1: %this is 2-column

2: %\documentclass[preprint2]{aastex}

3: 

4: %this is suitable for submission to ApJ fast track

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

6: 

7: %\documentclass[apjl]{emulateapj}

8: %\documentclass[apj]{emulateapj}

9: 

10: %this is one column, double-spaced

11: %\documentclass[manuscript]{aastex}

12: 

13: \usepackage{graphicx}

14: 

15: \usepackage{psfig}

16: 

17: %***************** BEGIN  Paco's  definitions

18: \newcommand{\vinf}{\mbox{$v_{\infty}$}}

19: \newcommand{\Msunyr}{\mbox{$M_{\odot}~{\rm yr^{-1}}$}}

20: \newcommand{\hoch}[2]{\mbox{$#1\cdot 10^{#2}$}}

21: \newcommand{\my}{$\mu$m}

22: \newcommand{\ha}{\frac{1}{2}}

23: \newcommand{\myvs}{$\mu$V/s}

24: \def\arcsec{\hbox{\rm \char '175}}

25: \def\opahei{\hbox{$\chi_{504}$}}

26: \def\pfga{\hbox{\rm Pf${\gamma}$}}

27: \def\bgam{\hbox{\rm B${\gamma}$}}

28: \def\hap{\hbox{\rm H${\alpha}$}}

29: \def\hab{\hbox{\rm H${\beta}$}}

30: \def\hag{\hbox{\rm H${\gamma}$}}

31: \def\had{\hbox{\rm H${\delta}$}}

32: \def\bap{\hbox{\rm B${\alpha}$}}

33: \def\pap{\hbox{\rm P${\alpha}$}}

34: \def\heone{\hbox{\rm \HeI\ 584~\AA}}

35: \def\hedubt{\hbox{\rm \HeI\ ${2.112~\mu}$m}}

36: \def\hedubi{\hbox{\rm \HeI\ ${2.112/3~\mu}$m}}

37: \def\hetrit{\hbox{\rm \HeI\ ${1.700~\mu}$m}}

38: \def\hetwot{\hbox{\rm \HeI\ ${2.058~\mu}$m}}

39: \def\fetwot{\hbox{\rm \FeII\ ${2.089~\mu}$m}}

40: \def\mgdubi{\hbox{\rm \MgII\ ${2.13/4~\mu}$m}}

41: \def\henue{\hbox{\HeI$_{1.909~\mu m}$}}

42: \def\henuet{\hbox{\rm \HeI\ ${1.909~\mu}$m}}

43: \def\hetwo{\hbox{\HeI$_{2.058~\mu m}$}}

44: \def\hetri{\hbox{\HeI$_{1.700~\mu m}$}}

45: \def\hedub{\hbox{\HeI$_{2.112~\mu m}$}}

46: \def\rheh{\hbox{\rm \HeI$_{2.06}/B{\gamma}$}}

47: \def\el{\hbox{\rm \={e}}}

48: \def\ne{\hbox{\rm n$_{\rm e}$}}

49: \def\np{\hbox{\rm n$_{\rm p}$}}

50: \def\pdos{\hbox{\rm $2^{1}P$}}

51: \def\pdose{\hbox{\rm $2P$}}

52: \def\pdop{\hbox{\rm \scriptsize {\pdos}}}

53: \def\pdope{\hbox{\rm \scriptsize {\pdose}}}

54: \def\sdos{\hbox{\rm $2^{1}S$}}

55: \def\sdost{\hbox{\rm $2^{3}S$}}

56: \def\sdose{\hbox{\rm $2S$}}

57: \def\suno{\hbox{\rm $1^{1}S$}}

58: \def\sunoe{\hbox{\rm $1S$}}

59: \def\ltres{\hbox{\rm \pdos-\sdos}}

60: \def\ltrese{\hbox{\rm \pdose-\sdose}}

61: \def\ltrep{\hbox{\rm \scriptsize {\ltres}}}

62: \def\ltrepe{\hbox{\rm \scriptsize {\ltrese}}}

63: \def\luno{\hbox{\rm \pdos-\suno}}

64: \def\lunoe{\hbox{\rm \pdose-\sunoe}}

65: \def\lunop{\hbox{\rm \scriptsize {\luno}}}

66: \def\lunope{\hbox{\rm \scriptsize {\lunoe}}}

67: \def\ofr{\sc (r)}

68: \def\tc504{\hbox{$\tau_{\rm 504}$}}

69: \def\ngst{\hbox{\rm $n_{1S}$}}

70: \def\taunu{\hbox{$\tau_{\nu}$}}

71: \def\taunur{\hbox{$\tau_{\nu}$(r)}}

72: \def\taues{\hbox{$\tau_{\rm es}$}}

73: \def\tauro{\hbox{$\tau_{\rm R}$}}

74: \def\tauL{\hbox{$\tau_{\rm L}$}}

75: \def\taumax{\hbox{$\tau_{\rm max}$}}

76: \def\tauo{\hbox{$\tau_{\rm o}$}}

77: \def\taucon{\hbox{$\tau_{\rm cont}$}}

78: \def\taurtt{{\footnotesize \hbox{$\tau$=2/3}}}

79: \def\taur23{ \hbox{$\tau$=2/3}}

80: \def\nuo{\hbox{$\nu_{\rm o}$}}

81: \def\Vo{\hbox{$v_{\rm o}$}}

82: \def\Vsound{\hbox{V$_{\rm s}$}}

83: \def\Vrad{\hbox{V$_{\rm rad}$}}

84: \def\Vesc{\hbox{V$_{\rm es}$}}

85: \def\Vcore{\hbox{V$_{\rm core}$}}

86: \def\Heff{\hbox{h$_{\rm eff}$}}

87: \def\Vray{\hbox{V$_{\rm ray}$}}

88: \def\Vnew{\hbox{V$_{\rm new}$}}

89: \def\Vold{\hbox{V$_{\rm old}$}}

90: \def\Rmax{\hbox{R$_{\rm max}$}}

91: \def\Bij{\hbox{\rm B$_{\rm ij}$}}

92: \def\Bji{\hbox{\rm B$_{\rm ji}$}}

93: \def\Aji{\hbox{\rm A$_{\rm ji}$}}

94: \def\Rij{\hbox{\rm R$_{\rm ij}$}}

95: \def\Rji{\hbox{\rm R$_{\rm ji}$}}

96: \def\Rik{\hbox{\rm R$_{\rm ik}$}}

97: \def\Rki{\hbox{\rm R$_{\rm ki}$}}

98: \def\Ripi{\hbox{\rm R$_{{\rm i}^{\prime}{\rm i}}$}}

99: \def\Riip{\hbox{\rm R$_{{\rm ii}^{\prime}}$}}

100: \def\Rippi{\hbox{\rm R$_{{\rm i}^{\prime\prime}{\rm i}}$}}

101: \def\Riipp{\hbox{\rm R$_{{\rm ii}^{\prime\prime}}$}}

102: \def\nip{\hbox{\rm n$_{{\rm i}^{\prime}}$}}

103: \def\nipp{\hbox{\rm n$_{{\rm i}^{\prime\prime}}$}}

104: \def\Cij{\hbox{\rm C$_{\rm ij}$}}

105: \def\Cji{\hbox{\rm C$_{\rm ji}$}}

106: \def\Cik{\hbox{\rm C$_{\rm ik}$}}

107: \def\Cki{\hbox{\rm C$_{\rm ki}$}}

108: \def\Cipi{\hbox{\rm C$_{{\rm i}^{\prime}{\rm i}}$}}

109: \def\Ciip{\hbox{\rm C$_{{\rm ii}^{\prime}}$}}

110: \def\Cippi{\hbox{\rm C$_{{\rm i}^{\prime\prime}{\rm i}}$}}

111: \def\Pij{\hbox{\rm P$_{\rm ij}$}}

112: \def\Pji{\hbox{\rm P$_{\rm ji}$}}

113: \def\Zji{\hbox{\rm Z$_{\rm ji}$}}

114: \def\Yij{\hbox{\rm Y$_{\rm ij}$}}

115: \def\SL{\hbox{S$_{\rm L}$}}

116: \def\Sij{\hbox{\rm S$_{\rm ij}$}}

117: \def\Snu{\hbox{S$_{\nu}$}}

118: \def\Sdma{\hbox{$ \bar{\mbox{\boldmath $S$}}_d$}}

119: \def\Ndma{\hbox{$ \bar{\mbox{\boldmath $N$}}_d$}}

120: \def\Jlma{\hbox{$ \bar{\mbox{\boldmath $J$}}_l$}}

121: \def\Jlksma{\hbox{$ \bar{\mbox{\boldmath $J$}}_{lks}$ }}

122: \def\Jlksmap{\hbox{$ \bar{\mbox{\boldmath $J$}}_{lks}^{'}$ }}

123: \def\Jlksmapp{\hbox{$ \bar{\mbox{\boldmath $J$}}_{lks}^{''}$ }}

124: \def\Jlkjma{\hbox{$ \bar{\mbox{\boldmath $J$}}_{lkj}$ }}

125: \def\Jlkjmap{\hbox{$ \bar{\mbox{\boldmath $J$}}_{lkj}^{'}$ }}

126: \def\Jlkjmapp{\hbox{$ \bar{\mbox{\boldmath $J$}}_{lkj}^{''}$ }}

127: \def\fij{\hbox{\rm f$_{\rm ij}$}}

128: \def\chiL{\hbox{$\chi_{\rm L}$}}

129: \def\chiij{\hbox{$\chi_{\rm ij}$}}

130: \def\chinu{\hbox{$\chi_{\nu}$}}

131: \def\chiros{\hbox{$\chi_{R}$}}

132: \def\chic{\hbox{$\chi_{\rm c}$}}

133: \def\chies{\hbox{$\chi_{\rm es}$}}

134: \def\etaL{\hbox{$\eta_{\rm L}$}}

135: \def\etaij{\hbox{$\eta_{\rm ij}$}}

136: \def\etanu{\hbox{$\eta_{\nu}$}}

137: \def\etac{\hbox{$\eta_{\rm c}$}}

138: \def\etaes{\hbox{$\eta_{\rm es}$}}

139: \def\Jbar{\hbox{$\overline{\rm J}$}}

140: \def\JbarL{\hbox{$\overline{{\rm J}_{\rm l}}$}}

141: \def\Jbarij{\hbox{$\overline{{\rm J}}_{\rm ij}$}}

142: \def\Jres{\hbox{J$_{\rm Res}$}}

143: \def\Jdel{\hbox{J$_{\Delta}$}}

144: \def\Jc{\hbox{J$_{\rm c}$}}

145: \def\Jnu{\hbox{J$_{\nu}$}}

146: \def\Inu{\hbox{I$_{\nu}$}}

147: \def\Imunu{\hbox{I$_{\mu\nu}$}}

148: \def\Inup{\hbox{I$_{\nu}^+$}}

149: \def\Inum{\hbox{I$_{\nu}^-$}}

150: \def\Hnu{\hbox{H$_{\nu}$}}

151: \def\Hc{\hbox{H$_{\rm c}$}}

152: \def\Knu{\hbox{K$_{\nu}$}}

153: \def\fnu{\hbox{f$_{\nu}$}}

154: \def\Bnu{\hbox{B$_{\nu}$}}

155: \def\nuij{\hbox{$\nu_{\rm ij}$}}

156: \def\nud{\hbox{$\nu_{\rm D}$}}

157: \def\dnud{\hbox{$\Delta\nu_{\rm D}$}}

158: \def\alphaij{\hbox{$\alpha_{\rm ij}$}}

159: \def\alphaijnu{\hbox{$\alpha_{\rm ij} (\nu)$}}

160: \def\alphaik{\hbox{$\alpha_{\rm ik}$}}

161: \def\alphaiknu{\hbox{$\alpha_{\rm ik} (\nu)$}}

162: \def\sigmaij{\hbox{$\sigma_{\rm ij}$}}

163: \def\sigmaijv{\hbox{$\sigma_{\rm ij} (v)$}}

164: \def\sigmaik{\hbox{$\sigma_{\rm ik}$}}

165: \def\sigmaikv{\hbox{$\sigma_{\rm ik} (v)$}}

166: \def\qij{\hbox{\rm q$_{\rm ij}$}}

167: \def\qijt{\hbox{\rm q$_{\rm ij} ({\rm T})$}}

168: \def\sige{\hbox{$\sigma_{\rm e}$}}

169: \def\phinu{\hbox{$\phi_{\nu}$}}

170: \def\psinu{\hbox{$\psi_{\nu}$}}

171: \def\qnu{\hbox{q$_{\nu}$}}

172: \def\bi{\hbox{\rm b$_{\rm i}$}}

173: \def\bj{\hbox{\rm b$_{\rm j}$}}

174: \def\gi{\hbox{\rm g$_{\rm i}$}}

175: \def\gijk{\hbox{\rm g$_{\rm ijk}$}}

176: \def\gojk{\hbox{\rm g$_{\rm 0jk}$}}

177: \def\gojpk{\hbox{\rm g$_{\rm 0j+1k}$}}

178: \def\gj{\hbox{\rm g$_{\rm j}$}}

179: \def\chiijk{\hbox{$\chi_{\rm ijk}$}}

180: \def\chiIjk{\hbox{$\chi_{\rm Ijk}$}}

181: \def\Ni{\hbox{\rm N$_{\rm i}$}}

182: \def\Nj{\hbox{\rm N$_{\rm j}$}}

183: \def\Njk{\hbox{\rm N$_{\rm jk}$}}

184: \def\Nk{\hbox{\rm N$_{\rm k}$}}

185: \def\ni{\hbox{\rm n$_{\rm i}$}}

186: \def\nj{\hbox{\rm n$_{\rm j}$}}

187: \def\nk{\hbox{\rm n$_{\rm k}$}}

188: \def\nijk{\hbox{\rm n$_{\rm ijk}$}}

189: \def\nojk{\hbox{\rm n$_{\rm 0jk}$}}

190: \def\nojpk{\hbox{\rm n$_{\rm 0j+1k}$}}

191: \def\nij{\hbox{\rm n$_{\rm ij}$}}

192: \def\nik{\hbox{\rm n$_{\rm ik}$}}

193: \def\njk{\hbox{\rm n$_{\rm jk}$}}

194: \def\NH{\hbox{N$_{\hbox{H}}$}}

195: \def\NHe{\hbox{N$_{\hbox{He}}$}}

196: \def\Fnu{\hbox{F$_{\nu}$}}

197: \def\Fc{\hbox{F$_{\rm c}$}}

198: \def\nheh{\hbox{n$_{\rm He}$/n$_{\rm H}$}}

199: \def\nnihe{\hbox{n$_{\rm N}$/n$_{\rm He}$}}

200: \def\nnih{\hbox{n$_{\rm N}$/n$_{\rm H}$}}

201: \def\nmax{\hbox{n$_{\rm max}$}}

202: \def\nl{\hbox{n$_{\rm l}$}}

203: \def\Mbol{\hbox{M$_{\hbox{\sc bol}}$}}

204: \def\EBV{\hbox{E$_{\hbox{\sc B-V}}$}}

205: \def\Mdot{\hbox{$\dot {\rm M}$}}

206: \def\Rsun{\hbox{R$_\odot$}}

207: \def\Mstar{\hbox{M$_*$}}

208: \def\Rstar{\hbox{R$_*$}}

209: \def\Rt{\hbox{R$_{\rm T}$}}

210: \def\Rro23{\hbox{R$_{\tauro=2/3}$}}

211: \def\R23{\hbox{R$_{2/3}$}}

212: \def\Reff{\hbox{r$_{\rm eff}$}}

213: \def\Zsun{\hbox{\it Z$_\odot$}}

214: \def\Lsun{\hbox{L$_\odot$}}

215: \def\Lstar{\hbox{L$_*$}}

216: \def\Msun{\hbox{M$_\odot$}}

217: \def\Msunyr{\hbox{M$_\odot\,$yr$^{-1}$}}

218: \def\Minit{\hbox{\it M$_{\rm initial}$}}

219: \def\Myr{\hbox{\it Myr}}

220: \def\Gyr{\hbox{\it Gyr}}

221: \def\Teff{\hbox{T$_{\rm eff}$}}

222: \def\Tstar{\hbox{T$_*$}}

223: \def\T23{\hbox{T$_{2/3}$}}

224: \def\Te{\hbox{T$_{\rm e}$}}

225: \def\Logg{\hbox{$\log\,{\rm g}$}}

226: \def\grad{\hbox{$g_{\rm rad}$}}

227: \def\gcont{\hbox{$g_{\rm cont}$}}

228: \def\gline{\hbox{$g_{\rm line}$}}

229: \def\Loggeff{\hbox{$\log\,{\rm g}_{\rm eff}$}}

230: \def\geff{\hbox{g$_{\rm eff}$}}

231: \def\Vtur{\hbox{V$_{\rm tur}$}}

232: \def\Vesc{\hbox{V$_{\rm esc}$}}

233: \def\Vinf{\hbox{$v_\infty$}}

234: \def\kms{\hbox{km$\,$s$^{-1}$}}

235: \def\kpc{\hbox{kpc}}

236: \def\Hz{\hbox{Hz}}

237: \def\mum{\hbox{$\mu$m}}

238: \def\yr{\hbox{yr$^{-1}$}}

239: \def\Gam{\hbox{$\Gamma$}}

240: \def\Game{\hbox{$\Gamma_{\rm e}$}}

241: \def\qdef{\hbox{$\delta_{\rm nlS}$}}

242: \def\qnef{\hbox{$\nu_{\rm nlS}$}}

243: \def\HeI{He\,{\sc i}}

244: \def\HeII{He\,{\sc ii}}

245: \def\HeIII{He\,{\sc iii}}

246: \def\HII{H\,{\sc ii}}

247: \def\HI{H\,{\sc i}}

248: \def\CI{C\,{\sc i}}

249: \def\CII{C\,{\sc ii}}

250: \def\CIII{C\,{\sc iii}}

251: \def\CIV{C\,{\sc iv}}

252: \def\NI{N\,{\sc i}}

253: \def\NII{N\,{\sc ii}}

254: \def\NIII{N\,{\sc iii}}

255: \def\NIV{N\,{\sc iv}}

256: \def\NV{N\,{\sc v}}

257: \def\OI{O\,{\sc i}}

258: \def\OII{O\,{\sc ii}}

259: \def\OIII{O\,{\sc iii}}

260: \def\OIV{O\,{\sc iv}}

261: \def\OV{O\,{\sc v}}

262: \def\OVI{O\,{\sc vi}}

263: \def\MgII{Mg\,{\sc ii}}

264: \def\FeII{Fe\,{\sc ii}}

265: \def\FeIII{Fe\,{\sc iii}}

266: \def\ArIII{Ar\,{\sc iii}}

267: \def\NiII{Ni\,{\sc ii}}

268: \def\SiIV{Si\,{\sc iv}}

269: \def\SiII{Si\,{\sc ii}}

270: \def\SiIII{Si\,{\sc iii}}

271: \def\SII{S\,{\sc ii}}

272: \def\CrII{Cr\,{\sc ii}}

273: \def\CaII{Ca\,{\sc ii}}

274: \def\FeXIV{Fe\,{\sc XIV}}

275: \def\ie{\hbox{i.e.,}} 

276: \def\eg{\hbox{e.g.,}} 

277: \def\etal{\hbox{et~al.}}

278: \def\etc{\hbox{etc.}}

279: \def\cm{\hbox{\rm cm}}

280: \def\Kel{\hbox{\rm K}}

281: \def\keV{\hbox{\rm keV}}

282: \def\eV{\hbox{\rm eV}}

283: \def\dim{\hbox{\rm erg$\,$cm$^{-2}\,$s$^{-1}$}}

284: \def\sec{\hbox{\rm  $s^{-1}$}}

285: \def\scm{\hbox{\rm  s$^{-1}\,$cm$^3 $}}

286: \def\cmcube{\hbox{\rm cm$^{-3}$}}

287: \def\etadim{\hbox{\rm erg$\,$cm$^{-3}\,$sr$^{-1}\,$Hz$^{-1}\,$s$^{-1}$}}

288: \def\etaldim{\hbox{\rm erg$\,$cm$^{-3}\,$sr$^{-1}\,$s$^{-1}$}}

289: \def\Ang{\hbox{\AA}} 

290: \def\mnk{\hbox{\it m$_{\rm F205W}$}}

291: \def\mnh{\hbox{\it m$_{\rm F160W}$}}

292: \def\mnj{\hbox{\it m$_{\rm F110W}$}}

293: \def\fnk{\hbox{\it F$_{\rm F205W}$}}

294: \def\fnh{\hbox{\it F$_{\rm F160W}$}}

295: \def\fnj{\hbox{\it F$_{\rm F110W}$}}

296: \def\Fnp{\hbox{\it F$_{\rm F187N}$}}

297: \def\Fnc{\hbox{\it F$_{\rm F190N}$}}

298: \def\Frh{\hbox{\it F$_{\rm 8.5~GHz}$}}

299: \def\Frl{\hbox{\it F$_{\rm 4.9~GHz}$}}

300: \def\ew{\hbox{\it EW$_{\rm 1.87~\micron}$}}

301: \def\arcsec{$^{''}$}

302: 

303: \def\mnheh{\hbox{n$_{\rm He}$/n$_{\rm H}$}}

304: \def\nnihe{\hbox{n$_{\rm N}$/n$_{\rm He}$}}

305: \def\Mdot{\hbox{$\dot {M}$}}

306: \def\Zdot{\hbox{$\dot {Z}$}}

307: \def\Rsun{\hbox{\it R$_\odot$}}

308: \def\Zsun{\hbox{\it Z$_\odot$}}

309: \def\Rstar{\hbox{\it R$_*$}}

310: \def\Lsun{\hbox{\it L$_\odot$}}

311: \def\Lstar{\hbox{\it L$_*$}}

312: \def\Msun{\hbox{\it M$_\odot$}}

313: \def\Minit{\hbox{\it M$_{\rm initial}$}}

314: \def\Msunyr{\hbox{\it M$_\odot\,$yr$^{-1}$}}

315: \def\Myr{\hbox{\it Myr}}

316: \def\Gyr{\hbox{\it Gyr}}

317: \def\Teff{\hbox{\it T$_{\rm eff}$}}

318: \def\Vinf{\hbox{$v_\infty$}}

319: \def\kms{\hbox{km$\,$s$^{-1}$}}

320: \def\AV{\hbox{\it A$_{\rm V}$}}

321: \def\AJ{\hbox{\it A$_{\rm J}$}}

322: \def\AH{\hbox{\it A$_{\rm H}$}}

323: \def\AK{\hbox{\it A$_{\rm K}$}}

324: \def\H{\hbox{\it H}}

325: \def\K{\hbox{\it K}}

326: \def\mk{\hbox{\it K}}

327: \def\AL{\hbox{\it A$_{\rm L}$}}

328: \def\BCK{\hbox{\it BC$_{\rm K}$}}

329: \def\BCV{\hbox{\it BC$_{\rm V}$}}

330: \def\simgr{\mathrel{\hbox{\rlap{\hbox{\lower4pt\hbox{$\sim$}}}\hbox{$>$}}}}

331: \def\HH{H{\sc ii}}	% HII region

332: \def\mnk{\hbox{\it m$_{\rm F205W}$}}

333: \def\mnh{\hbox{\it m$_{\rm F160W}$}}

334: \def\mnj{\hbox{\it m$_{\rm F110W}$}}

335: \def\fnk{\hbox{\it F$_{\rm F205W}$}}

336: \def\fnh{\hbox{\it F$_{\rm F160W}$}}

337: \def\fnj{\hbox{\it F$_{\rm F110W}$}}

338: \def\Fnp{\hbox{\it F$_{\rm F187N}$}}

339: \def\Fnc{\hbox{\it F$_{\rm F190N}$}}

340: \def\Frh{\hbox{\it F$_{\rm 8.5~GHz}$}}

341: \def\Frl{\hbox{\it F$_{\rm 4.9~GHz}$}}

342: \def\ew{\hbox{\it EW$_{\rm 1.87~\micron}$}}

343: \def\hh{H$_{2}$}

344: \def\cc{C$_{2}$}

345: \def\km/s{km~s$^{-1}$}

346: \def\um{${\mu}$m}

347: \def\Vlsr{v$_{LSR}$}

348: \def\wvnum{cm$^{-1}$}

349: \def\hhhp{H$_{3}^{+}$}

350: \def\hhp{H$_{2}^{+}$}

351: \def\kco{{\it k}$_{CO}$}

352: \def\ke{{\it k}$_{e}$}

353: \def\ne{n$_{e}$}

354: \def\CO{$^{12}$CO}

355: \def\co{$^{13}$CO}

356: \def\Lo{L$_\odot$}

357: \def\Vinf{\hbox{$V_\infty$}}

358: \def\HeI{He\,{\sc i}}

359: \def\HeII{He\,{\sc ii}}

360: \def\HII{H\,{\sc ii}}

361: \def\CII{C\,{\sc ii}}

362: \def\CIII{C\,{\sc iii}}

363: \def\CIV{C\,{\sc iv}}

364: \def\NI{N\,{\sc i}}

365: \def\NII{N\,{\sc ii}}

366: \def\NIII{N\,{\sc iii}}

367: \def\OII{O\,{\sc ii}}

368: \def\OIII{O\,{\sc iii}}

369: \def\NaI{Na\,{\sc i}}

370: \def\MgII{Mg\,{\sc ii}}

371: \def\FeII{Fe\,{\sc ii}}

372: \def\NiII{Ni\,{\sc ii}}

373: \def\SiIII{Si\,{\sc iii}}

374: \def\SiII{Si\,{\sc ii}}

375: \def\Mdot{\.{M}}

376: \def\FMM362{FMM362}

377: 

378: \defcitealias{naj04}{Paper I}

379: 

380: 

381: \shorttitle{Metallicity in the Quintuplet cluster}

382: \shortauthors{Najarro et al.}

383: 

384: \begin{document}

385: \title{Metallicity in the Galactic Center: \\ The Quintuplet cluster}

386: 

387: \author{

388: Francisco Najarro\altaffilmark{1}, Don F. Figer\altaffilmark{2}, 

389: D. John Hillier\altaffilmark{3},  T. R. Geballe\altaffilmark{4}, Rolf P. Kudritzki\altaffilmark{5}}

390: 

391: \email{najarro@damir.iem.csic.es}

392: 

393: \altaffiltext{1}{Instituto de Estructura de la Materia, CSIC, Serrano 121, 29006 Madrid, Spain }

394: \altaffiltext{2}{Chester F. Carlson Center for Imaging Science, Rochester Institute of Technology, 54 Lomb Memorial Drive, Rochester, NY 14623}

395: \altaffiltext{3}{Department of Physics and Astronomy, University of Pittsburgh, 3941 O'Hara Street, Pittsburgh, PA 15260}

396: \altaffiltext{4}{Gemini Observatory, Hilo, 670 N. A'ohoku Pl., HI 96720}

397: \altaffiltext{5}{Institute for Astronomy, University of Hawaii, 2680 Woodlawn Drive, Honolulu, HI 96822}

398: 

399: \begin{abstract}

400: 

401: We present a measurement of metallicity in the Galactic center Quintuplet

402: Cluster made using

403: quantitative spectral analysis of two Luminous Blue Variables (LBVs).

404: The analysis employs line-blanketed NLTE

405: wind/atmosphere models fit to high-resolution near-infrared spectra

406: containing lines of H, \HeI, \SiII, \MgII, and \FeII.  We are able to

407: break the H/He ratio vs. mass-loss rate degeneracy found in other LBVs and

408: to obtain robust estimates of the He content of both objects. Our results

409: indicate solar iron abundance and roughly twice solar abundance in the

410: $\alpha$-elements. These results are discussed within the framework of

411: recent measurements of oxygen and carbon composition in the nearby

412: Arches Cluster and iron abundances in red giants and supergiants

413: within the central 30~pc of the Galaxy. The relatively large enrichment of

414: $\alpha$-elements with respect to iron is consistent with a history of

415: more nucleosynthesis in high mass stars than the Galactic disk. \end

416: {abstract}

417: 

418: \keywords{Galaxy: abundances -- stars: abundances -- stars: individual

419: (Pistor Star, FMM362) -- infrared: stars -- Galaxy: center}

420: 

421: 

422: 

423: \section {Introduction}

424: 

425: Elements heavier than hydrogen and helium (``metals'') are primarily

426: created by nucleosynthesis in stars. Metals are important ingredients in

427: many astrophysical processes such as radiative cooling, and mass-loss

428: during star formation and at all stages of

429: stellar evolution. They also play a fundamental role in stellar evolution

430: through their influence on stellar opacities, and represent a historical

431: record of galactic chemical enrichment via stellar winds and supernovae

432: ejecta.

433: 

434: In the Galaxy metal abundance increases with decreasing galactocentric

435: radius, as seen in stars and gas

436: \citep{affler97,rud06,mac99,fuhr98,roll00,sma01,luck06}. Other galaxies show a similar

437: trend, having highest metal abundances in their nuclei

438: \citep{urba05,kenni03}.

439: 

440: Previous work \citep{fro99,fel00,car00,ram97,ram99,ram00} on the Galactic

441: center (GC) has indicated roughly solar stellar metal abundances, whereas

442: the analyses of interstellar emission lines \citep{shi94,mae00} have

443: suggested considerably higher abundances. It is not clear why the stellar

444: and gas-phase measurements should differ so greatly.

445: 

446: The GC contains three dense and massive star clusters that have recently

447: formed in the inner 50 pc, the Arches, Quintuplet, and Central clusters.

448: Using quantitative spectral analysis, \citet{naj04} (Paper I) determined

449: that the WNL stars in the very young (2-2.5~Myr) Arches Cluster have

450: roughly solar metallicities. Being more evolved ($\sim$4Myr), the

451: Quintuplet Cluster \citep{gla87,gla90,nag90,oku90,mon94} contains a

452: variety of massive stars, including WN, WC, WN9/Ofpe, luminous blue

453: variables (LBVs) and less evolved blue-supergiants

454: \citep{fig95,fig99a,fig99b}. Two LBVs in it are known, the Pistol Star

455: \citep{mon94,cot94,fig95,fig98,fig99c} and FMM362 \citep{fig99b,geb00},

456: each having an infrared spectrum rich in metal lines of \FeII, \SiII, \&\

457: \MgII.

458: 

459: In this paper, we use quantitative infrared spectroscopy of the two

460: Quintuplet LBVs to make direct determinations of metallicity in those

461: stars. We also use the derived $\alpha$-elements vs. Fe ratio to address

462: the dominance of massive stars on the IMF in this region.

463: 

464: 

465: \begin{figure*}

466: \epsscale{1.05}     %% submission

467: %\epsscale{1.15}      %% emulateapj 

468: \vspace{-.1cm}

469: \plotone{fig1.ps}

470: \caption{\label{fig:pistol}

471: Model fits ({\it dashed lines}) to the observed infrared diagnostic lines 

472: ({\it solid lines}) of the Pistol Star. The forbidden [\FeII] line at 

473: 1.677~$\mu$m was not included in the models.}

474: 

475: \end{figure*}

476: 

477: 

478: \section {Observational Data}

479: 

480: The data were obtained at UKIRT\footnote{The United Kingdom Infrared

481: Telescope (UKIRT) is operated by the Joint Astronomy Centre on behalf of

482: the Particle Physics and Astronomy} using CGS4.  The Pistol Star was

483: observed in April 1996 (P$\alpha$; R$\sim$3000), July 1997 (L;

484: R$\sim$16000) and April 1998 (H; R$\sim$5000 and K R$\sim$3000). Likewise,

485: spectra for \FMM362 were obtained in April (L) and May 1999 (H and K),

486: using CGS4 in medium resolution mode (R$\sim$5,000-6,500).  The slit width

487: was 0$\farcs$6 for all observations.  We used the photometric measurements

488: of \citet{fig98} for the Pistol Star, to scale the reduced spectra. For

489: \FMM362, given its photometric variability, we adopted the average value,

490: K=7.30, obtained by \citet{gla99} for the epoch closest to our

491: spectroscopic observations. This value agrees, within the 0.26 standard

492: deviation derived by \citet{gla99}, with the K=7.50 value adopted by

493: \citet{geb00} from flux-calibrated spectra. We assume the same extinction

494: for both objects and adopt the value of A$_K$=3.2 derived by \citet{fig98}

495: for the Pistol Star. The reader is referred to these papers for a detailed

496: discussion on the reduction of the observed spectra and photometry.

497: 

498: \begin{figure*}

499: \epsscale{1.00}

500: \vspace{-.1cm}

501: \plotone{fig2.ps}

502: \caption{\label{fig:362}

503: Model fits ({\it dashed lines}) to the observed infrared diagnostic lines 

504: ({\it solid lines}) of \FMM362.}

505: \end{figure*}

506: 

507: 

508: 

509: \section{Models}

510: 

511: To model the LBVs and estimate their physical parameters, we have used CMFGEN,

512: the iterative, non-LTE line blanketing method presented by \citet{hil98} which

513: solves the radiative transfer equation in the co-moving frame and in

514: spherical geometry for the expanding atmospheres of early-type stars. The

515: model is prescribed by the stellar radius, \Rstar, the stellar luminosity,

516: \Lstar, the mass-loss rate, \Mdot, the velocity field, $v(r)$ (defined by

517: \Vinf\ and $\beta$), the volume filling factor characterizing the clumping

518: of the stellar wind, {\it f(r)} (see Sec.~\ref{sub-clu}), and elemental

519: abundances. \citet{hil98,hil99} present a detailed discussion of the code.

520: For the present analysis, we have assumed the atmosphere to be composed of

521: H, He, C, N, O, Mg, Si, S, Fe and Ni. 

522: Given the parameter domain the LBVs are located, the $\tau=2/3$ radius 

523: is located close or above the sound speed, and therefore the assumed

524: hydrostatic structure plays no role. Thus, no spectroscopic information

525: about the mass of the object can be obtained.

526: The atomic data sources are described in detail \citet{hil01}.

527: Here we focus on the model atoms used for our abundance determinations \FeII,

528: \MgII\ and \SiII. Using the superlevel formalism 

529: \citep[NS/NF, number of superlevels vs. number of levels in the full atom, e.g.][]{hil98} we

530: we chose 233/709 (up to 109800~cm$^{-1}$) for \FeII,

531: 37/50 (up to 119400~cm$^{-1}$) for \MgII\ and 35/72 

532: (up to 125000~cm$^{-1}$) for \SiII.

533: The choice of the appropiate packing has been  extensively tested in

534: \citet{naj01}. We will revise the importance of this issue for the 

535: case of Mg in \ref{sub-meta}.

536: 

537: Observational constraints are

538: provided by the H, K and L-band spectra of the stars and the dereddened K

539: magnitudes from \citet{fig98}, \citet{geb00} and \citet{gla99}.

540: As in \citetalias{naj04},

541: a distance of 8~\kpc\ has been assumed. The validity of our technique has

542: been demonstrated in \citet{naj99} and \citet{naj01} by calibrating our

543: method against stars with similar spectral type such as P~Cygni and

544: HDE~316285 for which not only infrared but also optical and UV spectra

545: are available.

546: 

547: \input table1.tex

548: 

549: 

550: \section{Results}

551: \label{sec-results}

552: 

553: Table~1 gives the derived stellar parameters for both LBVs and

554: Figs.~\ref{fig:pistol} and \ref{fig:362} show model fits to the relevant

555: lines in the stars. Theoretical spectra have been convolved with the

556: instrumental resolution. We note that given the large number of parameters

557: involved in the analysis it is

558: unafordable to perform a full systematic error analysis in the whole

559: parameter domain. We rather proceed by estimating the range of values for the

560: main stellar parameters which provide acceptable fits to the observed spectra.

561: Once those ranges are set, we derive the corresponding abundances and their

562: errors.

563: From Table~1, it can be seen that the Pistol Star

564: and \FMM362 have very similar properties, with the exceptions of the

565: Pistol Star's significantly higher wind density (evidenced by the its

566: stronger spectral lines) and its higher He content. The latter may denote

567: a slightly advanced evolutionary stage for the Pistol Star (see below).  

568: Given the general resemblance of the spectra of the objects, we discuss

569: them together.

570: 

571: \subsection{Main Diagnostic Lines and  Stellar Properties}

572: \label{sub-prop}

573: 

574: Several spectral diagnostics constrain our estimates of the stellar

575: temperature, and thus the ionization structure, in particular the \HeI\

576: (5-4) components near 4.05~\um.

577: If helium is predominantly singly ionized, even for the most

578: favorable case with (minimum) cosmic helium abundance,

579: the observed ratio of H to  \HeI\ lines exceeds the expected values by 

580: large factors.

581: This indicates that \HeII\ must recombine to \HeI\ very close

582: to the photosphere, implying an upper limit of around 13,000~K for the

583: temperatures of these objects. We find a lower limit of 10,000~K for the

584: temperatures, as lower values would require non-detection of the \HeI\

585: components.  Also the strengths of the \hetrit\ and \hedubi\ lines are

586: very sensitive to temperature, so that they appear in emission above

587: 12,500~K and vanish below 10,500~K. These lower limits on the effective

588: temperature are also consistent with the non-detections of the \SiII\

589: 3s$^2$3p$^2$S$_{1/2}$-3s$^2$4p$^2$P$_{3/2}$~2.180~\um\ and

590: 3s$^2$3p$^2$S$_{1/2}$-3s$^2$4p$^2$P$_{1/2}$~2.209~\um\ intercombination

591: lines, as they are expected to appear strongly in absorption as soon as

592: the temperature drops below 10,000~K. Hence, $\Delta$T$\pm1500$~K are

593: conservative estimates of uncertainties for the temperatures of

594: the Pistol Star and \FMM362 given in Table~1.

595: 

596: \begin{figure*}

597: \epsscale{1.0}

598: \vspace{-.1cm}

599: \plotone{fig3.ps}

600: \vspace{-.1cm}

601: \caption{\label{fig:l362_clump}

602: Model spectra showing the effect of clumping parameter CL$_1$

603: (see Eq.\ref{eq:clump}) on diagnostic line profiles of \FMM362. Values of 

604: CL$_1$ are listed in the bottom left panel.} 

605: \end{figure*}

606: 

607: To estimate the terminal velocities, we make use of the \FeII]

608: (semi-forbidden \FeII) z$^{4}$F$_{9/2}$-c$^{4}$F$_{9/2}$ 1.688~\um\ line

609: \citep{geb00,fig98} that forms in the outer wind and has a weak oscillator

610: strength (gf$\sim$10$^{-5}$). This is because non-negligible continuum 

611: opacity effects at 4~\um\ may provide only lower limits if Br~$\alpha$ is 

612: used. The larger \Vinf\ derived for \FMM362\ can be clearly inferred from 

613: the width of this \FeII] line and  the obvious overlap at 

614: \bap\ between the \HI\ and \HeI\ components (see Fig.~\ref{fig:362}). For 

615: the Pistol Star (Fig.~\ref{fig:pistol}) the \bap\ components are fairly 

616: well separated.

617: 

618: Our analysis of wind density (\Mdot, $\beta$) gives values of $\beta$ that

619: agree with those inferred in the literature for other LBVs \citep{naj01}

620: and B-Supergiants \citep[e.g][]{cro06}. The values are fairly well

621: constrained by the shapes of the hydrogen lines, especially those of \bap\

622: and \bgam\, which are inconsistent with the same value of $\beta$

623: for both objects.

624: 

625: Although the wind density derived for the Pistol Star is much higher than

626: for \FMM362, the modified wind momenta D$_{mom}$ = log (\Mdot \Vinf

627: $\sqrt{R/\Rsun}$) \citep{kud00} of the two LBVs are nearly identical

628: (Table~1). This result is qualitatively consistent with the wind momentum

629: - luminosity relation \citep{kud00} which predicts the same modified

630: momenta for objects with the same stellar type and luminosity. Note that

631: we have used clumping-corrected values of \Mdot\ to compute the modified

632: momenta. If unclumped values were assumed, D$_{mom}$ would be closer to

633: 30.0. Interestingly, the latter agrees very well with the averaged

634: modified momentum of AG~Car at maximum (\ie\ at similar \Teff\ to our

635: objects) obtained using the values of \Mdot, \Vinf\ and \Rstar\ derived by

636: \citet{sta01} from fits to \hap. Those authors obtained

637: log(\Mdot)$\sim$-4.1 for about this temperature, while our unclumped

638: values are log(\Mdot)=-4.1 and -4.4 for the Pistol Star and \FMM362\,

639: respectively. Further, using radiation-driven wind models for LBVs,

640: \citet{vin02} were able to predict the \citet{sta01} \Mdot\ value for

641: AG~Car assuming \Vinf/\Vesc$\sim$1.3 and a current stellar mass of

642: 35\Msun. The same value of \Vinf/\Vesc\ for the LBVs would imply current

643: stellar masses of 27.5\Msun\ for the Pistol Star and 46\Msun\ for \FMM362.

644: Although these masses should be regarded with caution, they are consistent

645: with the Pistol Star being more evolved than \FMM362 (as inferred from

646: their He/H ratios) and hence having lost more mass during its evolution,

647: as indicated by the presence of a nebula around it.

648: 

649: Compared to the results obtained in \citet{fig98} by means of

650: non-blanketed models, the new blanketed models provide a significant

651: improvement in our knowledge of the physical properties of these two

652: stars. The degeneracy of the ``high'' and ``low'' luminosity (\Teff)

653: solutions for the Pistol Star presented by \citet{fig98} is broken by the

654: \SiII, \MgII\ and \FeII\ lines, which are clearly more consistent with the

655: ``low'' solution. We derived for this star a luminosity of

656: $\sim$1.6($10^6$)~\Lsun, an effective temperature of $\sim$11,800~\Kel,

657: and an initial mass of 100~\Msun. The stellar luminosity is reduced by a

658: factor of two compared with the previous estimate, illustrating the

659: importance of the new generation of line-blanketed models.  Below, we

660: discuss in detail the role of two additional stellar properties that are

661: derived using the new models, wind clumping and elemental abundances.

662: 

663: \subsection{Clumping}

664: \label{sub-clu}

665: 

666: Clumping is normally invoked in stellar winds to explain inconsistencies

667: arising between $\rho$ (density) and $\rho^2$ diagnostics. For a given

668: mass-loss, clumping causes an enhancement of $\rho^2$ processes while

669: leaving unaltered those which depend linearly on $\rho$.  Further, if

670: mass-loss rate and clumping are scaled without changing the

671: \Mdot$/f^{0.5}$ ratio, the $\rho$-dependent diagnostics vary while

672: the recombination lines profiles ($\propto \rho^2$) remain basically

673: unaltered.

674: 

675: To investigate the clumping we introduce the following clumping law:

676: 

677: \begin{equation}

678: \label{eq:clump}

679:   f = CL_1 + ( 1 -CL_1 )  e^{\frac{-V}{CL_2}} + ( CL_4 - CL_1 ) 

680: e^{\frac{(V-V_{\infty})}{CL_3}}

681: \end{equation}

682: 

683: \noindent where CL$_1$ and CL$_4$ are volume filling factors and CL$_2$ and CL$_3$

684: are velocity terms defining locations in the stellar wind where the

685: clumping structure changes. CL$_1$ sets the maximum degree of clumping

686: reached in the stellar wind (provided CL$_4$$>$CL$_1$) while CL$_2$

687: determines the velocity of the onset of clumping. CL$_3$ and CL$_4$

688: control the clumping structure in the outer wind. Hence, when the wind

689: velocity approaches \Vinf, so that (V-\Vinf)$\leq$CL$_3$, clumping starts

690: to migrate from CL$_1$ towards CL$_4$. If CL$_4$ is set to unity, the wind

691: will be unclumped in the outermost region.  Such behavior was already

692: suggested by \citet{nug98} and was utilized by \citet{fig02} and

693: \citet{naj04} for the analysis of the WNL stars in the Arches Cluster.

694: Recently, \citet{pul06} also have found similar behavior from \hap\ and

695: radio studies of OB stars with dense winds. Furthermore, our clumping

696: parametrization is consistent with results from hydrodynamical

697: calculations by \citet{run02}. From Eq.~\ref{eq:clump} we note that if

698: CL$_3$, and therefore CL$_4$, is not considered (CL$_3\rightarrow$0), we

699: recover the simpler variation proposed by \citet{hil99}. To avoid entering free

700: parameters heaven we set CL$_4$=1 in all of our investigations, aiming to

701: get an appropriate amount of leverage on the amount of non-constant

702: clumping in the outer wind regions.

703: 

704: 

705: \begin{figure*}

706: \epsscale{1.15}    %% emulateapj

707: %\epsscale{1.15}     %% submission

708: \vspace{-.1cm}

709: \plottwo{fig4a.ps}{fig4b.ps}

710: \vspace{-.1cm}

711: \plottwo{fig4c.ps}{fig4d.ps}

712: \caption{\label{fig:pistol_clump}

713: Influence of adopted clumping structure

714: (see Eq.~\ref{eq:clump})

715: in profiles.  Run of clumping

716: and line profile sensitivities for different values of CL$_2$

717: ({\it{left panel}}) and CL$_3$ ({\it{right panel}}), compared to observed 

718: profiles in the Pistol Star. See discussion in text.}

719: \end{figure*}

720: 

721: 

722: 

723: 

724: {\bf{CL$_1$: Estimating the wind clumpiness.}} Figure~\ref{fig:l362_clump}

725: illustrates the sensitivity to CL$_1$ of the main diagnostic lines

726: utilized to obtain the degree of clumping in the stellar winds of \FMM362.

727: For each value of CL$_1$ displayed in Fig.~\ref{fig:l362_clump} the

728: mass-loss rate in the model was scaled while keeping \Mdot$/f^{0.5}$

729: constant as described previously. Although there are some lines that

730: follow this scaling quite well (\bap, and also \bgam\ and some metal lines

731: not displayed in the figure), the \HeI\ lines and weak \HI\ lines react

732: quite sensitively to the absolute degree of wind clumping. It can be seen

733: that the \HeI\ lines only provide an upper limit to CL$_1$ (a bit lower

734: than 0.1) and do not react to lower values, but a unique value for

735: CL$_{1}$ can be selected by some \HI\ lines. The Hu$_{14}$ (\HI(14-6))

736: line displays the highest sensitivity to clumping.  Unfortunately, the

737: wavelength interval surrounding this line was not observed with

738: sufficiently high S/N in \FMM362\ and there is a large uncertainty in the

739: continuum value, which is critical for estimating CL$_{1}$. We could make

740: full use of the Hu$_{14}$ line to determine the clumping only for the

741: Pistol Star, where this line is relatively stronger in emission (see

742: Fig.~\ref{fig:pistol}). Nevertheless, Fig.~\ref{fig:l362_clump} shows that

743: CL$_1$ lies between 0.1 and 0.05. It must be stressed that only with a

744: well determined clumping may we address the He/H abundance issue (see

745: Sec.~\ref{sub-hhe}).

746: 

747: 

748: {\bf{CL$_2$ \&\ CL$_3$: Mapping the clumping structure.}} The upper panels

749: of Fig.~\ref{fig:pistol_clump} illustrate the behavior of the clumping

750: structure for different sets of CL$_2$ and CL$_3$ values, while the lower

751: panels display the influence of such behavior on diagnostic lines in the

752: spectrum of the Pistol Star. It is evident that for some spectral lines

753: \eg~\hedubt, the behavior of the profiles with clumping is far from being

754: monotonic. Furthermore, not only do lines of different ions react

755: differently to clumping, but also lines within the same ion, \eg~\HI,

756: behave differently. For example, \FeII\ does not respond in the same way

757: to changes in CL$_2$ and CL$_3$. Figure~\ref{fig:pistol_clump} shows the

758: great potential of the different IR lines to constrain the clumped

759: structure of the stellar wind and demands the following detailed

760: discussion.

761: 

762: The general impact of clumping on line profiles that was described at the

763: beginning of this section will occur provided the ionization equilibrium

764: is on the ``safe'' side. We consider the ``safe'' region to be where the

765: population of the next ionization stage clearly dominates over the one the

766: line belongs to (\ie\ \HII$\gg$\HI\ for the hydrogen lines). Noting,

767: however, that ionization depends linearly on density whereas recombination

768: is proportional to $\rho^2$, a ``changing'' ionization situation may

769: occur, where two adjacent ionization stages have similar populations. In

770: such a case clumping, which enhances recombination, will cause a net

771: reduction of the mean ionization. This will result in weaker lines.

772: 

773: Finally, in the infrared, via bound-free and free-free processes 

774: $\propto \rho^2$), not only the lines but also the continuum will depend 

775: on clumping, resulting in high sensitivity of the continuum-rectified

776: line profiles to CL$_1$ and CL$_2$. 

777: 

778: One may, therefore, distinguish between lines formed on the ``safe''

779: region and those arising from the ``changing'' region.  Within the

780: parameter domain of the two LBVs studied here we find that \HI, \SiII, and

781: \MgII\ lines and also the \FeII\ photospheric lines are formed in

782: ``safe'' regions, while \HeI\ and \FeII] lines arise from

783: ``changing'' regions.

784: 

785: Increasing the clumping (decreasing the CL$_1$ value), or alternatively

786: decreasing the velocity at which clumping sets in (decreasing CL$_2$)

787: results in stronger \SiII\ and \MgII\ lines, as shown in

788: Fig.~\ref{fig:pistol_clump}-left.  The strong \HI\ lines are formed

789: further out than the \SiII\ and \MgII\ lines and their strengths should in

790: principle show no sensitivity to CL$_2$. However, the continuum is clearly

791: affected by clumping. Thus the stronger the clumping (lower CL$_2$), the

792: stronger the continuum and the weaker the resulting line-to-continuum

793: ratio, as clearly shown by \bgam\ and \bap\ in

794: Fig.~\ref{fig:pistol_clump}-left.  On the other hand, weaker \HI\ lines

795: such as Hu$_{14}$ or B$_{10}$ and B$_{11}$ form much closer to the

796: photosphere and tend to brighten with increasing clumping. The weak \FeII\

797: lines formed close to the photosphere react in basically the same way as

798: the continuum and thus the normalized spectra of them show no changes with

799: changing clumping. Their near independency allows these lines to be used

800: as Fe abundance indicators (see below).  The \FeII] lines, formed even

801: beyond the \HI\ lines, are affected by two competing processes. On one

802: hand, increasing the extent of the clumped region (decreasing CL$_2$)

803: results in a reduction of the \FeIII/\FeII\ ratio in the wind. Since

804: \FeIII\ remains the dominant ionization stage in the \FeII] line formation

805: zone, this change will cause a slight increase in the strengths of the

806: \FeII] lines. On the other hand, as the continuum increases with

807: increasing clumping, the line-to-continuum ratio decreases.  Thus the two

808: processes counter-balance (see Fig.~\ref{fig:pistol_clump}-left). Finally,

809: the \HeI\ lines, which form close to the photosphere, show weak continuum

810: dependences, but high sensitivities to ionization/recombination. Thus,

811: starting with the model with the highest CL$_2$ values, the \HeI\ lines

812: are not affected by clumping, but the stellar parameters produce strong

813: ionization in the inner parts resulting in overly strong emission

814: (\hetrit) and line filling (\hedubt). However, as clumping is enhanced in

815: the line formation zone, recombination starts to dominate over ionization

816: and the \HeI\ line emission weakens, the lines are no longer as filled,

817: and start to appear in absorption.

818:   

819: Regarding clumping in the outer parts of the wind, it can be seen on the

820: right side of Fig.~\ref{fig:pistol_clump} that only the strong \HI\ and

821: \FeII] lines react to CL$_3$.  Note that \bap\, which forms further out than

822: \bgam\, is more sensitive to clumping and the observed ratio of the two

823: profiles may be used to determine CL$_3$. For winds of significantly lower

824: density, these lines will form further in and show little or no dependence

825: on CL$_3$ (\eg~\FMM362). The \FeII] lines are more sensitive to CL$_3$,

826: primarily due to the coupling of the Fe ionization structure with that of

827: hydrogen thru charge-exchange reactions in the outer wind zones where \HII\

828: starts to recombine. Due to the difference in H and Fe abundances, a small

829: and hardly noticeable change in the ionization of hydrogen will be amplified

830: in the \FeIII/\FeII\ ratio, resulting in a large change in iron

831: recombination. Thus, decreasing CL$_3$ dramatically enhances the \FeII]

832: lines, as shown in on the right side of Fig.~\ref{fig:pistol_clump}. The

833: \HI\ lines behave similarly to the way they do in ``safe'' regions.

834:   

835: For the Pistol Star the best fits to the various line profiles suggest a

836: decreasing clumping factor that becomes unity in the outer wind, whereas for

837: the less dense wind of \FMM362, the line profiles are best matched by a

838: constant clumping factor.  From the clumping estimates, we find

839: uncertainties of 0.10~dex in the mass-loss rates.  The value of CL$_1=0.08$

840: obtained for each star is somewhat low compared to the values derived for

841: other LBVs such P~Cygni \citep[CL$_1$=0.5,][]{naj01} or AG~Car

842: \citep[CL$_1$=0.25,][]{gro06} and is more consistent with those derived for

843: WR stars \citep[CL$_1$$\sim$0.1,][]{her01,naj04}. We note, however, that the

844: LBVs have higher He abundances than P~Cygni, pointing to a more evolved

845: status, closer to the WR phase.

846: 

847: 

848: \begin{figure}

849: \epsscale{1.00}    %% submission

850: %\epsscale{1.15}  %% emulateapj

851: \vspace{-.1cm}

852: \plotone{fig5.ps}

853: \caption{\label{fig:hei362}

854: Breakdown of the H/He degeneracy in \FMM362. Models with H/He ratios

855: ranging from 5.0 to 0.75 provide identical \HI\ and strong \HeI\ line

856: profiles, but weaker \HeI\ profiles can be used to determine the He

857: abundance (see text)}

858: \end{figure}

859: 

860: 

861: \subsection{H/He ratio. Breaking the degeneracy}

862: \label{sub-hhe} 

863:  

864: \citet{hil98b} showed that for HDE~316285, an LBV-like star with similar

865: \Teff\ and slightly higher wind density than the Pistol Star, a degeneracy

866: exists between the H/He ratio and the mass-loss rate.  In principle, fits of

867: virtually equal quality could be obtained with H/He ratios varying from 0.05

868: to 10 by scaling the mass-loss rate. Such a degeneracy, if present in the

869: Quintuplet LBVs, would imply that if the H/He ratio falls below a certain

870: value (H/He$\leq$2), the resulting metal abundances could be scaled down to

871: obtain the same line strengths. Hence one could obtain only an upper limit

872: on the metal abundances. breaking this degeneracy is crucial to

873: understanding the evolutionary status of these objects.

874:  

875: Because of the lower wind density of the Quintuplet LBVs and the

876: sensitivities of some of the infrared lines to the stellar parameters, we

877: are able to break the H/He degeneracy and obtain robust estimates of their

878: He content. Due to the high degree of clumping found in both objects, the

879: $\tau=2/3$ radius, where \Teff\ is defined, is reached at considerably lower

880: velocities than for classical LBVs. Thus, wind speeds roughly between half

881: and one-third of the sound speed are found in the Pistol Star and \FMM362\

882: while classical LBVs have wind speeds well above the speed of sound

883: \citep{hil98b,naj01}.  This enables quasi-photospheric absorption lines to

884: form. The  \hedubt\ line is the key in breaking the degeneracy. This is

885: shown for \FMM362\ in Fig.~\ref{fig:hei362}, which contains model spectra

886: computed for H/He ratios ranging from H/He=5.0 to 0.75, with mass-loss rates

887: and metal abundances scaled and the other stellar parameters fine-tuned to

888: reproduce the observed profiles of other lines.  The figure shows that while

889: identical \HI\ and \HeI\ (\hetrit) line profiles (also for the rest of

890: hydrogen and metal lines) are obtained for all H/He ratios considered, the

891: absorption {\it depths} of the \hedubt\ and \HeI~2.15~\my\ lines react

892: sensitively to the He abundance.  Both lines show that the best H/He value

893: must lie between 3.33 and 2.25, and we find a most likely value of 2.8 (see

894: Table~1). Similar behavior was found for the Pistol Star, where we obtain

895: H/He=1.5.

896: 

897: \subsection{Metal Abundances}

898: \label{sub-meta}

899: 

900: For the purpose of discussing metal abundances (see Table~2), we adopt the solar

901: composition of \citet{gre93}. Although their abundances have been recently

902: revised \citep{asp05,alle08} (but see also \citet{pin06}), they are the ones used

903: by \citet{igl96} to compute stellar interior opacities and adopted in the

904: most recent evolutionary models for massive stars with rotation from the

905: Geneva group \citep{mey03,mey05}, and the Padova tracks used for cooler,

906: less massive stars \citep{gir00,sal00}. Previously published evolutionary

907: tracks for massive stars \citep{scha92,mey94} used opacity tables calculated

908: with solar composition from \citet{and89}, which differ significantly from

909: \citet{gre93} only in Fe (A(Fe/H)=7.67 vs 7.50 in

910: \citet{gre93})\footnote{A(X/Y)=log[n(X)/n(Y)]+12} and very slightly in the

911: CNO ratios (A(C/H)=8.56, A(N/H)=8.05, A(O/H)=8.93 in \citet{and89} vs

912: A(C/H)=8.55, A(N/H)=7.97, A(O/H)=8.87 in \citet{gre93}). Nevertheless, we

913: have also listed in Table~1, in parentheses, the measured abundances with respect to

914: the solar Fe values from \citet{and89}. Si and Mg are the same in all

915: evolutionary models, and have been only slightly revised downward

916: ($\sim$0.05~dex) by \citet{asp05}. Thus, the reader should note that current

917: discussions found in the literature on the

918: derived $\alpha$-elements vs. Fe ratio may depend critically

919: on the assumed Fe solar abundance.

920: 

921: \begin{figure*}

922: \epsscale{1.0}

923: \vspace{-.1cm}

924: \plottwo{fig6a.ps}{fig6b.ps}

925: \caption{\label{fig:mg}

926: Influence of Ly$\beta$ fluorescence and choice of \MgII\  model atom on the

927:  \MgII\ K-band lines.}

928: \end{figure*}

929: 

930: {\bf{Iron.}} Two types of \FeII\ lines are found in the spectra

931: \footnote{The forbidden [\FeII] 1.677\um\ line present in the Pistol Star is

932: not included in our models}.

933: The first

934: are the strong semi-forbidden lines, including

935: z$^{4}$F$_{9/2}$-c$^{4}$F$_{9/2}$ 1.688~\um\ and

936: z$^{4}$F$_{3/2}$-c$^{4}$F$_{3/2}$~2.089~\um, that form in the outer wind and

937: have small oscillator strengths (gf$\sim$10$^{-5}$). The second are the

938: weak permitted (gf$\sim$1) lines connecting higher lying levels, such as the

939: 4de$^{6}$G-5p$^{6}$F lines near 1.733~\um\ or 6p$^6$D-6s$^6$D at 2.109~\um,

940: that form much closer to the photosphere.

941: 

942: 

943: The permitted lines are more robust iron abundance indicators, having only

944: weak dependences on other parameters, such us turbulent velocity. The

945: strengths of the semi-forbidden lines depend on the accuracy of their

946: weak gf values, the mass loss rate and the

947: run of the iron ionization structure in the outer wind, which is

948: sensitive to the hydrogen ionization structure due to the strong coupling to

949: the Fe/H charge-exchange reactions. Since a change in the run of the 

950: clumping factor in the outer wind regions modifies the ratio of

951: recombinations/ionizations in hydrogen, the semi-forbidden lines are

952: diagnostic of the behavior of clumping there.

953: From Fig.~\ref{fig:pistol} it

954: can be seen that that our model is able to simultaneously reproduce both

955: sets of lines, providing constraints on both clumping and abundance.\footnote{We think

956: that the slight missmatch in the \FeII] 2.117~\um\ line in both objects is

957: related to the accuracy of the f value.}

958: 

959: 

960: 

961: 

962: We obtain roughly solar iron abundances for both LBVs, with $\pm$0.15dex as

963: plausible uncertainties (see Fig.~\ref{fig:lbv_metal}).

964: Our results are similar to A(Fe/H)=7.59 recently

965: derived by \citet{cun07} from their analysis of a sample of luminous cool

966: stars within 30~pc of the Galactic Center. Note in Table~1 that the Fe

967: abundance ratio has significant uncertainty due to the uncertainty in the Fe

968: abundance in the Sun.

969: 

970: {\bf{Magnesium.}} The strongest \MgII\ lines observed in the H and K bands

971: share the 5p$^2$P level. Those lines with it as the upper level, the

972: 2.13/14~\um\ and 2.40/41~\um\ doublets (see Figs~\ref{fig:pistol} and

973: \ref{fig:362}) are much stronger than those with it as the lower level (H

974: band lines), revealing that pumping through the resonance 3s$^2$S-5p$^2$P

975: line must be a significant populator of the 5p$^2$P levels. Pumping through

976: the 3s$^2$S$[1/2]$-5p$^2$P$[3/2]$ 1025.968\AA\ transition is very efficient

977: due to Ly$\beta$ fluorescence. This was confirmed in models in which we

978: decoupled the 5p$^2$P$[3/2]$ and 5p$^2$P$[1/2]$ levels (see

979: Fig.~\ref{fig:mg}), resulting in \MgII\ 2.13/14~\um\ ratios much higher than

980: observed.

981: 

982: The relevance of this process can be easily followed in Fig.~\ref{fig:mg},

983: which displays the behavior of the doublet as a function of the choice of

984: the \MgII\ atom and the turbulent velocity. The latter refers to the fixed

985: Doppler width used in our models to compute the level populations. In the

986: left panel of Fig.~\ref{fig:mg} the levels are considered to be decoupled

987: (\ie\, the number of superlevels in the model atom, NS, is set to the total

988: number of levels in the full atom, NF). Because Ly$\beta$ lies closer to

989: 3s$^2$S$[1/2]$-5p$^2$P$[3/2]$ ($\Delta$v=72\kms) than to

990: 3s$^2$S$[1/2]$-5p$^2$P$[1/2]$ ($\Delta$v=114\kms), and because in these LBVS

991: the terminal velocities and wind densities determine the line formation

992: zones), only the 5p$^2$P$[3/2]$ is pumped thru fluorescence. Indeed, it can

993: be seen that as the turbulence velocity is increased, the overlap between

994: Ly$\beta$ and the \MgII\ line increases, as does the population of the

995: 5p$^2$P$[3/2]$ level and the strength of the \MgII\ 2.13~\um\ line

996: increases, while the longer wavelength \MgII\ 2.14~\um\ line is unaffected.  

997: On the other hand, if the \MgII\ model atom has both levels combined into a

998: superlevel (NS$\neq$NF, Fig.~\ref{fig:mg}{-{right}}), the observed ratio is

999: reproduced. Furthermore, increasing the turbulent velocity, and hence the

1000: effect of fluorescence, increases the pumping of both levels equally and

1001: thus increases the strength of the doublet with a constant ratio between its

1002: components. From Fig.~\ref{fig:mg} it can be seen that our assumed collision

1003: coefficients connecting the \MgII\ 5p$^2$P$[1/2]$ and 5p$^2$P$[3/2]$ levels

1004: may be too low. This comparison illustrates the importance of making the

1005: correct choice of model atoms for quantitative spectroscopic analysis.

1006: 

1007: Due to fluorescence coupling, the \MgII\ K-band lines show a stronger

1008: dependence on turbulent velocity than do the H-Band lines. We estimate about

1009: twice solar Mg abundance and an associated uncertainty 

1010: (see Fig.~\ref{fig:lbv_metal}) of about $\pm$0.25dex

1011: (due to uncertainties related to the fluorescence contribution).

1012: 

1013: \begin{figure}

1014: %\epsscale{1.15}    %% emulateapj

1015: \epsscale{0.90}     %% submission

1016: \vspace{-.1cm}

1017: \plotone{fig7a.ps}

1018: \vspace{-.3cm}

1019: \plotone{fig7b.ps}

1020: \vspace{-.3cm}

1021: \plotone{fig7c.ps}

1022: \caption{\label{fig:lbv_metal}

1023: Error estimates of Fe (upper-panel), Mg (middle-panel) and Si (lower-panel) abundandes. Dashed lines (red) correspond to

1024: our best model fitting the observed (black-solid) diagnostic lines of 

1025: \FMM362. Long-dashed (green) and dashed-dotted (blue) lines correspond to models

1026: where individual metal abundances have been set to the derived upper and lower estimates

1027: respectively.}

1028: \end{figure}

1029: 

1030: 

1031: 

1032: {\bf{Silicon.}} The \SiII\ doublet

1033: 5s$^{2}$S$_{1/2}$-5p$^{2}$P$_{3/2}$~1.691~\um\ and

1034: 5s$^{2}$S$_{1/2}$-5p$^{2}$P$_{1/2}$~1.698~\um\ constitutes a powerful

1035: diagnostic tool, as it appears in emission for only a very narrow range of

1036: stellar temperatures and wind density structures, indicating the presence of

1037: amplified NLTE effects. However, since it forms at the base of the wind, its

1038: strong dependence on the details of the velocity field there hinders a

1039: precise silicon abundance determination.

1040: 

1041: Instead, we use the well-behaved recombination line \SiII\

1042: 3s$^2$6g$^2$G-3s$^2$5f$^2$F~at~1.718~\um\ which shows a stronger dependence

1043: on the silicon abundance. Once again, a realistic mapping of full- to

1044: super-levels in our model atom is required. From our model fits (see

1045: Figs.~\ref{fig:pistol}~and~\ref{fig:362}) we derive roughly twice solar

1046: abundance ($\pm$0.20~dex) for silicon in each LBV, similar to magnesium 

1047: (see Fig.~\ref{fig:lbv_metal}).

1048: 

1049: {\bf{Other elements.}} One might expect a number of oxygen lines might be

1050: detectable in infrared spectra of LBVs: i.e., strong \OI\ lines at

1051: 2.763~\um, 2.893~\um\ and 3.098~\um\ and weaker lines at 1.8243~\um,

1052: 3.661~\um\ and 3.946~\um. Several of these, but not all, are problematical

1053: from ground-based observatories. Unfortunately our data set only encompasses

1054: the \OI~1.745~\um\ line which is blended with a stronger \MgII\ line. Thus,

1055: we defer an attempt to estimate the oxygen abundance until high resolution

1056: observations of unblended lines can be obtained. Determining the oxygen

1057: abundances in these objects will provide crucial constraints on their

1058: evolutionary status.  Models show that when H/He$<$1.50 oxygen has reached

1059: its maximum depletion within CNO equilibrium, while a significantly higher O

1060: content should be present on the stellar surface for H/He values around 3.

1061: The results in Table~1 then predict that the Pistol Star and \FMM362 have

1062: different oxygen abundances. On the other hand, if an LBV has a H/He$<$1.50

1063: but is still hydrogen rich, the oxygen abundance determination, expected to

1064: be $\sim$0.04 of the original value, will provide a measure of the

1065: metallicity of the natal cloud.  High resolution L-Band spectra of the

1066: Pistol Star should be able to address this issue.

1067: 

1068: The only detected sodium lines are the well-known doublet at 2.206/9~\um,

1069: from which we obtain a very high abundance, $\sim 20\times$solar. The strong

1070: observed emission of this doublet in the K-band spectra of other LBVs has

1071: been noted previously by \citet{hil98b}. Interestingly, our models display

1072: only a minor dependence of these lines on clumping. On the other hand, the

1073: strengths of the sodium lines might not indicate extraordinary sodium

1074: abundance if the lines are produced by fluorescence of circumstellar

1075: material, a component that we do not model.

1076: 

1077: {\section{Discussion.}}

1078: 

1079: Our results suggest solar Fe abundances and approximately twice-solar

1080: $\alpha$-element abundances for the Quintuplet LBVs.  Presumably, these

1081: abundances were the same in the gas that condensed to form these stars and

1082: the other stars in the Quintuplet cluster and indeed in the whole of the

1083: present-day Galactic center. The results can be discussed in the context of

1084: similar measurements of Galactic center objects and with respect to the

1085: trend one might expect if the region is an inward extension of the

1086: disk or the bulge. In addition, the ratio of Fe to $\alpha$-elements might

1087: be used to decipher the star formation history in the Galactic center.

1088: 

1089: \input table2.tex

1090: %\input table2_apj.tex

1091: 

1092: Table~2 displays a number of recent determinations of stellar 

1093: metal abundances in the

1094: Galactic Center together with the above mentioned three reference patterns

1095: for solar abundances.  The values derived for Fe abundances in cool stars

1096: agree with our result \citep{car00,ram97,ram99,ram00}.  \citet{cun07} find a

1097: very narrow range of Fe abundances clustered around the solar value for a

1098: population of cool stars in the central 30~pc. Of particular interest is

1099: star VR5-7 from \citet{cun07} sample which is located in the Quintuplet

1100: Cluster and shows A(Fe/H)=7.60 and A(Ca/H)=6.41

1101: 

1102: There are relatively few measurements of the $\alpha$-element abundances

1103: ([$\alpha$/Fe]) in GC stars.  \citet{naj04} find solar abundances (as

1104: defined in this paper) for hot

1105: stars in the Arches cluster based on the oxygen abundance and, to a lesser

1106: degree, carbon abundance, and adopting the canonical solar value of

1107: A(O/H)=8.93 \citep{and89}. Those estimates assume that nitrogen has reached

1108: its maximum surface abundance value. Evolutionary models indicate that 95\%

1109: of that value is already attained by the time that H/He$<$2 (by number).  

1110: \citet{naj04} followed the metallicity patterns from the Geneva evolutionary

1111: models and assumed no selective enrichment of CNO or $\alpha$-elements vs

1112: Fe, in concluding that the stars in the Arches Cluster have solar

1113: $\alpha$-element abundances. However, estimates of solar abundances have

1114: varied considerably over the past 15 years

1115: \citep[e.g.][, see also Table~2]{alle08}.

1116: Thus, depending on

1117: the assumed solar CNO composition, the derived nitrogen abundance by

1118: \citet{naj04} could imply solar \citep[][]{and89}, 1.2~$\times$~solar

1119: \citep[][]{gre93} or 2.0~$\times$~~solar \citep[][]{asp05} CNO composition.

1120: 

1121: Recently \citet{mar07a,mar07b} have analyzed a larger sample of hot stars in

1122: the Arches and Central Parsec clusters and find similar results (see

1123: Table~2). Interestingly, if one considers only the objects in \citet{mar07b}

1124: with He/H$>0.1$ and those with Z(C)$<0.05$, \ie\ fulfilling the condition to

1125: be close enough to Z(N)$_{max}$, the average value of Z(N) is 1.7.

1126: 

1127: \citet{geb06} estimate roughly solar oxygen abundance, A(O/H)=8.91, in IRS~8,

1128: an OIf supergiant near the central parsec.  \citet{cun07} find

1129: $<$A(O/H)$>$=9.04 ([0.37]) and $<$A(Fe/H)$>$=7.59 ([0.14]) for their sample

1130: of cool stars, where the numbers in brackets are the ratio with respect to

1131: the solar value in dex.  This implies [O/Fe]=0.22, i.e.\ a clear enhancement

1132: over the solar ratio. Again, the \citet{cun07} measurements could be

1133: interpreted as indicating solar ratios in O over Fe if the solar O abundance

1134: in evolutionary models is used.  

1135: It is crucial to have

1136: accurate solar abundances, and that values used in stellar

1137: evolution calculations should be consistent with these.

1138: An excellent example is

1139: attempting to determine whether the possible oxygen enhancement is due to a

1140: top-heavy IMF favoring $\alpha$-elements vs Fe enrichment, or simply an

1141: overall CNO and metal enhancement. Thus, taking the CNO abundances for the

1142: GC objects from \citet{cun07} and assuming C/N equilibrium values one can

1143: interprete their results either as solar CNO with mildly enhanced (30\%)

1144: oxygen \citep{and89} or a clearly supersolar environment with a factor of

1145: 1.7 enhancement for C and N and 2.5 for oxygen \citep{asp05}.

1146: 

1147: 

1148: Fortunately, there are other $\alpha$-elements whose adopted solar

1149: abundances have suffered basically no major revision.  Thus, we believe that

1150: the enhanced values obtained in this work for Mg and Si, roughly a factor of

1151: two solar, together with the enhancement of Ca found by \citet{cun07}, are a

1152: strong indication of the enrichment of $\alpha$-elements compared to Fe.

1153: 

1154: Our results run counter to the trend in the disk \citep{roll00,sma01,mar03},

1155: and are more consistent with the values found for the bulge

1156: \citep{fro99,fel00}. This may imply that the ISM in the disk does not extend

1157: inward to the GC, so that material is dragged into the central molecular

1158: zone from the bulge rather than from the disk. Another possibility is that

1159: the GC stars are forming out of an ISM that has an enrichment history

1160: distinctly different from that of the disk. At this point, further studies

1161: of the $\alpha$-elements vs Fe would be useful. Future high S/N and high

1162: resolution spectroscopy of the \OI\ lines in LBVs and K-band spectroscopy of

1163: WNL stars in the same cluster (Najarro et al. in prep.) will provide two

1164: independent

1165: measurements of the original oxygen content, and thus set definite

1166: constraints on metallicity.

1167: 

1168: The modest enrichment in $\alpha$-elements versus Fe that we find in the

1169: two Quintuplet LBVs is consistent with a top-heavy IMF in the GC

1170: \citep{fig99a}. In such a scenario, enhanced yields of $\alpha$-elements

1171: compared to Fe are expected through a higher than average ratio of the

1172: number of SNII vs SNIa events \citep{whe89,cun07}. 

1173: 

1174: 

1175: \acknowledgements

1176: 

1177: We thank Fabrice Martins and Katia Cuhna for usefull discussions.  F.~N. acknowledges

1178: AYA2004-08271-C02-02 and AYA2007-67456-C02-02 grants. The material in this paper is based on work

1179: supported by NASA under award NNG 05-GC37G, through the Long Term Space

1180: Astrophysics program. TRG's research is supported by the Gemini Observatory,

1181: which is operated by the Association of Universities for Research in

1182: Astronomy, Inc., on behalf of the international Gemini partnership of

1183: Argentina, Australia, Brazil, Canada, Chile, the United Kingdom, and the

1184: United States of America. D. Figer is supported by a NYSTAR Faculty

1185: Development Program grant.

1186: 

1187: 

1188: %\small

1189: %\clearpage

1190: \begin{thebibliography}{}

1191: \bibitem[Afflerbach et al.(1997)]{affler97} Afflerbach, A.,  Churchwell, E.~B. \&\ Werner, M.W., 1997, \apj, 478, 190

1192: \bibitem[Allende Prieto(2008)]{alle08} Allende Prieto, C.\ 2008, 14th Cambridge Workshop on Cool Stars, Stellar Systems, and the Sun, 384, 39 

1193: \bibitem[Anders \& Grevesse(1989)]{and89} Anders, E. \&\ Grevesse, N., 1989, Geochimica et  Cosmochimica Acta, Vol. 53, 197

1194: \bibitem[Asplund et al.(2005)]{asp05} Asplund, M., Grevesse, N., \& Sauval, A.~J.\ 2005, Cosmic Abundances as Records of Stellar Evolution and Nucleosynthesis, ASPC, 336, 25 

1195: %\bibitem[Blum et al.(2001)]{blu01} Blum, R.\ D., Schaerer, D., Pasquali, A., Heydari-Malayeri, M., Conti, P.\ S., Schmutz, W. 2001, \aj, accepted

1196: \bibitem[Carr et al.(2000)]{car00} Carr, J.S., Sellgren, K. \& Balachandran, S.C. 2000, \apj, 530, 307

1197: \bibitem[Charbonnel et al.(1993)]{charb93} Charbonnel, C., Meynet, G., Maeder, A., Schaller, G., \& Schaerer, D.\ 1993, \aaps, 101, 415 

1198: \bibitem[Cunha et al.(2007)]{cun07} Cunha, K., Sellgren, K., Smith, V.~V., Ramirez, S.~V., Blum, R.~D., \& Terndrup, D.~M.\ 2007, ArXiv e-prints, 707, arXiv:0707.2610 

1199: %\bibitem[Chiosi \& Maeder(1986)]{chi86} Chiosi, C.~\& Maeder, A.\ 1986, \araa, 24, 329 

1200: %\bibitem[Cotera et al.(1992)]{cot92} Cotera, A.~S., Erickson, E.~F., Simpson, J.~P., Colgan, S.~W.~J., Allen, D.~A., \& Burton, M.~G.\ 1992, American Astronomical Society Meeting, 181, 8702

1201: \bibitem[Cotera et al.(1994)]{cot94} Cotera, A.~S., Erickson, E.~F., Allen, D.~A., Colgan, S.~W.~J., Simpson, J.~P., \& Burton, M.~G.\ 1994, NATO ASIC Proc.~445: The Nuclei of Normal Galaxies: Lessons from the Galactic Center, 217 

1202: \bibitem[Crowther et al.(2006)]{cro06} Crowther, P.~A., Lennon, D.~J., \& Walborn, N.~R.\ 2006, \aap, 446, 279

1203: %\bibitem[Cotera(1995)]{cot95} Cotera, A.\ S.\ 1995, Ph.D.\ Thesis, Stanford University

1204: %\bibitem[Cotera et al.(1996)]{cot96} Cotera, A. S., Erickson, E. F., Colgan, S. W. J., Simpson, J. P., Allen, D. A., \& Burton, M. G. 1996, \apj, 461, 750

1205: \bibitem[Feltzing \& Gilmore(2000)]{fel00} Feltzing, S., \& Gilmore, G.\ 2000, \aap, 355, 949

1206: \bibitem[Figer, McLean, \& Morris(1995)]{fig95} Figer, D.~F., McLean, I.~S., \& Morris, M.\ 1995, \apjl, 447, L29

1207: \bibitem[Figer et al.(1998)]{fig98} Figer, D. F., Najarro, F., Morris, M., McLean, I. S., Geballe, T. R., Ghez, A. M., \& Langer, N. 1998, \apj, 506, 384

1208: \bibitem[Figer, McLean, \& Najarro(1997)]{fig97} Figer, D.~F., McLean, I.~S., \& Najarro, F.\ 1997, \apj, 486, 420 

1209: \bibitem[Figer et al.(1999a)]{fig99a} Figer, D.\ F., Kim, S.\ S., Morris, M., Serabyn, E., Rich, R.\ M., \& McLean, I.\ S.\ 1999a, \apj, 525, 750

1210: \bibitem[Figer et al.(1999b)]{fig99b} Figer, D. F., McLean, I. S., \& Morris, M. 1999, \apj, 514, 202 

1211: \bibitem[Figer et al.(1999c)]{fig99c} Figer, D.\ F., Morris, M., Geballe, T.\ R., Rich, R.\ M., Serabyn, E., McLean, I.\ S., Puetter, R.\ C., \& Yahil, A.\ 1999b, \apj, 525, 759 

1212: \bibitem[Figer et al.(2002)]{fig02} Figer, D.~F., et al.\ 2002, \apj, 581, 258 

1213: \bibitem[Frogel et al.(1999)]{fro99} Frogel, J.A., Tiede, G.P., \& Kuchinski, L.E.\ 1999, \aj, 117, 2296

1214: \bibitem[Fuhrmann (1998)]{fuhr98} Fuhrmann, K., 1998, \aap, 338, 161

1215: \bibitem[Geballe, Najarro, \& Figer (2000)]{geb00} Geballe, T.~R., Najarro, F., \& Figer, D.~F.\ 2000, \apjl, 530, L97 

1216: \bibitem[Geballe et al.(2006)]{geb06} Geballe, T.~R., Najarro, F., Rigaut, F., \& Roy, J.-R.\ 2006, \apj, 652, 370 

1217: \bibitem[Girardi et al.(2000)]{gir00} Girardi, L., Bressan, A., Bertelli, G., \& Chiosi, C.\ 2000, \aaps, 141, 371 

1218: \bibitem[Glass et al.(1987)]{gla87} Glass, I.~S., Catchpole, R.~M., \& Whitelock, P.~A.\ 1987, \mnras, 227, 373 

1219: \bibitem[Glass et al.(1990)]{gla90} Glass, I.~S., Moneti, A., \& Moorwood, A.~F.~M.\ 1990, \mnras, 242, 55P 

1220: \bibitem[Glass et al.(1999)]{gla99} Glass, I.~S., Matsumoto, S., Carter, B.~S., \& Sekiguchi, K.\ 1999, \mnras, 304, L10

1221: \bibitem[Grevesse \& Noels(1993)]{gre93} Grevesse, N., \& Noels, A.\ 1993, Origin and evolution of the elements: proceedings of a symposium in honour of H.~Reeves, held in Paris, June 22-25, 1992.~Edited by N.~Prantzos, E.~Vangioni-Flam and M.~Casse.~Published by Cambridge University Press, Cambridge, England, 1993, p.14, 14 

1222: \bibitem[Grevesse \& Sauval(1998)]{gre98} Grevesse, N., \& Sauval, A.~J.\ 1998, Space Science Reviews, 85, 161 

1223: \bibitem[Groh et al.(2006)]{gro06} Groh, J.~H., Hillier, D.~J., \& Damineli, A.\ 2006, \apjl, 638, L33

1224: \bibitem[Herald et al.(2001)]{her01} Herald, J.~E., Hillier, D.~J., \& Schulte-Ladbeck, R.~E.\ 2001, \apj, 548, 932 

1225: \bibitem[Hillier \& Miller(1998)]{hil98} Hillier, D.~J.~\& Miller, D.~L.\ 1998, \apj, 496, 407

1226: \bibitem[Hillier et al.(1998)]{hil98b}  Hillier, D.J., Crowther, P.A., Najarro, F., Fullerton, A.W., 1998b, \aap, 340, 483

1227: \bibitem[Hillier \& Miller(1999)]{hil99} Hillier, D.~J.~\& Miller, D.~L.\ 1999, \apj, 519, 354

1228: \bibitem[Hillier et al.(2001)]{hil01} Hillier, D.~J., Davidson, K., Ishibashi, K., \& Gull, T.\ 2001, \apj, 553, 837

1229: \bibitem[Iglesias \& Rogers(1996)]{igl96} Iglesias, C.~A., \& Rogers, F.~J.\ 1996, \apj, 464, 943 

1230: \bibitem[Kennicutt et. al.(2003)]{kenni03} Kennicutt, R.C., Bresolin, F., Garnett, D.R. 2003, \apj, 591, 544

1231: \bibitem[Kudritzki \& Puls(2000)]{kud00} Kudritzki, R.-P., \& Puls, J.\ 2000, \araa, 38, 613 

1232: \bibitem[Luck et al.(2006)]{luck06} Luck, R.~E., Kovtyukh, V.~V., \& Andrievsky, S.~M.\ 2006, \aj, 132, 902 

1233: \bibitem[Maciel \&\ Quireza (1999)]{mac99} Maciel, W.J., \&\ Quireza, C. \ 1999, \aap, 345, 629

1234: \bibitem[Maeda et al.(2000)]{mae00} Maeda, Y., et. al.,  2002, \apj, 570, 671

1235: \bibitem[Mart{\'{i}}n-Hern\'andez et al.(2003)]{mar03} Mart{\'{i}}n-Hern\'andez, N.L., van der Hulst, J.M., \&\ Tielens, A.G.G.M.,  2003, \aap, 407, 957

1236: \bibitem[Martins et al.(2007a)]{mar07a} Martins, F., Genzel, R., Hillier, D.~J., Eisenhauer, F., Paumard, T., Gillessen, S., Ott, T., \& Trippe, S.\ 2007, \aap, 468, 233

1237: \bibitem[Martins et al.(2007b)]{mar07b} Martins, F., Hillier, D.~J., et al., 2007, \aap, (submitted)

1238: %\bibitem[McLean et al.(1998)]{mcl98} McLean et al.\ 1998, SPIE Vol. 3354, 566

1239: \bibitem[Meynet et al.(1994)]{mey94} Meynet, G., Maeder, A., Schaller, G., Schaerer, D., \& Charbonnel, C. 1994, \aap\ Supp., 103, 97

1240: \bibitem[Meynet \& Maeder(2003)]{mey03} Meynet, G., \& Maeder, A.\ 2003, \aap, 404, 975 

1241: \bibitem[Meynet \& Maeder(2005)]{mey05} Meynet, G., \& Maeder, A.\ 2005, \aap, 429, 581

1242: %\bibitem[Meynet \&\ Maeder(2004)]{mey04} Meynet, G.,\&| Maeder, A., 2004, \aap, 404, 975

1243: \bibitem[Moneti et al.(1994)]{mon94} Moneti, A., Glass, I. S. \& Moorwood, A. F. M. 1994, \mnras, 268, 194

1244: \bibitem[Nagata et al.(1990)]{nag90} Nagata, T., Woodward, C.~E., Shure, M., Pipher, J.~L., \& Okuda, H.\ 1990, \apj, 351, 83 

1245: %\bibitem[Nagata et al.(1995)]{nag95} Nagata, T., Woodward, C.\ E., Shure, M., \& Kobayashi, N.\ 1995, \aj, 109, 1676 

1246: \bibitem[Najarro et al.(1999)]{naj99} Najarro, F., Hillier, D.~J., Figer, D.~F., \& Geballe, T.~R.\ 1999, The Central Parsecs of the Galaxy, ASPC, 186, 340 

1247: \bibitem[Najarro (2001)]{naj01} Najarro, F.\ 2001, P Cygni  2000: 400 Years of Progress, ASPC, 233, 133 

1248: \bibitem[Najarro et al.(2004)]{naj04} Najarro, F., Figer, D. F., Hillier, D. J., \& Kudritzki, R. P.\ 2004 (Paper I), \apjl, 611, L105 

1249: %\bibitem[][Paper~I]{naj04} Najarro, F., Figer, D. F., Hillier, D. J., \& Kudritzki, R. P.\ 2004 (Paper I), \apjl, 611, L105 

1250: \bibitem[Nugis et al.(1998)]{nug98} Nugis, T., Crowther, P.~A., \& Willis, A.~J.\ 1998, \aap, 333, 956

1251: \bibitem[Okuda et al.(1990)]{oku90} Okuda, H., et al.\ 1990, \apj, 351, 89 

1252: \bibitem[Pinsonneault \& Delahaye(2006)]{pin06} Pinsonneault, M.~H., \& Delahaye, F.\ 2006, ArXiv Astrophysics e-prints, arXiv:astro-ph/0606077 

1253: \bibitem[Plez(1992)]{ple92} Plez, B.\ 1992, \aaps, 94, 527

1254: \bibitem[Puls et al.(2006)]{pul06} Puls, J., Markova, N., Scuderi, S., Stanghellini, C., Taranova, O.~G., Burnley, A.~W., \& Howarth, I.~D.\ 2006, \aap, 454, 625

1255: \bibitem[Ramirez et al.(1997)]{ram97} Ramirez, S.V., Carr, J.S., Balachandran, S., Blum, R., \& Terndrup, D.M.\ 1997, in IAU Symp. 184, The Central Regions of the Galaxy and Galaxies, ed. Y. Sofue (Dordrecht: Kluwer), 28

1256: \bibitem[Ramirez et al.(1999)]{ram99} Ramirez, S.V., Sellgren, K., Carr, J.S., Balachandran, S., Blum, R., \& Terndrup, D.M.\ 1999, ASP Conf.~Ser.~186: The Central Parsecs of the Galaxy

1257: \bibitem[Ramirez et al.(2000)]{ram00} Ramirez, S.V., Sellgren, K., Carr, J.S., Balachandran, S., Blum, R., Terndrup, D.M., \&\ Steed, A., 2000, \apj, 537,205

1258: \bibitem[Rolleston et. al.(2000)]{roll00} Rolleston, W.R., Smartt, S.J., Dufton, P.L., \&\ Ryans, R.S.I., 2000, \aap, 363, 537

1259: \bibitem[Runacres \& Owocki(2002)]{run02} Runacres, M.~C.,\& Owocki, S.~P.\ 2002, \aap, 381, 1015 

1260: \bibitem[Rudolph et al.(2006)]{rud06} Rudolph, A.~L., Fich, M., Bell, G.~R., Norsen, T., Simpson, J.~P., Haas, M.~R., \& Erickson, E.~F.\ 2006, \apjs, 162, 346 

1261: \bibitem[Salasnich et al.(2000)]{sal00} Salasnich, B., Girardi, L., Weiss, A., \& Chiosi, C.\ 2000, \aap, 361, 1023 

1262: \bibitem[Schaller et al.(1992)]{scha92} Schaller, G., Schaerer, D., Meynet, G., \& Maeder, A.\ 1992, \aaps, 96, 269

1263: \bibitem[Shields \&\ Ferland (1994)]{shi94}Shields, J.C.,  \&\ Ferland, G.J., 1994, \apj, 430, 236

1264: \bibitem[Smartt et. al.(2001)]{sma01} Smartt, S.J., Venn, K.A.,  Dufton, P.L., Lennon, D.J., Rolleston, W.R.,  \&\ Keenan, F.P., 2001, \aap, 367, 86

1265: \bibitem[Stahl et al.(2001)]{sta01} Stahl, O., Jankovics, I., Kov{\'a}cs, J., Wolf, B., Schmutz, W., Kaufer, A., Rivinius, T., \& Szeifert, T.\ 2001, \aap, 375, 54 

1266: \bibitem[Urbaneja et al.(2005)]{urba05} Urbaneja, M.A., Herrero, A., Bresolin, F., Kudritzki, R.P., Gieren, W., Puls, J., Przybilla, N., Najarro, F., Pietrzynski, G., 2005, \apj, 622, 862

1267: %\bibitem[Vink et al.(2000)]{vin00} Vink, J.~S., de Koter, A., \& Lamers, H.~J.~G.~L.~M.\ 2000, \aap, 362, 295 

1268: \bibitem[Vink \& de Koter(2002)]{vin02} Vink, J.~S., \& de Koter, A.\ 2002, \aap, 393, 543 

1269: \bibitem[Wheeler et al.(1989)]{whe89} Wheeler, J.~C., Sneden, C., \& Truran, J.~W., Jr.\ 1989, \araa, 27, 279

1270: \end{thebibliography}

1271: 

1272: 

1273: 

1274: \end{document}

1275: 

1276: 

1277: