/2005& 694.13&   8.69&  $16.572\pm0.116$ &$16.671\pm0.021$

1482: &$15.840\pm0.013$& $ 15.553\pm 0.011$&  $15.427\pm 0.025$\\

1483: 22/11/2005& 697.06&  11.58&  $17.212\pm0.037$ &$17.132\pm0.012$

1484: &$16.019\pm0.008$& $ 15.649\pm 0.015$&  $15.466\pm 0.015$\\

1485: 01/12/2005& 706.06&  20.48&  $18.620\pm0.021$ &$18.260\pm0.021$

1486: &$16.704\pm0.009$& $ 16.126\pm 0.015$&  $15.796\pm 0.019$\\

1487: 05/12/2005& 710.03&  24.40&  $              $ &$18.494\pm0.011$

1488: &$16.925\pm0.011$& $ 16.364\pm 0.013$&  $15.970\pm 0.019$\\

1489: 08/12/2005& 713.02&  27.36&  $              $ &$18.600\pm0.019$

1490: &$17.076\pm0.010$& $ 16.501\pm 0.017$&  $16.106\pm 0.009$\\

1491: 12/12/2005& 717.03&  31.32&  $              $ &$18.758\pm0.033$

1492: &$17.237\pm0.024$& $ 16.646\pm 0.025$&  $16.251\pm 0.018$\\

1493: 17/12/2005& 722.06&  36.29&  $              $ &$18.832\pm0.016$

1494: &$17.362\pm0.012$& $ 16.826\pm 0.013$&                   \\

1495: 23/12/2005& 728.13&  42.29&  $              $ &$18.846\pm0.096$

1496: &$17.607\pm0.051$& $ 17.053\pm 0.041$&  $16.593\pm 0.044$\\

1497: %26/12/2005\tablenotemark{\dagger} & 730.79  &  44.92  &  $              $

1498: %&$18.859\pm0.010$ &                & $ 17.150\pm 0.017$&\\

1499: 26/12/2005 & 730.79  &  44.92  &  $              $

1500: &$18.859\pm0.010$ &                & $ 17.150\pm 0.017$&\\

1501: 26/12/2005& 731.14&  45.27&  $              $ &$18.944\pm0.024$

1502: &$17.605\pm0.010$& $ 17.101\pm 0.009$&  $16.698\pm 0.023$\\

1503: 28/12/2005& 733.07&  47.17&  $              $ &$19.017\pm0.014$

1504: &$17.618\pm0.013$& $ 17.158\pm 0.008$&  $16.749\pm 0.008$\\

1505: 12/01/2006& 748.06&  61.99&  $              $ &$19.204\pm0.031$

1506: &$17.920\pm0.014$& $ 17.501\pm 0.011$&  $17.142\pm 0.027$\\

1507: 14/01/2006& 750.05&  63.96&  $              $ &$19.325\pm0.045$

1508: &$18.038\pm0.035$& $ 17.548\pm 0.027$&  $17.203\pm 0.015$\\

1509: %30/06/2006\tablenotemark{\dagger} & 917.04 & 230.03&  $              $

1510: %&$21.578\pm0.033$ &$              $& $ 20.140\pm 0.022$& $          $  \\

1511: 30/06/2006 & 917.04 & 230.03&  $              $

1512: &$21.578\pm0.033$ &$              $& $ 20.140\pm 0.022$& $          $  \\

1513: 04/07/2006& 921.40& 233.31&  $              $ &$              $

1514: &$20.940\pm0.044$& $               $&  $               $\\

1515: 18/07/2006& 935.40& 247.14&  $              $ &$              $ &$

1516:      $& $ 20.474\pm0.036$&  $               $\\

1517: 20/09/2006& 999.39& 310.38&  $              $ &$              $

1518: &$22.733\pm0.168$& $ 21.633\pm0.099$&  $               $\\

1519: %27/11/2006\tablenotemark{\dagger}&1066.73& 376.98&  $              $

1520: %&$24.206\pm0.124$ &                 & $ 21.776\pm0.036$&  $$\\

1521: 27/11/2006&1066.73& 376.98&  $              $

1522: &$24.206\pm0.124$ &                 & $ 21.776\pm0.036$&  $$\\

1523: \enddata

1524: \tablenotetext{\rlap{*}}{{ \,Observed phase with respect to the epoch of

1525: maximum in $B$ band (JD 2453685.34), corrected for the (1+z) time-dilation

1526: factor using the host galaxy redshift z = 0.0118}}

1527: %\tablenotetext{\dagger}{Subaru photometric observations}

1528: \label{tab:mag}

1529: \end{deluxetable}

1530: 

1531: 

1532: %\clearpage

1533: 

1534: \begin{deluxetable}{lcrcl}

1535: \tablewidth{0pt}

1536: \tablecaption{Log of spectroscopic observations of SN 2005hk}

1537: \tablehead{

1538: Date & J.D. & \multicolumn{1}{c}{Phase\rlap{*}} & Range & Telescope\\

1539:      & 2453000+ & \multicolumn{1}{c}{(days)} & \AA\ &

1540: }

1541: \startdata

1542: 04/11/05 & 679.10 & $ -6.17$   & 3500-7000; 5200-9100 & HCT\\

1543: 05/11/05 & 680.06 & $ -5.22$   & 3500-7000; 5200-9100 & HCT\\

1544: 06/11/05 & 681.06 & $ -4.23$   & 3500-7000; 5200-9100 & HCT\\

1545: 13/11/05 & 688.06 & $  2.69$   & 3500-7000; 5200-9100 & HCT\\

1546: 15/11/05 & 690.25 & $  4.85$   & 3500-7000; 5200-9100 & HCT\\

1547: 17/11/05 & 692.10 & $  6.68$   & 3500-7000; 5200-9100 & HCT\\

1548: 20/11/05 & 695.27 & $  9.91$   & 3500-7000; 5200-9100 & HCT\\

1549: 23/11/05 & 698.19 & $ 12.70$   & 3500-7000; 5200-9100 & HCT\\

1550: 24/11/05 & 699.18 & $ 13.68$   & 3500-7000; 5200-9100 & HCT\\

1551: 01/12/05 & 706.09 & $ 20.50$   & 3500-7000; 5200-9100 & HCT\\

1552: 05/12/05 & 710.08 & $ 24.45$   & 3500-7000; 5200-9100 & HCT\\

1553: 13/12/05 & 718.02 & $ 32.30$   & 3500-7000; 5200-9100 & HCT\\

1554: 26/12/05 & 730.81 & $ 44.94$    & 4000-9000 & Subaru\\

1555: 12/01/06 & 748.09 & $ 62.02$   & 3500-7000; 5200-9100 & HCT\\

1556: 30/06/06 & 916.5  & $ 228.46$  & 4700-9000 & Subaru\\

1557: 27/11/06 & 1067.05 & $ 377.26$  & 4700-9000 & Subaru\\

1558: \enddata

1559: \tablenotetext{*}{{ \,Observed phase with respect to the epoch of

1560: maximum in $B$ band (JD 2453685.34), corrected for the (1+z) time-dilation

1561: factor using the host galaxy redshift z = 0.0118}}

1562: \label{tab:spec}

1563: \end{deluxetable}

1564: 

1565: 

1566: %\clearpage

1567: 

1568: \begin{deluxetable}{lrrrr}

1569: \tablecaption{Photometric parameters for SN 2005hk}

1570: \tablehead{

1571: Data  &  $B$  &  $V$ &  $R$ & $I$

1572: }

1573: \startdata

1574: epoch of max\tablenotemark{*}  & $685.34\pm 0.4$ & $689.49\pm 0.8$ &

1575: $691.78\pm0.2$ &$693.93\pm 0.3$ \\

1576: magnitude at max   & $15.91\pm 0.03$ & $15.71\pm 0.04$ &

1577: $15.55\pm 0.02$ & $15.43\pm 0.03$\\

1578: 

1579: $\Delta m_{15}(B)$  & $1.68\pm 0.05$ & \\

1580: \\

1581: colors at B max\tablenotemark{**} & & $B-V$ & $V-R$ & $R-I$\\

1582:                  & & $-0.03\pm 0.04$ & $0.08\pm 0.03$ &

1583: $-0.02\pm 0.06$\\

1584: \\

1585: decline rate days 20-45 (mag day$^{-1}$) & 0.021 & 0.027 & 0.031 & 0.031\\

1586: decline rate days 230-380 (mag day$^{-1}$) & & 0.015 & 0.011 &  \\

1587: 

1588: \enddata

1589: \tablenotetext{*}{JD 2453000+}

1590: \tablenotetext{**}{colors are corrected for reddening

1591: $E(B-V)_{total}=0.11$}

1592: \label{tab:lcparam}

1593: \end{deluxetable}

1594: 

1595: 

1596: 

1597: %\clearpage

1598: 

1599: 

1600: \begin{deluxetable}{lcccccc}

1601: \tablewidth{0pt}

1602: \tablecaption{Characteristic Quantities of the Models}

1603: \tablehead{

1604: Model &

1605: $\KE$ \tablenotemark{a} &

1606: $\Mni$ \tablenotemark{b} &

1607: $M ({\rm C})$ \tablenotemark{c} &

1608: $M ({\rm O})$ \tablenotemark{d} &

1609: $M ({\rm Si})$ \tablenotemark{e} &

1610: $M ({\rm Fe\ group})$ \tablenotemark{f}

1611: }

1612: \startdata

1613: W7 (92A)   & 1.2  & 0.17  & 0.038  & 0.12  & 0.16 & 0.95\\

1614: W7 (91bg)  & 1.2  & 0.088 & 0.038  & 0.12  & 0.16 & 0.95\\

1615: E03        & 0.30 & 0.18  & 0.36  & 0.44  & 0.074 & 0.45 \\ 

1616: E008       & 0.08 & 0.09  & 0.44  & 0.53  & 0.053 & 0.32 \\ \hline

1617: Spectra    & --   & --    & $<$ 0.011  & 0.84  & 0.023 & 0.44 \\

1618: \enddata

1619: \tablenotetext{a}{Kinetic energy of the explosion ($10^{51}$ \ergs) }

1620: \tablenotetext{b}{Total mass of \Nifs\ ($\Msun$)}

1621: \tablenotetext{c}{Mass of C contained in the ejecta ($\Msun$)}

1622: \tablenotetext{d}{Mass of O contained in the ejecta ($\Msun$)}

1623: \tablenotetext{e}{Mass of Si contained in the ejecta ($\Msun$)}

1624: \tablenotetext{f}{Total mass of Fe group elements including \Nifs\ ($\Msun$)}

1625: \label{tab:param}

1626: \end{deluxetable}

1627: 

1628: %\clearpage

1629: 

1630: 

1631: 

1632: \begin{deluxetable}{ccccccccccc}

1633: \tablewidth{0pt}

1634: \tablecaption{Parameters of Spectral fitting}

1635: \tablehead{

1636: Epoch \tablenotemark{a} &

1637: $\vph$ \tablenotemark{b} &

1638: $M_{r, {\rm ph}}$ \tablenotemark{c} &

1639: $\log (L)$ \tablenotemark{d} &

1640: $X ({\rm C})$ &

1641: $X ({\rm O})$ &

1642: $X ({\rm Si})$ &

1643: $X ({\rm S})$  &

1644: $X ({\rm Ca})$ &

1645: $X ({\rm Fe})$ \tablenotemark{e}&

1646: $X ({\rm ^{56}Ni})$\tablenotemark{f}

1647: }

1648: \startdata

1649: -6  &  6500 & 1.11 & 42.56 & $<$ 0.001 & 0.86 & 0.002 & 0.021 & 0.001  & 0.020 &

1650: 0.035  \\

1651: +3  &  6000 & 1.05 & 42.66 & $<$ 0.001 & 0.77 & 0.020 & 0.002 & 0.001  & 0.020 &

1652: 0.14  \\

1653: +14 &  3500 & 0.583 & 42.44 & 0.010 (?) & 0.55 & 0.020 & 0.002 & 0.004  & 0.010 &

1654: 0.39  \\

1655: +24 &  1500 & 0.128 & 42.19 & 0.010 (?) & 0.53 & 0.020 & 0.022 & 0.004  &

1656: \multicolumn{2}{c}{0.40\tablenotemark{ g}}  \\

1657: +45 &  1000 & 4.90 $\times 10^{-2}$ & 41.88 & -- & 0.53 & 0.020 & 0.022 & 0.004  &

1658: \multicolumn{2}{c}{0.40\tablenotemark{ g}}  \\

1659: +228 & 250 & 1.18$\times 10^{-3}$  & 040.86 & -- & 0.53  & 0.020 & 0.022 & 0.004  &

1660: \multicolumn{2}{c}{0.40\tablenotemark{ g}}  \\

1661: \enddata

1662: \tablenotetext{a}{Days from $B$ maximum }

1663: \tablenotetext{b}{Photospheric velocity (\kms)}

1664: \tablenotetext{c}{Mass coordinate at the photosphere ($\Msun$)}

1665: \tablenotetext{d}{Bolometric luminosity (\ergs) in logarithm}

1666: \tablenotetext{e}{Mass fraction of stable Fe}

1667: \tablenotetext{f}{Mass fraction of radioactive \Nifs}

1668: \tablenotetext{g}{The sum of stable Fe and $^{56}$Ni}

1669: \label{tab:spparam}

1670: \end{deluxetable}

1671: 

1672: 

1673: 

1674: \end{document}

1675: 

1676: 

1677: 

1678: 

1679: 

1680: 

1681: 

1682: 

1683: \clearpage

1684: 

1685: \begin{figure}

1686: \epsscale{1.0}

1687: \plotone{f1.eps}

1688: \caption{Identification chart for SN 2005hk. The stars used as local

1689: standards are marked as numbers 1-11.

1690: \label{fig:field} }

1691: \end{figure}

1692: 

1693: \clearpage

1694: 

1695: 

1696: \begin{figure}

1697: \epsscale{1.0}

1698: \plotone{f2.eps}

1699: \caption{The $UBVRI$ \  LCs of SN 2005hk. The LCs have been shifted by the

1700: amount indicated in the legend.

1701: \label{fig:lc} }

1702: \end{figure}

1703: 

1704: \clearpage

1705: 

1706: 

1707: \begin{figure}

1708: \epsscale{1.0}

1709: \plotone{f3.eps}

1710: \caption {Early phase ($< 50$~days) LC of SN 2005hk together with those of 

1711: SNe 2002cx, 1991bg, 1991T,  2003du and  1994D. The ordinate in each panel is the

1712: magnitude below the respective maximum, and the abscissae represent the days

1713: since respective $B$ maximum. For clarity, the LCs have been shifted by the amount

1714: indicated in the legend.

1715: \label{fig:compu} }

1716: \end{figure}

1717: 

1718: \clearpage

1719: 

1720: \begin{figure}

1721: \epsscale{1.0}

1722: \plotone{f4.eps}

1723: \caption{Late phase ($>50$~days) LC of SN 2005hk together with those of 

1724: SNe 2002cx, 1991bg, 1991T, 2003du and  1994D. The ordinate in each panel is the

1725: magnitude below the respective maximum, and the abscissae represent the days

1726: since respective $B$ maximum. For clarity, the LCs have been shifted by the amount

1727: indicated in the legend.

1728: \label{fig:compb}}

1729: \end{figure}

1730: 

1731: 

1732: \begin{figure}

1733: \epsscale{1.0}

1734: \plotone{f5.eps}

1735: \caption{The $U-B$, $B-V$, $V-R$ and $V-I$ color curves of SN 2005hk. Also

1736: shown for comparison are the color curves of SNe 2002cx, 1991T, 1991bg, 2003du and

1737: 1994D. The abscissae correspond to days since respective $B$ maximum.

1738: \label{fig:color}}

1739: \end{figure}

1740: 

1741: \clearpage

1742: 

1743: \begin{figure}

1744: \epsscale{1.0}

1745: \plotone{f6.eps}

1746: \caption{

1747:  Bolometric LC of SN 2005hk (red circles) compared with a synthetic LC

1748: with the E03 model (red line).

1749: Black squares and blue triangles are bolometric LCs of

1750: SNe 1992A and 1991bg (Suntzeff 1996), respectively.

1751: Model LCs for these two SNe are shown in black and blue lines, respectively

1752: (based on W7 but different mass of \Nifs, see Table \ref{tab:param}).

1753: Green triangles are the bolometric LC of SN 2002cx.

1754: In order to fix the rise time of the LCs of SNe 1992A and 1991bg,

1755: $B$ band rise time is assumed as $t_{rise} = 19.5 \times s$ days

1756: (Conley et al. 2006),

1757: where $s$ is stretch factor of $B$ band LC (Goldhaber et al. 2001).

1758: The $B$ band rise times of SNe 1992A ($s = 0.80$) and 1991bg ($s = 0.68$)

1759: are estimated as 15.6 and 13.2 days, respectively.

1760: Since the LC of SN 2005hk is deviated from the normal sequence of SNe Ia,

1761: we simply assume  $t_{rise} = 17$ days as in Phillips et al. (2007).

1762:  The magenta line shows the synthetic LC with the E008 model.

1763: \label{fig:modelLC}}

1764: \end{figure}

1765: 

1766: 

1767: \clearpage

1768: 

1769: \begin{figure}

1770: \epsscale{1.0}

1771: \plotone{f7.eps}

1772: \caption{

1773: The density structure of model E03 compared with that of W7

1774: (1 day after the explosion).

1775: The kinetic energy of model E03 is $\KE = 0.3 \times 10^{51} {\rm ergs}$.

1776: The density structure is scaled from that of W7 homologously.

1777: This makes the density at the inner part higher than the original.

1778: \label{fig:dens}}

1779: \end{figure}

1780: 

1781: 

1782: \clearpage

1783: 

1784: \begin{figure}

1785: \epsscale{1.0}

1786: \plotone{f8.eps}

1787: \caption{Pre-maximum spectral evolution  of SN 2005hk. The phases marked

1788: are relative to date of $B$ maximum. The spectra have  been corrected for

1789: the redshift of the host galaxy, but not corrected for the reddening. For

1790: clarity the spectra have been shifted vertically. The telluric lines have 

1791: not been

1792: removed from the spectra and are marked with the symbol {\small $\oplus$}.

1793: \label{fig:spec1}}

1794: \end{figure}

1795: 

1796: \clearpage

1797: 

1798: \begin{figure}

1799: \epsscale{1.0}

1800: \plotone{f9.eps}

1801: \caption{Pre-maximum spectrum of SN 2005hk compared with those of 

1802: Type Ia SNe 1991T and  2003du at similar epoch.  The phases marked are

1803: relative to date of $B$ maximum. The spectra have  been corrected for

1804: the redshift of the host galaxy, but not corrected for the reddening.

1805: For clarity the spectra have been shifted vertically. The telluric lines 

1806: have not been removed from the spectra of SNe 2005hk and  1991T, they are 

1807: marked with  the symbol  {\small $\oplus$}.

1808: \label{fig:comp_photo_03duu1}}

1809: \end{figure}

1810: 

1811: \clearpage

1812: 

1813: \begin{figure}

1814: \epsscale{1.0}

1815: \plotone{f10.eps}

1816: \caption{Spectral evolution  of SN 2005hk from phase $+3$ days to

1817: $+14$ days relative to date of $B$ maximum. The spectra have  been

1818: corrected for the redshift of the host galaxy, but not corrected for

1819: the reddening. For clarity the spectra have been shifted vertically.

1820: The telluric lines 

1821: have not been removed from the spectra  and are marked with 

1822: the symbol {\small $\oplus$}. Line identification is based on

1823: Branch et al. (2004)

1824: \label{fig:spec2}}

1825: \end{figure}

1826: 

1827: \clearpage

1828: 

1829: 

1830: \begin{figure}

1831: \epsscale{1.0}

1832: \plotone{f11.eps}

1833: \caption{Post-maximum spectrum   of SN 2005hk compared with those

1834: of Type Ia SNe 1991T and  2003du at similar epochs.  The phases

1835: marked are relative to date of $B$ maximum. The spectra have  been

1836: corrected for the redshift of the host galaxy, but not corrected for

1837: the reddening. For clarity the spectra have been shifted vertically.

1838: The telluric lines  have not been removed from these spectra 

1839: and are marked with the symbol {\small $\oplus$}.

1840: \label{fig:comp_photo_03duu2}}

1841: \end{figure}

1842: 

1843: \clearpage

1844: 

1845: 

1846: \begin{figure}

1847: \epsscale{1.0}

1848: \plotone{f12.eps}

1849: \caption{Spectral evolution  of SN 2005hk from phase $+21$ days to $+62$ days

1850: relative to date of $B$ maximum. The spectra have been corrected for the

1851: redshift of the host galaxy, but not corrected for the reddening. For clarity

1852: the spectra have been shifted vertically.  The telluric lines 

1853: have not been removed from the spectra of SN 2005hk taken with HCT and are 

1854: marked with the symbol {\small $\oplus$}. The line identification is based on

1855: Branch et al. (2004).

1856: \label{fig:spec3}}

1857: \end{figure}

1858: 

1859: \clearpage

1860: 

1861: \begin{figure}

1862: \epsscale{1.0}

1863: \plotone{f13.eps}

1864: \caption{Late phase spectrum of SN 2005hk at $+228$ days and $+377$ days

1865: relative to date of $B$ maximum. The spectra have  been corrected for

1866: the redshift of the host galaxy, but not corrected for the reddening.

1867: For clarity the spectra have been shifted vertically. The telluric 

1868: lines have been removed from the spectrum.

1869: The line identification is based on Jha et al. (2006).

1870: \label{fig:spec4}}

1871: \end{figure}

1872: 

1873: \clearpage

1874: 

1875: \begin{figure}

1876: \epsscale{1.0}

1877: \plotone{f14.eps}

1878: \caption{Late phase spectrum of SN 2005hk at $+228$ days and $+377$ days

1879: relative to date of $B$ maximum compared with those of other Type Ia SNe 2003du, 2002er, 1998bu, 1991T and 1991bg.

1880: The spectra have  been corrected for the redshift of the host galaxy, but

1881: not corrected for the reddening. For clarity the spectra have been shifted

1882: vertically.

1883: \label{fig:comp_nebular}}

1884: \end{figure}

1885: 

1886: \clearpage

1887: 

1888: 

1889: \begin{figure}

1890: \epsscale{1.0}

1891: \plotone{f15.eps}

1892: \figcaption{

1893: Comparison between the observed early phase spectra of SN 2005hk at

1894: ($-6$, $+3$, $+14$, $+24$ and $+45$ days from $B$ maximum,

1895: from top to bottom) and the synthetic spectra.

1896: Lines that make a major contribution to the absorptions are marked.

1897: \label{fig:early}}

1898: \end{figure}

1899: 

1900: \clearpage

1901: 

1902: 

1903: \begin{figure}

1904: \plotone{f16.eps}

1905: \caption{Spectrum of SN 2005hk at +228 days with respect

1906: to date of  $B$ maximum (black line, $3 \times 10^{-17}$

1907: erg cm$^{-2}$ s$^{-1}$ \AA$^{-1}$ added to the flux for

1908: presentation), as compared to

1909: the synthetic "combined" spectrum (red line, the same amount added to

1910: the flux). The synthetic spectrum is a combination of

1911: the photospheric (blue line) and the nebular (green line) components.

1912: \label{fig:nebmodel}}

1913: \end{figure}

1914: 

1915: 

1916: \end{document}

1917: 

1918: 

1919: 

1920: 

1921: 

1922: 

1923: