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

2: 

3: %\usepackage{graphics}

4: %% manuscript produces a one-column, double-spaced document:

5: % \documentclass[manuscript]{aastex}

6: %% preprint2 produces a double-column, single-spaced document:

7: % \documentclass[preprint2]{aastex}

8: 

9: 

10: \newcommand{\vdag}{(v)^\dagger}

11: \newcommand{\myemail}{jgrimes@pha.jhu.edu}

12: \newcommand{\ch}{{\it Chandra}}

13: \newcommand{\xmm}{{\it XMM-Newton}}

14: \newcommand{\fu}{{\it FUSE }}

15: \newcommand{\ulg}{ULIRGs}

16: \newcommand{\ul}{ULIRG}

17: \newcommand{\cps} {counts s$^{-1}$}

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

19: \newcommand{\ergcms} {erg cm$^{-2}$ s$^{-1}$}

20: \newcommand{\chsq}{$\chi^2$}

21: \newcommand{\feka}{$Fe-K\alpha$}

22: \newcommand{\al}{$\alpha$}

23: \newcommand{\afe}{$\alpha/\rm{Fe}$}

24: \newcommand{\alphe}{$\alpha$-element}

25: \newcommand{\ha}{$\rm{H}\-\alpha$}

26: \newcommand{\lya}{$\rm{Ly}\-\alpha$}

27: \newcommand{\ang}{\rm{\AA} }

28: \newcommand{\vv}{VV 114}

29: \newcommand{\lfir}{$\rm{L_{FIR}}$}

30: \newcommand{\ovi}{\ion{O}{6}}

31: \newcommand{\hoo}{Haro\,11}

32: \newcommand{\ks}{\rm{km\,s^{-1}}}

33: \newcommand{\kms}{$\rm{km\,s^{-1}}$}

34: 

35: 

36: 

37: 

38: 

39: \shorttitle{\fu\,\&\ch\,Observations of Haro 11}

40: \shortauthors{J Grimes et al.}

41: 

42: \slugcomment{Accepted for Publication in ApJ}

43: 

44: \begin{document}

45: 

46: \title{Feedback in the local LBG Analog Haro 11 as probed by far-UV and X-ray observations}

47: 

48: \author{J. P. Grimes\altaffilmark{1}, T. Heckman\altaffilmark{1}, 

49: D. Strickland\altaffilmark{1}, W. V. Dixon\altaffilmark{1},

50: K. Sembach\altaffilmark{2},

51: R. Overzier\altaffilmark{1}, C. Hoopes\altaffilmark{1},

52: A. Aloisi\altaffilmark{2}, and A. Ptak\altaffilmark{1}}

53: \altaffiltext{1}{Center for Astrophysical Sciences, Johns Hopkins University,

54:     3400 N. Charles St, Baltimore, MD 21218; jgrimes@pha.jhu.edu, heckman@pha.jhu.edu,

55:     dks@pha.jhu.edu,ptak@pha.jhu.edu}

56: \altaffiltext{2}{Space Telescope Science Institute, 3700 San Martin Drive,

57:    Baltimore, MD, 21218; sembach@stsci.edu,aloisi@stsci.edu}

58: 

59: 

60: \begin{abstract}

61: 

62: We have re-analyzed \fu\, data and obtained new \ch~observations of \hoo, 

63: a local ($\rm{D_L}$=88 Mpc) UV luminous galaxy.  

64: \hoo\, has a similar  far-UV luminosity ($10^{10.3}\,\rm{L}_{\odot}$), UV surface 

65: brightness ($10^{9.4}\,\rm{L_{\odot}\,kpc^{-2}}$), SFR, and metallicity to that observed

66: in Lyman Break Galaxies (LBGs).  We show that \hoo\, has extended, soft thermal 

67: ($\rm{kT\sim0.68\,keV}$) X-ray emission with a luminosity and size which scales with 

68: the physical properties (e.g. SFR, stellar mass) of the host galaxy. 

69:  An enhanced \afe\, ratio of $\sim4$ relative to solar abundance

70: suggests significant supernovae enrichment.  These results are consistent with

71: the X-ray emission being produced in a shock between a supernovae driven outflow

72: and the ambient material.

73: The FUV spectra show strong absorption lines similar to those observed in LBG spectra.

74: A blueshifted absorption component is identified as a wind outflowing at $\sim200-280$ \kms.

75: \ion{O}{6}$\lambda\lambda$1032,1038 emission, the dominant cooling mechanism for coronal 

76: gas at $\rm{T\sim10^{5.5}K}$ is also observed.  

77: If associated with the outflow, the luminosity of the \ovi\, emission 

78: suggests that $\lesssim20\%$ of the total mechanical energy 

79: from the supernovae and solar winds is being radiated away.  

80: This implies that 

81: radiative cooling through \ovi\, is not significantly inhibiting the growth of

82: the outflowing gas.

83: In contradiction to the findings of \citet{berg06}, we find 

84: no convincing evidence of Lyman continuum leakage

85: in \hoo.  We conclude that the wind has not created a ``tunnel'' allowing the escape of a significant 

86: fraction of Lyman continuum photons and place a limit on the escape

87: fraction of $\rm{f_{esc}}\lesssim2\%$.

88: Overall, both \hoo\, and a previously observed LBG analogue \vv, provide

89: an invaluable insight into the X-ray and FUV properties of 

90: high redshift LBGs.

91: 

92: \end{abstract}

93: 

94: 

95: \keywords{ galaxies: starburst  ---

96:              galaxies: halos ---

97:              UV: galaxies ---

98:              galaxies: individual (Haro11)}

99: 

100: \section{Introduction}

101: 

102: 

103: Recent observations, headlined by the {\it WMAP} satellite, 

104: have ushered in a new era of ``precision cosmology'' \citep{benn03}.  

105: Measurements of the age, geometry, density, and 

106: composition of the universe have led to

107: a robust understanding of the evolution of its large scale structure

108: (dominated by dark matter and dark energy).   

109: However, significant problems in $\Lambda$CDM 

110: simulations emerge when examining

111: small scale structure and regions of high density where the complex range of baryonic physics becomes important \citep{klyp99,rob05,som99}.   

112: The frontier of cosmology is now in

113: understanding the physical processes involved in 

114: the interactions between the intergalactic medium (IGM) and the gas, stars,

115: and black holes that comprise galaxies.

116: 

117: 

118: At the forefront of

119: this endeavor is understanding the role of feedback with which a

120: galaxy mediates both its own evolution and that of the

121: next generation of galaxies.  Supernovae driven galactic scale winds

122: appear to be an essential piece of galaxy evolution \citep{veill05}.

123: In galaxies undergoing intense star formation, these winds 

124: drive gas, energy, and metals out of galaxies and possibly into the IGM \citep{heck98}. 

125: Although common in the local universe,

126: these winds were of particular importance in the past,

127: due to the significantly higher overall star formation rate (SFR) 

128: in the early universe \citep{bunk04}. 

129: 

130: Such winds at high redshift almost

131: certainly had a profound

132: impact on the energetics, ionization, and chemical composition

133: of the IGM \citep{ag05}.  For example, star-forming galaxies are thought to be

134: an important source of ionizing photons in the early universe

135: and could play a vital role in the reionization

136: of the IGM.  This effect should be

137: enhanced by winds as they could

138: clear paths through the neutral

139: gas and dust surrounding the galaxy,

140: allowing the ionizing photons to

141: escape into the IGM \citep{dov00}.  A reionized IGM

142: strongly impacts galaxy formation

143: as it suppresses gas cooling and subsequent infall into low mass

144: halos \citep{bark06}.

145: 

146: 

147: Multiwavelength observations are required to understand the complex multiphase

148: nature of galactic winds \citep{strick04b}. The coronal 

149: gas ($T \sim 10^5$ to $10^6$ K) and hot X-ray gas ($T \sim 10^6$ to $10^8$ K) in 

150: galactic winds are particularly important.  These phases

151: are intimately connected to the mechanical/thermal energy that drives the 

152: outflows and to the metals the outflows carry.

153: Far-ultraviolet (FUV) observations 

154: of the coronal gas provide important insights into the cooling 

155: and kinematics of the galactic winds \citep{heck01,hoop03}.  X-ray observations 

156: probe the slowly cooling hot gas that is most likely to escape the

157: galaxy \citep{strick00}.

158: 

159: The most widely studied population of high redshift star forming galaxies are 

160: the Lyman Break Galaxies \citep[LBGs,][]{steid99}.  These galaxies can be efficiently detected for 

161: $2\lesssim z \lesssim 6$ using

162: the Lyman-break technique, which picks out spectral band dropouts caused by the

163: 912 $\ang$ Lyman continuum discontinuity. They constitute a significant (and possibly dominant) fraction of the population of star forming galaxies during this important cosmic epoch \citep{peac00}.    The LBGs trace the most overdense regions of the universe, which are believed to be the progenitors of present day galaxy groups and clusters \citep{giav02,ouchi04}. They therefore play an important role in understanding the evolution of such clusters.  

164: Galactic winds appear to be a ubiquitous property of LBGs \citep{shap03}.

165: In particular, winds from LBGs could be the mechanism that heats

166: and chemically enriches the Intra-cluster Medium \citep[ICM, ][]{hels00,tam04}.

167: 

168: Unfortunately, our knowledge of the properties of the winds in LBGs is limited to what can be inferred from the interstellar UV absorption lines longward of Lyman $\alpha$. Because of their great distance, X-ray detections are almost impossible using

169: current X-ray observatories, although stacking techniques have been used to

170: construct luminosity weighted average X-ray spectra \citep{nan02,lehm05}.  

171: A more fundamental problem is that the large redshifts of the LBGs mean that soft X-ray observations with {\it Chandra} or \xmm ~are observing at rest-frame energies above a few keV. In local star-forming galaxies the emission in this band is dominated by the population of X-ray binaries, and the thermal emission from the galactic wind is negligible \citep{colb04}. Likewise, observations of the coronal phase gas is very difficult in LBGs. The most accessible probe of such gas is the  \ion{O}{6}~$\lambda\lambda$1032,1038 doublet, which lies deep within the Lyman $\alpha$ forest in the spectra of LBGs. 

172: 

173: It is clear that directly studying gas hotter than $10^5\,$K in high-redshift LBGs will continue to be very difficult or even impossible for the forseeable future.  Therefore, an important step in understanding galactic winds in LBGs would be to identify the best local analogs to LBGs and to then investigate their winds using the full suite of observations that are possible at low redshift.

174: 

175: Recently \citet{heck05} created a catalog of low redshift galaxies 

176: using {\it GALEX} \citep{mart05} UV observations of galaxies with spectra taken by the Sloan Digital Sky Survey \citep[SDSS, ][]{york00}. This catalog has now been significantly expanded based on additional {\it GALEX} data by Hoopes et al. (2006), who used this matched catalog to select a sample of extraordinarily rare local (z $<$ 0.3) galaxies having the same UV luminosities, sizes, and surface brightnesses as typical LBGs. They then showed that these galaxies have the same star formation rates, galaxy masses, velocity dispersions, and chemical compositions as typical LBGs. 

177: 

178: In the {\it GALEX} sample, \vv~   is the 

179: closest ($\rm{D_L\sim 86\, Mpc}$) known LBG analog.  We have previously 

180: analyzed X-ray and FUV observations of

181: \vv\, \citep{grim06}.  Diffuse thermal X-ray emission

182: encompassing \vv~was observed by \ch\, and \xmm.  This hot

183: ($\rm{kT}\sim 0.59~\rm{keV}$) gas has an enhanced

184: \afe~element ratio relative to solar abundances and follows the

185: same relations as typical starbursts between its properties (luminosity, size, and temperature) and those of the starburst galaxy

186: (SFR, dust temperature, galaxy mass). 

187: These results are consistent with the X-ray gas having been produced by shocks driven by a galactic superwind.

188: The \fu\, observations of \vv~show strong, broad interstellar absorption lines with a pronounced blueshifted component

189: (similar to what is seen in LBGs). This implies 

190: an outflow of material moving at $\sim300-400~\ks$ relative to \vv.  The properties 

191: of the strong \ovi~absorption line are consistent with radiative

192: cooling at the interface between the hot outrushing gas seen in X-rays and

193: the cooler material seen in the other outflowing ions in the \fu\, data.

194: We also showed that the wind in \vv~has {\it not} created a ``tunnel'' that would enable more than a 

195: small fraction ($<$ few percent) of the ionizing photons from \vv ~to escape into the IGM.

196: 

197: The next closest LBG analog in the {\it GALEX} sample is \hoo.  \hoo\, is particularly interesting,

198: as \citet{berg06} have claimed the detection of Lyman continuum radiation in the \fu~spectra.  

199: \citet{berg06} derived an escape fraction of $\rm{f_{esc}}\sim4-10\%$.

200: This would make it the only known galaxy with Lyman continuum emission in the local universe

201: and thus an ideal candidate to study this intriguing property of some high redshift LBGs \citep{shap06}.

202: 

203: %00^{\rm h}36^{\rm m}52\fs  and -33\degr33\arcmin19\arcsec

204: 

205: \hoo, also  known as ESO 350-IG38 ($\alpha_{2000}=00^{\rm h}36^{\rm m}52\fs5$,

206: $\delta_{2000}=-33\degr33\arcmin19\arcsec$), is a starbursting blue compact galaxy 

207: located $\sim$88 Mpc.  This distance was derived using $\rm{v_{sys}}=6180\, \ks$

208: \citep{berg06} and assuming $\rm{H_0=71\,km\,s^{-1}\,Mpc^{-1}}$.

209: \hoo's far-UV luminosity ($10^{10.3}\,\rm{L}_{\odot}$) and effective surface 

210: brightness ($10^{9.4}\,\rm{L_{\odot}\,kpc^{-2}}$)

211: are typical of LBGs \citep{heck05}.  An ACS far-UV continuum filter 

212: image of the center of \hoo\, is shown in Figure \ref{acs}.

213: Three separate starbursting nuclei are observed in the image. 

214: In deeper images, \hoo\,

215: appears to have an irregular morphology

216: suggestive of a merger.  This is consistent with  \ha~velocity

217: observations by \citet{ost01}

218: that show that \hoo\,is not dynamically relaxed.

219: IRAS measured a high dust temperature for

220: \hoo\, with 

221: $\rm{F_{60\micron}/F_{100\micron}}~\sim 1.3$ \citep{sand03}, implying an 

222: unusually intense UV radiation field associated with

223: a high star-formation rate per unit area. 

224: \citet{berg00} 

225: suggest that the lack of cold gas could imply that \hoo\,is

226: about to run out of the gas required to fuel the starburst. Lastly, it has

227: a very low gas phase metallicity of $\sim$20\% solar

228: \citep{berg02}. This low metallicity is presumably related

229: to the relatively low amount of dust extinction.  Galaxy properties of \hoo\, and, for comparison, \vv\, are

230: summarized in Table \ref{prop}.  While \vv\, has roughly four times the stellar mass and SFR, \hoo\, 

231: has a significantly higher UV surface brightness due to its smaller size.

232: 

233: %only for vv114w

234: %neutral h poor

235: 

236: 

237: \section{The Data}

238: 

239: \subsection{\fu~Observations\label{fusedata}}

240: 

241: \fu\, was launched on June 25th, 1999.  The \fu\,instrument is composed of 

242: four separate mirrors each having its own Rowland 

243: circle spectrograph \citep{moos00}.  Two mirrors

244: are coated with Al+LiF and two with SiC.  The LiF 

245: mirrors have better sensitivity in the $1000~\ang \lesssim \lambda  \lesssim 1180~\ang$

246: wavelength region, while the SiC are optimized for observations in the range

247: between $900~\ang$ to $1000~\ang$.  

248: During the period when our observations were taken,

249: pointing stability was $\sim0.5 \arcsec$.

250: The spectra are imaged onto

251: two micro-channel plate (MCP) detectors each having an A and B side.  

252: Each side produces 

253: both a LiF and SiC spectrum, for a total of 8 spectra

254: for every \fu\, observation.

255: 

256: 

257: Six time-tag mode exposures of \hoo\, were obtained through 

258: the LWRS aperture ($30\arcsec \times 30\arcsec$) on October 12, 2001.  

259: We have followed the standard data reduction methods for low S/N spectra.

260: The raw exposure data was run through the CalFUSE 3.1.8 data pipeline to produce

261: intermediate data (IDF) files and then the calibrated flux data.  As thermally-induced

262: rotations of the spectrograph gratings and mirror misalignment can cause

263: small zero point shifts in the wavelength calibration, we then cross correlate the

264: individual exposures.  For \hoo\, the largest zero-point wavelength

265: shift was very small at $ < 0.039$\,\AA.  These wavelength shifts were then retroactively applied to

266: the individual IDF files.  We  combined the IDF files for the individual exposures

267: to create one IDF file for each instrument channel.  

268: Extracting calibrated spectra from the longer, combined IDF files 

269: allows the CalFUSE pipeline to fit (and not just scale)

270: the various backgrounds.  

271: In order to minimize the detector background, we have extracted our

272: spectra using the smaller, point source detector regions

273: instead of the usual extended source regions \citep{dix07}.

274: This is possible as a visual examination of the

275: 2-D detector image shows no evidence of galaxy emission outside

276: this smaller, point source extraction region.  

277: Although the exposure time is channel dependent,

278: the final exposure lengths are similar for all channels.

279: For reference,  the LiF 1A channel had 14953 seconds of 

280: total exposure time with 11940 seconds falling during orbital night.

281: 

282: 

283: \subsection{\ch~Observations\label{chandraobs}}

284: 

285: \hoo\, was observed by \ch's ACIS instrument on October 28, 2006.  

286: The target was observed in very faint mode with \hoo\, centered on the S3 aimpoint.

287: We  reprocessed the data with {\it ciao} 3.4.1 and followed the standard

288: \ch\, data reduction threads \citep{ciao}.  For spectra extraction we have used the {\it ciao}

289: command {\it specextract} and a local background from a region surrounding the galaxy.  

290: Approximately 1851 counts are observed from \hoo.

291: As described previously in \citet{grim05} we used a rescaled background

292: image included in {\it CALDB} 3.3.0.1 to create smoothed background subtracted

293: images.  The rescaled background image was also used to create an X-ray radial profile

294: of \hoo\, based on the 0.3 - 1.0 keV count data.

295: 

296: \section{Discussion}

297: 

298: \subsection{Physical Properties of the Hot X-ray Gas}

299: 

300: Figure  \ref{chandrafalse} is a three color adaptively smoothed X-ray image of \hoo\,

301: with 0.3-1.0 keV in red, 1.0-2.0 keV in green, and 2.0-8.0 keV in blue.  

302: Soft X-ray emission surrounds the galaxy and extends up to 6 kpc from

303: the galaxy.  The soft X-ray emission is also seen to have

304: a similar morphology to the \ha\,emission in the adjacent image \citep{schmitt06}.

305:   

306: Knot B shows the strongest emission in the X-ray and \ha\, images but is the weakest

307: knot in the UV.  \citet{kun03} observed that knot B is the reddest knot and exhibited

308: \lya\, absorption.  In both the X-ray and UV imaging data there appears to be an absorption

309: feature running across knot B from the NE to SW.  In all three wavebands Knot C appears

310: to be a single luminous star cluster while the UV image of knot A shows it

311: to be an assembly of several resolved star clusters.  This could explain knot C's 

312: relatively low X-ray surface brightness as compared to knots A and B.

313: 

314: A faint point source is visible SE of knot C in the X-ray image.

315: It is coincident with knot D which was identified by \cite{kun03} in their

316: \lya\,emission image of \hoo\, and therefore is unlikely to be a background

317: AGN.  We have extracted spectra of the point source using a local background region which

318: includes the diffuse thermal emission surrounding the source.

319: We detect  a total of $\sim 22$ counts from the source.  We fit the spectra with a powerlaw

320: and find $\Gamma\sim1.6$ and a luminosity of  

321: $\rm{L_{0.3-8.0\, keV}\sim2.2\times10^{39}\,erg\,s^{-1}}$.   The source's

322: luminosity and location near a star forming region suggest 

323: a possible ultraluminous X-ray source (ULX) identification but

324: are not conclusive.

325: 

326: The complete X-ray spectrum of \hoo\, is plotted in Figure \ref{chandraspec}.

327: As we have done in previous work \citep{grim05,grim06},

328: the spectrum was fit with an xspec model of \linebreak

329: $\rm{wabs(vmekal+zwabs\times powerlaw)}$.

330:  We find a hot gas temperature of

331: $0.68$ keV (see Table \ref{chandraspec} 

332: for the derived fit parameters).  Using

333: these fits we derive luminosities of \linebreak

334: $\rm{L_{0.3-2.0\,keV,thermal}\sim7.2\times10^{40}\,erg\,s^{-1}}$

335: for the soft thermal component and \linebreak

336: $\rm{L_{2.0-8.0\,keV,powerlaw}\sim1.0\times10^{41}\,erg\,s^{-1}}$

337: for the hard X-ray emission.  We estimate 

338: a SFR from the powerlaw fit to the hard X-ray emission of $25\,\rm{M_\odot\,yr^{-1}}$

339: using the relation of

340: \citet{franc03}.  This is in excellent agreement

341: with the \ha, FIR, and radio continuum derived values given  in \cite{schmitt06}.

342: 

343: 

344:  We also find an enhanced \afe\, ratio 

345:  with a value of $\sim4$ times the solar ratio.

346: This is similar to the \afe\, ratios seen in in other starburst galaxies \citep[e.g. ][]{grim05}.

347: The enhanced \afe\, ratio is consistent with 

348: chemical enrichment of the hot gas by supernovae ejecta \citep{strick04b}.

349: Due to the usual degeneracy of the abundance parameters, 

350: it is impossible to place strong constraints

351: on the  absolute \alphe\, and Fe abundances.  

352: Figure \ref{contour} shows the 1-3 $\sigma$ contours of $\chi^2$ as a function of

353: \alphe\,and Fe abundances (relative to solar) from

354: our fit to the X-ray spectra.

355: The 20\% oxygen gas phase metallicity

356: derived in \citet{berg02} is significantly below

357: our allowed range for the  \alphe\, abundances, 

358: again suggesting SN enrichment of the hot gas.

359: Our best fit values for the absolute abundances of the hot gas are 

360: 1 and .25 solar for the  \alphe s\, and Fe respectively.

361: If we assume an oxygen enrichment of $\sim 6$ (relative to solar) 

362: into the wind from supernovae \citep{lim06} we can estimate $\beta$, the ratio of total

363: mass added to the hot wind compared to that from stellar wind ejecta and SN 

364: \citep{strick05}.  Our results suggest a mass 

365: loading in the hot wind fluid of $\rm{\beta\sim7}$.

366: 

367: 

368: 

369: It is also worth estimating the mass of material entrained in the hot X-ray gas.

370: We can use the parameters for the fit to the total Chandra 

371: X-ray spectra of the diffuse gas in \hoo~

372: to derive estimates for the basic physical properties of the hot phase

373: of the wind. The normalization for the VMEKAL

374: component implies an emission integral (the volume integral of density

375: squared) of $5.6 \times 10^{63}$ cm$^{-3}$. For the geometrical volume

376: of the emitting region we take a sphere whose radius encompasses

377: 90\% of the soft X-ray emission ($\sim$2.7 kpc). This then implies

378: a mean gas density of $n_e \sim 5\times 10^{-2} f^{-1/2}$ cm$^{-3}$  and

379: a gas mass of $M \sim 1 \times 10^{8} f^{1/2}$ M$_{\odot}$

380: (where $f$ is the volume filling factor of the X-ray emitting material).

381: For kT = 0.68 keV

382: the mean thermal pressure is $P = 1.1 \times 10^{-10} f^{-1/2}$ dyne cm$^{-2}$

383: and the total thermal energy content of the hot gas is $E = 2.5 \times 10^{56}

384: f^{1/2}\,\rm{erg}$.

385: 

386: Taking the characteristic timescale to be the above radius (2.7 kpc) divided by the

387: sound speed in the hot gas ($\sim$ 450 \kms) yields an outflow age of 6 Myr. The implied

388: outflow rates in the hot gas are then $\dot{M} \sim 18 f^{1/2}$ M$_{\odot}\,\rm{year}^{-1}$ and

389: $\dot{E} \sim 1.3 \times 10^{42} f^{1/2}$ \ergs. 

390: Since this neglects the kinetic energy in the wind material

391: it probably underestimates the total energy transport rate by a

392: factor of a few \citep{strick00}.

393: For volume filling factors similar

394: to those estimated for other starburst winds \citep[$f \sim$ 0.1 to 1, ][]{strick00b},

395: the implied mass outflow rate in \hoo~is thus comparable

396: to the total SFR of $\sim$ 25 M$_{\odot}\,\rm{year}^{-1}$ \citep{schmitt06}.

397: The rate of energy transport is similar to the total

398: rate at which we estimate that mechanical energy 

399: would be supplied by supernovae and stellar winds

400: in the starburst of $\sim 5 \times 10^{42} $ \ergs. 

401: This value was calculated using Starburst99 v5.1

402: assuming a Salpeter IMF from $1-100\,\rm{M_\odot}$ \citep{vaz05}.

403: These outflow rates significantly exceed

404: the rates for the cooler gas derived from the \fu\, data below, but 

405: are typical of winds in powerful local

406: starburst galaxies \citep[e.g.][]{heck03}. 

407: 

408: The Starburst99 calculations estimate a supernovae outflow

409: rate of $\rm{\dot{M}_{SN}\sim3.8\, M_{\odot}\,\rm{year}^{-1}}$.  

410: From the outflow rate of hot gas of 

411: $\dot{M} \sim 18 f^{1/2}$ M$_{\odot}\,\rm{year}^{-1}$,

412: we find a mass loading of $\beta \sim5\,f^{1/2}$.    This value agrees well

413: with our previous estimate of $\rm{\beta\sim7}$.

414:  It is also worth noting that the 

415: pressure and density derived for \hoo\, are almost identical to those in \vv\, 

416: with $P = 9.2 \times 10^{-11} f^{-1/2}$ dyne cm$^{-2}$ and

417: $n_e \sim 5\times 10^{-2} f^{-1/2}$ cm$^{-3}$  \citep{grim06}.

418: 

419: 

420: 

421: 

422: \subsection{\fu Observations of the Wind}

423: The \fu\, data reveals a large number of absorption features, as shown in Figure  

424: \ref{fullspectra}.

425: These absorption lines probe the stellar population, inter-stellar medium (ISM), 

426: and outflowing gas of \hoo.  We have fit absorption lines to most of the most 

427: promiment features using the IRAF 

428: \citep{tod96}

429:  tool {\it specfit} \citep{kriss94}.  Each line was fit using a freely varying powerlaw for 

430:  the continuum and a symmetric gaussian absorption line.  When absorption lines 

431:  are  blended (e.g. \ion{O}{1}~$\lambda$989 and \ion{N}{3}~$\lambda$990), we have 

432:  fit them simultaneously.  For the majority of absorption lines we have used the LiF 

433:  channels to make two independent measurements of the equivalent width, velocity, and 

434: full-width half-maximum (FWHM).  For lines with only one

435: measurement, the line either fell in the gap between LiF 

436: channels or was in the region (1125-1160\AA) 

437: affected by partial light blocking by a wire grid (e.g. the worm) on the LiF 1B channel.  We have ignored the SiC channels above 1000 \AA\,

438: due to their lower S/N.  

439:  Tables \ref{absdata} and \ref{phabsdata} display

440: the results of our fits to the data for the ISM lines and stellar photospheric lines respectively.  

441: Errors are $1\sigma$ and are

442: calculated from the minimization error matrix which is 

443: re-scaled by the reduced $\chi^2$ value 

444: \citep{kriss94}.  

445: 

446: 

447: We use the stellar photospheric lines as an independent measurement of the

448: system velocity of \hoo.  Previous work by \citet{berg06} used \ion{Ca}{2} absorption lines

449: to derive a velocity of 6180 \kms.  This compares favorably with our own measurements of 

450: 6163 - 6202 \kms\, from \ion{Si}{4} and \ion{P}{5} stellar lines.  Therefore we will

451: use the value of 6180 \kms\, as the galaxy's systemic velocity.

452: 

453:  The ISM lines in Table \ref{absdata} suggest a centroid blueshift relative to the

454:  galaxy of $\sim100$ \kms.  A closer examination of

455:  the line profiles shows that they are consistently asymmetric.  

456:  This mild asymmetry in the line profile can be seen in Figure \ref{absprofiles}.  

457:  There appears to be excess absorption on the blue side.  The gaussian models are 

458:  overestimating the FWHM in order to fit the absorption line profiles.  In keeping 

459:  with our previous work on \vv,

460:  we have fit the strongest absorption lines with a second, blue absorption line.  The results of our 

461:  two-component fits are listed in Table \ref{twoabsdata}.

462:  

463: We associate the stronger absorption feature in our fits with the host galaxy or a low

464: velocity outflow.  In general, the

465: centroids of these lines are at $\sim6100$\,\kms.  The central velocity is 

466: blueshifted $\sim80$\,\kms\, relative to the host galaxy.  The stronger absorption

467: lines have FWHMs of $\sim300$\,\kms, suggesting that the  lines do

468: include the galaxy's ISM.

469: The weaker absorption lines are generally at $\sim5900$\,\kms. 

470: We identify the second set of absorption lines with a high velocity outflow of material

471:  moving at $200-280$\,\kms\, away from the galaxy.  Interestingly, a 

472:  comparison of the equivalent widths of \ion{C}{2} to \ion{C}{3} and \ion{N}{2} to  \ion{N}{3} show that 

473:  the higher ionization lines are relatively stronger in the outflow as compared to the 

474:  lower velocity absorption features.  This could be a product

475:  of shocks between the outflow and the ambient material.

476: 

477:  

478:  

479: %First just a general discussion of the line profile (two components in

480: %absorption plus a red-shifted emission-line). You could then comment about

481: %how the properties of the absorption line compares to VV 114 and other

482: %starbursts. You might even place Haro 11 on the plot of N_OVI vs. b like in

483: 

484: 

485:  

486: In Figure \ref{absprofiles} it is clear that the 

487: coronal gas line \ovi$\lambda1032$ has a different profile 

488: than the other ISM absorption lines.  

489: While the \ovi\,centroid velocity of $6043\pm34$\,\kms\,

490: is roughly consistent with those of the ISM absorption lines,

491: there are significant discrepancies.

492: First, a single absorption

493: line fits  \ovi\, poorly,  due to a weak blue wing

494: extending several hundred \kms\, away from the

495:  galaxy.  While this blue wing is relatively weak

496:  compared to the primary feature, it extends to much lower

497:  velocities (e.g. higher outflow velocities) than the 

498:  blue wings we observe in some of the

499:  other absorption profiles.

500: Secondly and more intriguingly, \ovi\,

501:  emission appears to be present, forming part of a 

502:  P-Cygni profile.  An emission feature is observed 

503:  at $6303\pm21$\,\kms.  This identification is supported by a 

504:  corresponding emission feature for \ovi$\lambda1038$

505:  with a similar centroid at $6282\pm30$\,\kms.  

506: They also have similar FWHMs of

507: $185\pm47$ and $151\pm55$ \kms\,for \ovi$\lambda1032$

508: and \ovi$\lambda1038$ respectively.  

509:  An absorption feature

510:  is not observed for \ovi$\lambda1038$ but it is likely hidden in the

511:  broad \ion{C}{2}$\lambda1036$ absorption feature.  

512:  In fact, the previously described two component absorption line profile

513:  for \ion{C}{2} fits the red wing particularly poorly (Figure \ref{absprofiles}) and has a large

514: FWHM, consistent with  contamination by an \ovi$\lambda1038$ 

515:  absorption feature.  The \ovi$\lambda\lambda1032,1038$ doublet

516:  is the only ion that we unambiguously identify in emission

517:   although there is a possible emission feature for \ion{C}{3}$\lambda977$

518:  at 6356\,\kms.  The \ovi\, P-Cygni profile has been previously identified by

519:  \citet{berg06}.

520:  

521:  %the VV 114 paper. I assume it will be "normal". You could talk about the

522: %generic idea that OVI arises in regions where the hot X-ray gas and cold gas

523: %overlap, possibly through the ablation of material off the clouds (as in

524: %Marcolini et al).

525:  

526: Broad P-Cygni profiles

527: are common for \ovi\, in spectra of O stars \citep{pell02}.

528: While a P-Cygni profile has  not been observed in other starburst galaxies,

529: the high UV surface brightness, low UV attenuation,

530: and high SFR of \hoo\,could suggest that we are detecting an aggregate

531: stellar profile.  

532: However, the observed \ovi\,emission and absorption features are significantly

533: narrower than those seen in stellar spectra.  

534: In particular, narrow \ovi$\lambda\lambda1032,1038$ emission is not observed

535: in the stars in the  sample of  \citet{pell02}.  The narrow \ovi\,feature is

536: also inconsistent with the synthetic starburst spectra generated by \citet{rob03}.

537: We therefore explore the possibility that the \ovi\, profile is not a stellar feature

538: but is instead due to interstellar gas.

539: 

540: In previous analyses of starburst galaxies 

541: \citep[e.g.][]{heck01,grim06} the \ovi\,absorption has been the most

542: blueshifted and broadest feature.  

543: The extreme outflow velocities of the \ovi\,absorption suggests

544: that in those galaxies, \ovi\, is not produced 

545: directly by the cooler outflowing clouds responsible

546: for the other absorption lines.   Instead, \citet{heck01} attribute the 

547: production of  \ovi~

548: to the intermediate temperature regions created by the 

549: hydrodynamical interaction between hot outrushing gas

550: and the cool fragments of the ruptured superbubble seen 

551: in H$\alpha$ images. Such a situation is predicted to be

552: created as an overpressured superbubble 

553: accelerates and then fragments as it expands out of the galaxy \citep{heck01}.  

554: 

555: \citet{heck02} derive a simple and general relationship between the \ovi~

556: column density and absorption line width which holds whenever 

557: there is a radiatively cooling gas flow passing through the coronal

558: temperature regime.  They showed that this simple model 

559: accounted for the properties of \ovi ~ absorption line systems as 

560: diverse as clouds in the disk and halo of the

561: Milky Way, high velocity clouds, the Magellanic Clouds, 

562: starburst outflows, and some of the clouds in the IGM (but see Tripp et al. 2007,

563: in prep, for a more thorough test of the relation for IGM clouds).  

564: From our fit to the \ovi$\lambda1032$ absorption feature we have derived

565: a column density of $\rm{N_{OVI}=4.2\times10^{14}}\,cm^{-2}$.

566: \hoo\, does fall within the scatter in the relationship, but we would expect

567: a slightly higher \ovi\,column density

568: for the measured line width of $\rm{\sqrt{2}\sigma=b=106\,\ks}$.

569: While the equivalent width and breadth of the \ovi\,absorption  are 

570: roughly consistent with \ovi\, production in

571: the interface between the outrushing gas and the 

572: cool shell fragments \citep{heck01}, the kinematics of \ovi\,

573: are inconclusive.  Specifically, unlike our previous 

574: observations of \vv\, and NGC\,1705 \citep{grim06,heck01}, 

575: the primary \ovi\, absorption feature is not blueshifted relative to the cooler, ISM 

576: lines.  

577: 

578:  

579:  %stellar or cospatial or ...

580:  

581: %Then you could focus on the emission line. You could say there are two ways

582: %alternative ways to interpret this. The first is that this is just

583: %radiatively cooling gas. You can then turn the extinction-corrected OVI

584: %luminosity into a constraint on cooling. This is essentially what you have

585: %done now. 

586: 

587:  \ovi\, emission has previously been observed in only

588:  two spiral galaxies, NGC\,4631 \citep{ott03} and our own

589:  Milky Way \citep[][and references therein]{dix06}.  The \ovi\, emission in the

590:  halo of NGC\,4631 was attributed to cooling gas from

591:  a galactic chimney \citep{ott03}.  Similarly,

592:  for \hoo\,, we could attribute the \ovi\, emission

593: to radiatively cooling gas

594:  associated with the outflow. 

595:  As \ovi\, is the dominant coolant for gas at temperatures 

596:  $\rm{T\sim10^{5.5}\,K}$, 

597:  \ovi\,cooling could inhibit the ability of the outflowing gas to

598: escape the galaxy.  Therefore it is interesting 

599: to estimate the implied energy losses associated with

600: the \ovi\, emission.

601: 

602: The observed flux in \ovi\, emission is

603: $2.3\pm0.8\times10^{-14}\,\rm{erg\,s^{-1}\,cm^{-2}}$ and

604: $1.4\pm0.5\times10^{-14}\,\rm{erg\,s^{-1}\,cm^{-2}}$

605: for the $\lambda1032$ and $\lambda1038$ lines respectively.

606: While the \ovi$\lambda1038$ flux could be an underestimate due to

607: \ion{O}{1}$\lambda1039$ absorption, these measured values

608: are near the theoretical upper limit

609: of $\rm{F_{OVI\lambda1032}/F_{OVI\lambda1038}}=2$

610: found in an optically thin gas.

611: We correct the fluxes for both 

612: galactic extinction and intrinisic

613: dust attenuation.

614: To correct for galactic extinction we 

615: use E(B-V)=0.011 \citep{schleg98} and assume

616: the extinction law of \citet{card89}. 

617: For intrinsic attenuation we start with 

618: work by \citet{seib05}.  They used

619: {\it GALEX} and {\it IRAS} data to

620: study the relationship between

621: UV attenuation (1500 \AA) and the IR excess.

622: From Table \ref{prop}

623: we find an IR excess for \hoo\, of $\rm{\log(L_{FIR}/L_{UV}})= 0.53$.  This IR

624: excess correspond to $\rm{A_{1500\,\AA}\sim1.6\,mag}$ \citep{seib05}.

625: We then extend this dust attenuation to 1035 \AA\,

626: following the work by \citet{leith02}. 

627: They used Hopkins Ultraviolet Telescope (HUT) 

628: observations of star-forming galaxies to study 

629: dust attenuation at shorter wavelengths.

630: They found a factor of 1.3 increase in

631: the dust attenuation from 1500 to 1035 \AA.

632: We therefore calculate $\rm{A_{1035\,\AA}\sim2.1\, mag}$. 

633: Finally we estimate a total corrected \ovi~luminosity

634: of $\sim2.8\times10^{41}$ \ergs.

635: \ovi\, is responsible for $\sim30\%$ of the

636: cooling in coronal gas \citep{heck01} so 

637: $\rm{L_{coronal}\sim9.3\times10^{41}}$ \ergs.

638: This luminosity is larger

639: than \hoo's thermal X-ray luminosity ($7.2 \times 10^{40}$ \ergs) 

640: and

641: is comparable to both the energy outflow rate of

642: the hot gas ($\dot{E} \sim 1.3 \times 10^{42} f^{1/2}$ \ergs)

643: and the

644: mechanical energy being input by

645: supernovae ($\sim 5 \times 10^{42} $ \ergs).  

646: Therefore, if the \ovi\, emission is 

647: produced by radiative cooling of

648: the outflowing gas, it suggests that a significant

649: percentage (20\%) of the energy input from the supernovae could

650: be radiated away by the coronal gas.

651: 

652: 

653: %You could then show that the ratio of L_OVI/L_x is very high in

654: %Haro 11 compared to other starbursts (M 82, NGC 4631, NGC 1705). So maybe

655: %there is something else going on! You could then note that the line profile

656: %really is suggestive of a classic P-Cygni profile where the emission

657: %component is actually _scattering_ of continuum photons off the outflowing

658: %OVI on the backside of the flow. In this case, the OVI luminosity is not

659: %really telling you about cooling rates (it's just telling you about covering

660: %factors and column densities). You could note that the lack of the emission

661: %component in the other lines would then require that somehow we see the

662: %backside of the outflow only in OVI (with the emission from the other ions

663: %being strongly attenuated by dust). You can then speculate that perhaps this

664: %means that the OVI in the flow is arising in a distinctly different region

665: %than the other lower ions. Perhaps larger? You could note that this does not

666: %readily fit into the model where the OVI is co-spatial with the cool clouds.

667: 

668: Several previous works have discussed the  $\rm{L_{OVI}/L_{X-ray}}$ ratio.

669: In their detection of \ovi\, emission in NGC\,4631, \citet{ott03} estimated 

670: $\rm{L_{OVI}/L_{X-ray}}\sim0.11$.  

671: Although \ovi\, emission was not observed in NGC\,1705 and M82, lower limits

672: of   $\rm{L_{OVI}/L_{X-ray}}\lesssim3$ \citep{heck02} and $\lesssim0.4$ \citep{hoop03} 

673: were determined respectively.  Our own result, $\rm{L_{OVI}/L_{X-ray}}\sim3.9$ seems

674: anomalously high in comparison to these previous results.

675: 

676: It is possible that the \ovi\, emission line in \hoo\, is not tracing the

677: rate at which coronal gas is radiatively cooling. Instead it may arise

678: through resonant scattering of far-UV continuum photons from the starburst

679: off \ovi\, ions in the outflow. The redshift of the emission component is a

680: natural consequence of such a picture (this is a classic P-Cygni profile).

681: There is a shell of coronal gas flowing out from the starburst, and is

682: traced by \ovi\, ions. The front side of the flow is seen as blue-shifted

683: absorption against the far-UV continuum of the starburst, while resonant

684: scattering by \ovi\, ions on the back side of the flow produces a red-shifted

685: emission line. For a symmetric flow geometry, and in the absence of

686: significant differential dust extinction of the OVI emission from the back

687: side, this would predict that the equivalent widths of the blueshifted

688: absorption and redshifted emission lines would be equal and opposite. This

689: is roughly what is observed in Haro 11.

690: 

691: If this is the correct interpretation, the question is then why a redshifted

692: emission component is not seen in the other ions (lower ionization states).

693: This would have to mean that the \ovi\, and cooler gas are not co-spatial:

694: redshifted emission from the back side of an outflow is obscured by dust for

695: the low ions, but not for \ovi\,. This could be the case (for instance) if the

696: size scale of the flow were significantly larger in OVI so that more of the

697: back side lay beyond the region of significant dust extinction. 

698: 

699: With the present data it is not possible to determine whether the \ovi\,

700: emission is produced by radiative cooling, or resonant scattering. If it's

701: the latter, then the true contribution to radiative cooling by the coronal

702: gas discussed above must be regarded as an upper limit. This reinforces the

703: idea that radiative cooling is not significantly influencing the dynamics of

704: the wind.

705: 

706: 

707: %

708: 

709: %Secondly, the kinematics of the \ovi\,emission, redshifted $\sim110$\,\kms\,

710: %relative to the galaxy, also suggest that the \ovi\, emission is not 

711: %radiatively cooling gas associated with the outflow.  

712: %If we are observing \ovi\, emission

713: %from the backside of the outflow, where is the

714: % \ovi\, emission from the

715: %near side?    

716: 

717: %As discussed earlier, the full \ovi\, feature is suggestive of a P-Cygni profile, particularly as

718: %the midpoint of the P-Cygni profile is consistent with the

719: %centroid velocity of the galaxy ($6180$\,\kms).  

720: %The lack of emission and the different kinematics

721: %of \ovi\,  suggests that the \ovi\, feature arises in

722: %a different physical region than the outflowing lower ionization state

723: %ions.  

724: %We therefore think it unlikely the \ovi\, emission is probing the

725: %cooling of the outflowing gas.  

726: %If the \ovi\, was produced in the

727: %outflow it appears to be moving approximately $\gtrsim-120$\,\kms\, and $\gtrsim+120$\,\kms \,

728: %relative to the galaxy for the absorption and emission features respectively.

729: %Therefore the \ovi\, could be produced in an outflow of gas expanding

730: %at $\sim120$\,\kms.  It is difficult to interpret this possibility as we are

731: %unable to determine the geometry of the outflow (e.g. spherical, bipolar)

732: % from the \fu\, data so we have no handle on the opening angle of the outflowing gas.

733: %We conclude that, although we are unable to determine

734: %the source of the \ovi\, emission, radiative cooling

735: %by \ovi\, emission is not significantly inhibiting the growth of the outflow.

736: 

737: 

738: %You can then speculate that perhaps this

739: %means that the OVI in the flow is arising in a distinctly different region

740: %than the other lower ions. Perhaps larger? You could note that this does not

741: %readily fit into the model where the OVI is co-spatial with the cool clouds.

742: 

743: 

744: 

745: 

746: %A P-Cygni profile is generically attributed to 

747: %an expanding shell of gas.  \hoo\,

748: %shows morphological evidence of a recent merger and is not dynamically relaxed.

749: %Unfortunately, we are unable to

750: % unable to determine the geometry of the outflow (e.g. spherical, bipolar)

751: % from the \fu\, data which hinders our ability to interpret the \ovi\, emission

752: % as we have no handle on the opening angle of the outflowing gas.

753: 

754: % in a starburst galaxy \citep{hoop03}.  This is probably

755: %not a result of \hoo's small FUV attenuation or low

756: %metallicity as \ovi\, emission was not seen

757: %in the dwarf starburst NGC\,1705 which has an even lower

758: %attenuation and metallicity \citep{heck01}.

759: % 

760: 

761: %

762: % If the outflow drives into the ISM with a velocity of  

763: % $\sim 150$\,\kms\,,  \ovi\, is produced

764: % behind the shock \citep{dop96}.

765: % In our observations, the \ovi\, ions themselves appear to be moving at 

766: % approximately $\gtrsim-120$\,\kms\, and $\gtrsim+120$\,\kms \,

767: % relative to the galaxy for the absorption and emission features respectively.

768: % However, we know that there is an outflow component moving at

769: % $\gtrsim+200$\,\kms\, so their is enough kinetic energy

770: % in the shock to produce postshock temperatures $\rm{T\gtrsim10^{5.5}\,K}$.

771: %In comparison both M82 and NGC\,1705 have outflow

772: %velocities below  150 \kms\, which could

773: % explain their lack of \ovi\, emission.  \vv\, however has a higher outflow velocity

774: %of 300-400 \kms.  The range of temperature at which 

775: %\ovi\, is an important coolant is relatively narrow and the high

776: %outflow velocities in \vv\, might  produce postshock temperatures

777: %above that range.  Secondly, \vv\,has significant FUV attenuation

778: %so \ovi\, emission could be blocked.  We therefore attribute

779: %the majority of the \ovi\, in \hoo\, to having been produced in

780: %the postshock region of the outflow as it is now

781: %entrained with the lower ionization state outflowing gas.

782:    

783: 

784: 

785: 

786: %%no disk but if there is we might be looking at 90

787: 

788: \subsection{Escape of Ionizing Radiation?}

789: 

790: 

791: Recent observations imply that the epoch of reionization 

792: occurred at a redshifts $\rm{6<z<20}$ 

793: (Becker et al. 2001, Fan et al. 2002, Spergel et al. 2003). Several mechanisms have been proposed

794: for reionization including UV photons from star formation or active galactic nuclei (AGN), 

795: X-ray photons 

796: from supernovae and mini-quasars (Oh 2001), and even sterile neutrino decay 

797: (Hansen \& Haiman 2004).  However X-ray background and cosmic microwave 

798: background (CMB) observations

799: rule out significant contributions by X-ray photons and neutrino decays 

800: (Mapelli \& Ferrara 2005; Dijkstra et al. 2004). UV photons from quasars

801: likewise appear to be inadequate due to the rapidly

802: falling space density of quasars at $\rm{z>3}$ (Fan et al. 2001).

803: While recent work suggests that

804: UV photons from stars could reionize the universe 

805: (Yan \& Windhorst 2004), this inference is quite uncertain. First, the cosmic star formation

806: at the relevant redshifts is still uncertain. Second, the fraction of ionizing photons that

807: are able to escape to the IGM ($f_{esc}$) is even more uncertain.

808: 

809: 

810: The escape of ionizing continuum radiation from star forming galaxies 

811: was first observed by Steidel et al. (2001) in a sample of 

812: $\rm{z\sim 3.4}$ LBGs.  

813: It has been suggested that the lower

814: luminosity LBGs (e.g. $\rm{L < L_*}$) could play an

815: important role in reionization (Bouwens et al. 2006).

816: Galactic winds appear to be a global property of LBGs

817: (Shapley et al. 2003), and this may explain why a significant fraction of their ionizing radiation

818: is able to escape.

819: 

820: 

821: More recent work by \citet{shap06} focused on deep rest-frame UV spectroscopy of

822: a sample of 14 LBGs.  They detected significant 

823: Lyman continuum emission in only two of the galaxies.  

824: They found a sample-averaged relative Lyman continuum

825: escape fraction of 14\% but were unable to find

826: a galaxy property that controlled

827: the detection of the

828: Lyman continuum. Their results suggest that

829: even at high redshifts ($\rm{z\sim3}$) significant Lyman leakage is

830:  the exception.

831: 

832: 

833: Most recently \citet{berg06} published an analysis showing 

834: the detection of Lyman continuum emission in \hoo.  \hoo's low 

835: neutral hydrogen mass limit of $\lesssim10^8\rm{M}_\odot$ \citep{berg06}, 

836: high FUV surface brightness, irregular morphology, and low metallicity \citep{berg02} all suggest

837: that \hoo\, is an ideal local candidate for  Lyman continuum emission.  

838: \hoo's Lyman continuum leakage was in fact our prime motivation

839: for obtaining \ch\, data of the galaxy.  As part of our analysis we have

840: re-analyzed the \fu\, data and have come to a different conclusion than the

841: previous work by \citep{berg06}.

842: 

843: 

844: For the following discussion we  focus

845: primarily on the night only data.  While we lose 4 ks (12 ks total night exposure) of observing time,

846: the night background level is considerably lower and relatively well behaved. 

847: In Figure \ref{sic1b2aorig} we have plotted the extracted spectra for the SiC 2A and SiC 1B detectors.

848: As discussed in section \ref{fusedata} we have used the standard 

849: extraction methods  for 

850: low S/N spectra to produce these plots.   

851: The SiC 1B spectra show a negative flux, a common feature 

852: of low S/N LWRS  spectra at these wavelengths.  

853: This is related to the difficulty in fitting the background models to the LWRS aperture region

854: as it is near the bottom edge of the detector (see Figure \ref{detector}).  Unlike the SiC 1B spectra, the SiC 2A spectra have

855:  a small positive net flux.  

856: 

857: The oversubtracted  SiC 1B spectra suggests the relative importance of the background 

858: subtraction.  Figure \ref{detector} shows images of the two micro-channel plates (MCPs) in the regions

859: of interest.  We have overlaid the point source extraction region and wavelength scales on the image. 

860: While galactic emission is clearly visible at wavelengths longward of \hoo's Lyman break ($\sim 930.8\,\rm{\AA}$),

861: there is no convincing visual evidence of Lyman continuum emission in these images.  More importantly, possible

862: emission features in SiC 2A are not replicated in the SiC 1B image.  

863: In order to examine the spectra

864: more directly we have extracted

865: the total counts in 1\AA\, bins for a rectangular region approximating the source aperture.  We have also

866: extracted the same information from adjacent regions of the same size directly above and below the source aperture (Figure \ref{radialprofile}).  These background  profiles are 

867: similar to the source spectrum 

868: for wavelengths below the

869: Lyman limit.  If present, any Lyman continuum flux 

870: appears to be fainter than the variance present in the background regions.

871: 

872: In Figure \ref{sic2aairglow} we show

873: the regions flagged by the CalFUSE pipeline as 

874: possibly contaminated by geocoronal or scattered solar light.  

875: While these features are significantly weaker during orbital night, the region of interest between 

876: $920 - 930 \rm{\AA}$ is almost fully covered by these features.  

877: Determining the background in these regions is thus particularly difficult.

878: Regions of background contamination were interpolated across when 

879: the \fu\, scattered light background files

880: were created \citep{dix07}.

881: 

882: There are several components to the \fu\, background.  The intrinsic background of

883: $\sim\, 0.5\, \rm{counts\, cm^{-2}\, s^{-1}}$ consists of beta decays from $^{40}\rm{K}$ on the MCP glass  

884: and cosmic rays \citep{dix07}.  In addition, a contribution from geocoronal light varies from

885: 20-300\% of the intrinsic background during the orbit.  The \fu\, pipeline uses independent

886: measurements of the dark count, night geocoronal scattered light, and day geocoronal scattered light

887: to fit the background observed on {\bf unilluminated} regions of the detector.  An 

888: optimal extraction algorithm then extracts the source counts

889: from the background \citep{horn86}.  Figure \ref{sic2acounts} shows

890: the total counts and fitted background model for the SiC 2A night time data.

891: The background clearly dominates the observed counts in the regions below

892: the Lyman limit.  While the fitted background is generally reasonable, it does

893: over-subtract in some regions.  This again highlights the difficulties in subtracting

894: a temporally and spatially varying background from relatively low S/N data.

895: 

896: We have derived total fitted source and background fluxes for the SiC 2A night data

897: in two wavelength ranges below the Lyman limit.  From 

898: 920-925 \AA\, we found a

899: source flux of $1.9\times10^{-15}\,\rm{erg\,cm^{-2}\,s^{-1}\,\AA^{-1}}$ 

900: compared to a background

901: flux of $8.7\times10^{-15}\,\rm{erg\,cm^{-2}\,s^{-1}\,\AA^{-1}}$.  

902: For comparison, we also examined

903: the 925-930 \AA\, region within which we would expect less airglow contamination (e.g. Figure \ref{sic2aairglow}).

904: This region has a significantly lower source flux of 

905: $5.6\times10^{-16}\,\rm{erg\,cm^{-2}\,s^{-1}\,\AA^{-1}}$

906: with a comparable background of $6.1\times10^{-15}\,\rm{erg\,cm^{-2}\,s^{-1}\,\AA^{-1}}$.  

907: The source flux varies by a factor of $\sim 3$ between these two regions although we would not expect

908: the Lyman continuum to vary strongly \citep{berg06}.  

909: A 20\% error in the background subtraction would erase any source flux even in the 

910: 920-925 \AA\, region.  

911: 

912: %BKGD  2A  bkgd2a013.fit  52120.000  0

913: %BKGD  2A  bkgd2a014.fit  52202.000  0

914:  %52194

915: 

916: As we have stated, the \fu\, background varies significantly

917: during an observation.  Airglow and geocoronal scattered light

918: both contribute to the background and vary depending on

919: the pointing of the satellite relative to the earth.  

920: As the \fu\,data is time-resolved, we can study

921: the detector count rate as a function of wavelength and

922: earth limb angle.  In Figure \ref{earthlimb} we have

923: plotted the earth limb angle versus detector count rate 

924: for three wavelength regions.

925: This plot includes both night and day data

926: so that we could include the lowest earth limb angles.  

927: The Lyman continuum region is represented by

928: the 920-930 \AA\, points while the 930-940 \AA\, and

929: 940-940 \AA\, regions contain some emission intrinsic to

930: \hoo.  The Lyman continuum region

931: has the lowest count rate for all earth limb angles.

932: More importantly, as the earth limb angle increases, the count rate 

933: continues to fall even at the highest

934: earth limb angles.  We estimate an expected dark count rate from cosmic rays and $^{40}\rm{K}$ 

935: of $\sim0.05\,\rm{counts\,s^{-1}}$  which closely matches the count rate in the Lyman

936: continuum region observed at the 

937: highest earth limb angles.  This again suggests that the observed Lyman continuum emission

938: is likely a systematic artifact of extracting low S/N spectra from a dominant

939: and variable background.

940: 

941: For Figure \ref{sic1b2afix} we have rescaled the SiC 1B background subtraction to

942: match the observed SiC 2A spectra.    In comparing the two spectra it is worthwhile

943: to note that in the regions below 930\,\AA\,, possible spectral features are not replicated

944: in both spectra.  This also suggests that the detected emission is largely a combination

945: of detector noise and systematic effects from the background subtraction.

946: 

947: A comparison to the results by \cite{berg06} of the SiC 2A data

948: show a positive offset of the continuum level by  

949: $\sim1\times10^{-14}\,\rm{erg\,cm^{-2}\,s^{-1}\,\AA^{-1}}$.  

950: The continuum shapes however are similar.  

951: In the \citet{berg06} spectra it is worthwhile noting that none of the

952: absorption lines are black.  This includes two narrow Milky

953: Way absorption lines \ion{C}{3} $\lambda 977$ and \ion{C}{2} $\lambda 1036$

954: which would be expected to be saturated at line center, independent of

955: the properties of \hoo.  These lines are black in our spectra as seen in

956: Figure \ref{fullspectra}.

957: 

958: As we are analyzing the same observational data, we have three 

959: possible reasons  to explain the discrepancy between our results and those of \citet{berg06}.  

960: First, we are using a 

961: newer version of the CalFUSE pipeline (3.1.8 versus 3.0) which includes

962: several improvements in the background analysis.  In particular

963: v3.0 did not properly account for regions of the detector excluded because of 

964: airglow contamination and therefore frequently underestimated the total background.

965: Secondly, we suspect that the extended source extraction region was used for the

966: prior analysis by \citet{berg06}.  For \hoo\, the point source extraction region includes all of the 

967: source flux and has

968: a higher S/N due to the smaller extraction region on the detector.  Lastly, it is possible that the

969: individual exposures were separately background subtracted.  For low S/N spectra

970: it is preferable to combine the exposures and allow the CalFUSE pipeline to fit

971: a background model to the merged dataset.  If the background subtraction

972: is performed on the individual exposures of \hoo\,, the CalFUSE pipeline will

973: simply scale the subtracted background by the exposure time.  As the background

974: is varying both spatially and temporally, this causes 

975: systematic offsets in the background subtraction.  This explanation is likely

976: as it would also explain the offset seen even in the relatively high S/N LiF 1A spectra 

977: of \citet{berg06}.

978: 

979: Although we find no convincing evidence for Lyman continuum emission it is useful

980: to derive an upper limit.  Following \citet{berg06} we use the  920-925 \AA\, flux 

981: and correct for galactic extinction.  We derive

982: $\rm{f_{900}\lesssim2.3\pm0.7\times\,10^{-15}\, erg\,cm^{-2}\,s^{-1}\,\AA^{-1}}$

983: as compared to the previous result of 

984: $\rm{f_{900}=1.1\pm0.1\times\,10^{-14}\, erg\,cm^{-2}\,s^{-1}\,\AA^{-1}}$

985: \citep{berg06}.  Our $1\sigma$ errors are purely statistical and do not include the

986: dominant systematic errors.  Using our derived value of $\rm{f_{900}}$

987: and the formulism for deriving $\rm{f_{esc}}$ derived by

988: \citet{berg06}, we find $\rm{f_{esc}\lesssim2\%}$.

989: 

990: 

991: %%e(b-v)=0.011

992: 

993: 

994: \subsection{X-ray Properties of LBG Analogues}

995: 

996: 

997: In \citet{grim05}  we analyzed \ch\, observations of the

998: diffuse emission in 23 star-forming galaxies.  We showed that the properties 

999: of the hot gas  (e.g.

1000: temperature, metal enrichment, and luminosity)  were remarkably 

1001: consistent throughout the sample.  These findings suggested that the

1002: same physical mechanism was

1003: producing the diffuse X-ray emission throughout the sample.

1004: We found that shocks, driven by a galactic ``superwind'' powered by the kinetic

1005: energy collectively supplied by stellar winds and supernovae in

1006: the starburst,  were the likely explanation for this correspondence. 

1007: As we have now obtained deep \ch\, observations of two

1008: LBG analogues, it is worth examining how they fit within

1009: this framework.

1010: While it is 'dangerous' to make

1011: general conclusions based on such a small sample size,  

1012: it is worth discussing as few LBG analogues are 

1013: currently observable in the X-ray without unreasonable

1014: exposure times.  

1015: 

1016: \hoo\, and \vv\, were selected due to their high UV luminosities

1017: and compact size.  This suggests a high SFR 

1018: compared to galaxies of similar mass (see \citet{heck05}).  Given the selection criteria,  

1019: we find that the properties of the hot gas follow accordingly.

1020: The left panel of Figure \ref{div_L_K} is a plot of  $\rm{L_{FIR} + L_{UV}}$ 

1021: (a proxy for SFR) 

1022: versus the thermal 0.3-2.0 keV 

1023: X-ray luminosity.  This plot

1024: is originally taken from \citet{grim05} and includes

1025: dwarf starbursts, starbursts, and ULIRGs.

1026: There is a clear correlation (possibly non linear) 

1027: between the thermal X-ray emission and the SFR.

1028: \hoo\, and \vv\, fall between the starbursts 

1029: and ULIRGs as expected for LBG analogues.

1030: In the right-hand panel we have divided the thermal X-ray 

1031: and FIR+UV luminosities by the 

1032: $K$-band luminosity (a proxy for stellar mass).

1033: This shows a fundamental connection between the 

1034: level of star formation and the luminosity of the

1035: thermal X-ray gas.  Unsurprisingly, both \vv\,

1036: and \hoo\, have high SFRs and soft X-ray luminosities given

1037: their stellar mass.

1038: 

1039: In \citet{grim05} we showed that both the $K$-band and FIR luminosities

1040: scaled with the radial extent of the soft X-ray emission.  

1041: In Figure \ref{r90} we have plotted the 0.3-1.0 keV 90\% flux

1042: enclosed radius versus the FIR and $K$-band luminosities.

1043: Due to the high SFRs and compact sizes, both

1044: \hoo\, and \vv\, fall below the 

1045: FIR relation for other starbursts.  Previously, the $K$-band

1046: relation had been only marginally preferred

1047: to the FIR-band relation based on

1048: Kendall's rank-order correlation coefficient.

1049: The chances of a spurious correlation

1050: were $2.7\times10^{-5}$ versus

1051: $3.4\times10^{-5}$ for $K$-band and FIR respectively.

1052: With the two new data points 

1053: the chance of a spurious relation is unchanged for the 

1054: FIR data.  However the K-Band relation improved with

1055: a chance of a spurious correlation of 

1056: $4.4\times10^{-6}$.

1057: This suggests that the host galaxy has a strong

1058: effect on the spatial extent of the X-ray emission, 

1059: although the starburst drives

1060: the energetics and composition of the hot gas.

1061: 

1062: Figure \ref{F60o100vkt} has a plot of the X-ray temperature

1063: versus the dust temperature 

1064: (as traced by the IRAS F60$\micron$/F100$\micron$ ratio).

1065: While most of the galaxies show a rough

1066: correlation \hoo\, with F60$\micron$/F100$\micron$=1.29 \citep{soif89} 

1067: is one of the outliers.  Physically, as the 

1068: FIR emission is reprocessed UV light, 

1069: F60$\micron$/F100$\micron$ corresponds to the

1070: luminosity-weighted mean dust temperature

1071: which is set by the mean FUV intensity in the

1072: starburst ( $\propto$ SFR per unit area).

1073: Knot B appears to be a compact region

1074: of intense star formation as it is the brightest feature

1075: in \ha\,, X-ray, Spitzer IRAC and MIPS, and 3.6 cm radio data \citep{schmitt06}. 

1076: Knot B is probably dominating the observed 

1077: IRAS F60$\micron$/F100$\micron$ ratio

1078: and driving the anomalously high SFR per area.

1079: 

1080: %driving the high   F60$\micron$/F100$\micron$ ratio.

1081: 

1082: 

1083: \section{Conclusions}

1084: 

1085: Star formation driven feedback is an essential ingredient to understanding

1086: galaxy evolution and the IGM.  Observations of local starbursts 

1087: have shown that galactic winds, driven by the kinetic energy from

1088: supernovae and stellar winds, are the strongest manifestation

1089: of star formation driven feedback.  These outflows are complex multi-phase phenomena whose physical, chemical, and dynamical properties can only be understood through complementary observations at many wavebands and detailed modeling \citep{veill05}. Observations of the coronal ($10^5$ to $10^6$ K) and hot ($10^6$ to $10^7$\,K) gas are particularly important, as they provide essential information about the importance of radiative cooling of the outflow and about the dynamics and energy content of the wind. The coronal gas is best traced through the \ion{O}{6} doublet in the far-UV and the hot gas is best observed via its soft X-ray emission.

1090: 

1091: The strong overall cosmic evolution in the global SFR  \citep[e.g. ][]{bunk04} implies that the bulk of the feedback from galactic winds occurred at early times ($z >$ 1). Indeed, the direct signature of galactic winds -- the presence of broad, blueshifted interstellar absorption lines in the rest-frame UV -- is generically present in the LBGs \citep{shap03}. These are the best-studied population of high-redshift star forming galaxies  \citep{steid99}. However, such data provide only a narrow range of information about galactic winds. Unfortunately, direct observations of the hotter gas in the rest-frame soft X-ray and FUV regions for high redshift galaxies are extremely difficult or even impossible to obtain, which hinders progress in understanding 

1092: this complex phenomenon.  Finding and studying

1093: analogs to LBGs is therefore very important

1094: as a proxy for studying the outflow phenomenon.

1095: 

1096: 

1097: 

1098: In this paper we have described FUV and soft X-ray observations of \hoo, the second Lyman break analogue

1099: observed by \ch.  Our \fu~observations in the FUV show strong and broad interstellar absorption lines. The lines with the highest S/N show two kinematic components. One is centered near the galaxy's  systemic velocity and apparently includes the ISM. The second component is strongly blueshifted,  suggesting an outflowing wind with a velocity of  $\sim200-280\,\ks$.

1100: This is consistent with what is seen at slightly longer rest wavelengths

1101: in high redshift LBGs and further establishes the similarity of \hoo~to the LBGs.  

1102: The \ovi\, feature has a P-Cygni profile that is not seen in other

1103: starbursting galaxies.  While we are not able to clearly determine

1104: whether the source of the \ovi\, emission

1105: is radiatively cooling or resonance scattering, 

1106: we estimate that $\lesssim20\%$ of the available

1107: energy from supernovae could be lost to radiative cooling.  This suggests that 

1108: \ovi\,radiative losses do not significantly inhibit the growth of the outflow in \hoo.

1109: 

1110: The wind in \hoo ~ might in principle be able to carve a channel in the ISM through

1111: which ionizing photons could escape from the starburst to the IGM. However,

1112: our re-analysis of the \fu\, observations shows that there is no convincing

1113: evidence of Lyman continuum leakage.  This result is in contradiction to a previous

1114: report by \citet{berg06}.

1115: 

1116: Observations with \ch~of the hot diffuse gas 

1117: in \hoo~are consistent with those seen in a local sample of star forming galaxies with galactic outflows. As expected based on its far-IR luminosity and implied SFR, \hoo ~has X-ray properties intermediate between those of \ulg ~and those of more typical present-day starbursts. It has X-ray properties very similar to that of \vv, the only other LBG analogue mapped 

1118: with \ch.  This suggests

1119: that diffuse thermal X-ray emission should be a common feature of LBGs created in

1120: the shocks between the outflowing wind material and surrounding medium \citep{marc05}.

1121: 

1122: 

1123: 

1124: \acknowledgements

1125: J. G. thanks Henrique Schmitt for providing the \ha\, image of \hoo\,

1126: and Claus Leitherer for helpful discussions on the UV properties of stellar winds.   

1127: Funding for this research was provided

1128: by NASA through Chandra Proposal 8610240.

1129: 

1130: 

1131: \begin{thebibliography}{}

1132: 

1133: \bibitem[Aguirre et al.(2005)]{ag05} Aguirre, A., Schaye, 

1134: J., Hernquist, L., Kay, S., Springel, V., \& Theuns, T.\ 2005, \apjl, 620, 

1135: L13 

1136: 

1137: \bibitem[Anders \& Grevesse(1989)]{and89} Anders, E., \& 

1138: Grevesse, N.\ 1989, \gca, 53, 197

1139: 

1140: 

1141: 

1142: \bibitem[Asplund et al.(2005)]{asp05} Asplund, M., Grevesse, 

1143: N., \& Sauval, A.~J.\ 2005, ASP Conf.~Ser.~336: Cosmic Abundances as 

1144: Records of Stellar Evolution and Nucleosynthesis, 336, 25 

1145: 

1146: %\bibitem[Bauer et al.(2002)]{bauer02} Bauer, F.~E., Alexander, 

1147: %D.~M., Brandt, W.~N., Hornschemeier, A.~E., Vignali, C., Garmire, G.~P., \& 

1148: %Schneider, D.~P.\ 2002, \aj, 124, 2351 

1149: 

1150: \bibitem[Barkana \& Loeb(2006)]{bark06} Barkana, R., \& Loeb, 

1151: A.\ 2006, \mnras, 371, 395 

1152: 

1153: \bibitem[Bennett et al.(2003)]{benn03} Bennett, C.~L., et al.\ 

1154: 2003, \apjs, 148, 1 

1155: 

1156: \bibitem[Bergvall et al.(2000)]{berg00} Bergvall, N., 

1157: Masegosa, J., {\"O}stlin, G., \& Cernicharo, J.\ 2000, \aap, 359, 41 

1158: 

1159: 

1160: \bibitem[Bergvall \& $\rm{\ddot{O}}$stlin(2002)]{berg02} Bergvall, N., 

1161: 	$\rm{\ddot{O}}$stlin, G.\ 2002, \aap, 390, 891

1162: 

1163: \bibitem[Bergvall et al.(2006)]{berg06} Bergvall, N., Zackrisson, E., Andersson, B.-G., Arnberg, D., Masegoas, J., \& Ostlin, G.\

1164: 2006, A\&A, 448, 513

1165: 

1166: \bibitem[Bunker et al.(2004)]{bunk04} Bunker, A.~J., Stanway, 

1167: E.~R., Ellis, R.~S., \& McMahon, R.~G.\ 2004, \mnras, 355, 374 

1168: 

1169: 

1170: \bibitem[Cardelli et al.(1989)]{card89} Cardelli, J.~A., 

1171: Clayton, G.~C., \& Mathis, J.~S.\ 1989, \apj, 345, 245 

1172: 

1173: \bibitem[Carpenter(2001)]{carp01} Carpenter, J.~M.\ 2001, \aj,

1174: 121, 2851   

1175: 

1176: \bibitem[Chandra Interactive Analysis of Observations (CIAO)]{ciao}

1177: Chandra Interactive Analysis of Observations (CIAO), http://cxc.harvard.edu/ciao/

1178: 

1179: \bibitem[Charlot \& Longhetti(2001)]{char01} Charlot, S., \& 

1180: Longhetti, M.\ 2001, \mnras, 323, 887 

1181: 

1182: \bibitem[Colbert et al.(2004)]{colb04} Colbert, E.~J.~M., 

1183: Heckman, T.~M., Ptak, A.~F., Strickland, D.~K., \& Weaver, K.~A.\ 2004, 

1184: \apj, 602, 231 

1185: 

1186: %\bibitem[Condon et al.(1991)]{con91} Condon, J.~J., Huang, 

1187: %Z.-P., Yin, Q.~F., \& Thuan, T.~X.\ 1991, \apj, 378, 65 

1188: 

1189: %\bibitem[Davis et al.(1985)]{davis85} Davis, M., Efstathiou, 

1190: %G., Frenk, C.~S., \& White, S.~D.~M.\ 1985, \apj, 292, 371 

1191: 

1192: %\bibitem[Deharveng et al.(2001)]{deh01} Deharveng, J.-M., 

1193: %Buat, V., Le Brun, V., Milliard, B., Kunth, D., Shull, J.~M., \& Gry, C.\ 

1194: %2001, \aap, 375, 805 

1195: 

1196: %\bibitem[Dixon \& Sahnow(2003)]{dix03} Dixon, W.~V., \& 

1197: %Sahnow, D.~J.\ 2003, ASP Conf.~Ser.~295: Astronomical Data Analysis 

1198: %Software and Systems XII, 295, 241 

1199: 

1200: \bibitem[Dixon et al.(2006)]{dix06} Dixon, W.~V.~D., Sankrit, 

1201: R., \& Otte, B.\ 2006, \apj, 647, 328 

1202: 

1203: \bibitem[Dixon et al.(2007)]{dix07} Dixon, W.~V., et al.\ 

1204: 2007, \pasp, 119, 0

1205: 

1206: 

1207: %\bibitem[Dopita \& Sutherland(1996)]{dop96} Dopita, M.~A., \& 

1208: %Sutherland, R.~S.\ 1996, \apjs, 102, 161 

1209: 

1210: \bibitem[Dove et al.(2000)]{dov00} Dove, J.~B., Shull, J.~M., 

1211: \& Ferrara, A.\ 2000, \apj, 531, 846 

1212: 

1213: %\bibitem[Doyon et al.(1995)]{doy95} Doyon, R., Nadeau, D., 

1214: %Joseph, R.~D., Goldader, J.~D., Sanders, D.~B., \& Rowlands, N.\ 1995, 

1215: %\apj, 450, 111

1216: 

1217: 

1218: \bibitem[Franceschini et al.(2003)]{franc03} Franceschini, A., 

1219: et al.\ 2003, \mnras, 343, 1181 

1220: 

1221: %\bibitem[Fujita et al.(2003)]{fuj03} Fujita, A., Martin, 

1222: %C.~L., Mac Low, M.-M., \& Abel, T.\ 2003, \apj, 599, 50

1223: 

1224: %\bibitem[Fujita et al.(2004)]{fuj04} Fujita, A., Mac Low, 

1225: %M.-M., Ferrara, A., \& Meiksin, A.\ 2004, \apj, 613, 159 

1226: 

1227: \bibitem[Giavalisco(2002)]{giav02} Giavalisco, M.\ 2002, 

1228: \araa, 40, 579 

1229: 

1230: 

1231: 

1232: 

1233: %\bibitem[Gilfanov et al.(2004)]{gilf04} Gilfanov, M., Grimm, 

1234: %H.-J., \& Sunyaev, R.\ 2004, Nuclear Physics B Proceedings Supplements, 

1235: %132, 369 

1236: 

1237: 

1238: \bibitem[Grimes et al.(2006)]{grim06} Grimes, J.~P., Heckman, 

1239: T., Hoopes, C., Strickland, D., Aloisi, A., Meurer, G., \& Ptak, A.\ 2006, 

1240: \apj, 648, 310 

1241: 

1242: \bibitem[Grimes et al.(2005)]{grim05} Grimes, J.~P., Heckman, 

1243: T., Strickland, D., \& Ptak, A.\ 2005, \apj, 628, 187 

1244: 

1245: %\bibitem[Heckman, Armus, \& Miley(1990)]{heck90} Heckman,

1246: %T.~M., Armus, L., \& Miley, G.~K.\ 1990, \apjs, 74, 833

1247: 

1248: \bibitem[Heckman et al.(1998)]{heck98} Heckman, T.~M., Robert,

1249: C., Leitherer, C., Garnett, D.~R., \& van der Rydt, F.\ 1998, \apj, 503,

1250: 646

1251: 

1252: %\bibitem[Heckman et al.(2000)]{heck00} Heckman, T.~M., Lehnert, M., Strickland, D., \& Armus, L.\ 2000,

1253: %\apjs, 129, 493

1254: 

1255: \bibitem[Heckman et al.(2001a)]{heck01} Heckman, T.~M., 

1256: Sembach, K.~R., Meurer, G.~R., Strickland, D.~K., Martin, C.~L., Calzetti, 

1257: D., \& Leitherer, C.\ 2001, \apj, 554, 1021 

1258: 

1259: 

1260: 

1261: %\bibitem[Heckman et al.(2001b)]{heck01b} Heckman, T.~M., 

1262: %Sembach, K.~R., Meurer, G.~R., Leitherer, C., Calzetti, D., \& Martin, 

1263: %C.~L.\ 2001, \apj, 558, 56 

1264: 

1265: \bibitem[Heckman et al.(2002)]{heck02} Heckman, T.~M., Norman, 

1266: C.~A., Strickland, D.~K., \& Sembach, K.~R.\ 2002, \apj, 577, 691

1267: 

1268: \bibitem[Heckman(2003)]{heck03} Heckman, T.~M.\ 2003, Revista 

1269: Mexicana de Astronomia y Astrofisica Conference Series, 17, 47 

1270: 

1271: \bibitem[Heckman et al.(2005)]{heck05} Heckman, T.~M., et al.\ 

1272: 2005, \apjl, 619, L35 

1273: %

1274: \bibitem[Helsdon \& Ponman(2000)]{hels00} Helsdon, S.~F., \& 

1275: Ponman, T.~J.\ 2000, \mnras, 315, 356 

1276: 

1277: 

1278: 

1279: \bibitem[Hoopes, Heckman, Strickland, \& Howk(2003)]{hoop03}

1280: Hoopes, C.~G., Heckman, T.~M., Strickland, D.~K., \& Howk, J.~C.\ 2003,

1281: \apjl, 596, L175

1282: 

1283: %\bibitem[Hoopes et al.(2006)]{hoop06}

1284: %Hoopes, C.~G., et al. \ 2006, in prep

1285: 

1286: \bibitem[Horne(1986)]{horn86} Horne, K.\ 1986, \pasp, 98, 609 

1287: 

1288: %\bibitem[Hurwitz et al.(1997)]{hurw97} Hurwitz, M., Jelinsky, 

1289: %P., \& Dixon, W.~V.~D.\ 1997, \apjl, 481, L31

1290: 

1291: %\bibitem[Iono et al.(2004)]{iono04} Iono, D., Ho, P.~T.~P., 

1292: %Yun, M.~S., Matsushita, S., Peck, A.~B., \& Sakamoto, K.\ 2004, \apjl, 616, 

1293: %L63 

1294: 

1295: \bibitem[Jarrett et al.(2003)]{jarr03} Jarrett, T.~H.,

1296: Chester, T., Cutri, R., Schneider, S.~E., \& Huchra, J.~P.\ 2003, \aj,

1297: 125, 525

1298: 

1299: %\bibitem[Kennicutt(1998)]{kenn98} Kennicutt, R.~C.\ 1998, 

1300: %\apj, 498, 541 

1301: 

1302: \bibitem[Kim et al.(1995)]{kim95} Kim, D.-C., Sanders, D.~B., 

1303: Veilleux, S., Mazzarella, J.~M., \& Soifer, B.~T.\ 1995, \apjs, 98, 129

1304: 

1305: \bibitem[Klypin et al.(1999)]{klyp99} Klypin, A., Kravtsov, 

1306: A.~V., Valenzuela, O., \& Prada, F.\ 1999, \apj, 522, 82 

1307: 

1308: \bibitem[Kunth et al.(2003)]{kun03} Kunth, D., Leitherer, C., 

1309: Mas-Hesse, J.~M., {\"O}stlin, G., \& Petrosian, A.\ 2003, \apj, 597, 263

1310: 

1311: \bibitem[Kriss(1994)]{kriss94} Kriss, G.\ 1994, ASP Conf.~Ser.~ 

1312: 61: Astronomical Data Analysis Software and Systems III, 61, 437

1313: 

1314: \bibitem[Lehmer et al.(2005)]{lehm05} Lehmer, B.~D., et al.\ 

1315: 2005, \aj, 129, 1 

1316: 

1317: %\bibitem[Lehnert \& Heckman(1996)]{lehn96} Lehnert, M.~D.~\&

1318: %Heckman, T.~M.\ 1996, \apj, 462, 651

1319: 

1320: %\bibitem[Levenson et al.(2006)]{lev06} Levenson, N.~A.,

1321: %Heckman, T.~M., Krolik, J.~H., Weaver, K.~A., \& Zycki, P.,~T. \ 2006, submitted

1322: %

1323: %\bibitem[Leitherer et al.(1995)]{leith95} Leitherer, C., 

1324: %Ferguson, H.~C., Heckman, T.~M., \& Lowenthal, J.~D.\ 1995, \apjl, 454, L19 

1325: 

1326: \bibitem[Leitherer et al.(2002)]{leith02} Leitherer, C., Li, 

1327: I.-H., Calzetti, D., \& Heckman, T.~M.\ 2002, \apjs, 140, 303 

1328: 

1329: \bibitem[Limongi \& Chieffi(2006)]{lim06} Limongi, M., \& 

1330: Chieffi, A.\ 2006, ArXiv Astrophysics e-prints, arXiv:astro-ph/0611140 

1331: 

1332: %\bibitem[Loeb \& Barkana(2001)]{loeb01} Loeb, A., \& Barkana, 

1333: %R.\ 2001, \araa, 39, 19 

1334: 

1335: %\bibitem[Malkan et al.(2003)]{malk03} Malkan, M., Webb, W., \& 

1336: %Konopacky, Q.\ 2003, \apj, 598, 878 

1337: 

1338: \bibitem[Marcolini et al.(2005)]{marc05} Marcolini, A., 

1339: Strickland, D.~K., D'Ercole, A., Heckman, T.~M., \& Hoopes, C.~G.\ 2005, 

1340: \mnras, 362, 626 

1341: 

1342: %\bibitem[Martin(2005)]{mart05b} Martin, C.~L.\ 2005, \apj, 621, 227

1343: 

1344: \bibitem[Martin et al.(2005)]{mart05} Martin, D.~C., et al.\ 

1345: 2005, \apjl, 619, L1 

1346: 

1347: \bibitem[Moos et al.(2000)]{moos00} Moos, H.~W., et al.\ 2000, 

1348: \apjl, 538, L1 

1349: 

1350: \bibitem[Nandra et al.(2002)]{nan02} Nandra, K., Mushotzky, 

1351: R.~F., Arnaud, K., Steidel, C.~C., Adelberger, K.~L., Gardner, J.~P., 

1352: Teplitz, H.~I., \& Windhorst, R.~A.\ 2002, \apj, 576, 625 

1353: 

1354: \bibitem[{\"O}stlin et al.(2001)]{ost01} {\"O}stlin, G., 

1355: Amram, P., Bergvall, N., Masegosa, J., Boulesteix, J., \& M{\'a}rquez, I.\ 

1356: 2001, \aap, 374, 800 

1357: 

1358: %\bibitem[Oskinova et al.(2006)]{osk06} Oskinova, L.~M.,

1359: %Feldmeier, A, Hamann, W.~R. (astro-ph 0603286)

1360: 

1361: \bibitem[Otte et al.(2003)]{ott03} Otte, B., Murphy, E.~M., 

1362: Howk, J.~C., Wang, Q.~D., Oegerle, W.~R., \& Sembach, K.~R.\ 2003, \apj, 

1363: 591, 821 

1364: 

1365: \bibitem[Ouchi et al.(2004)]{ouchi04} Ouchi, M., et al.\ 2004, 

1366: \apj, 611, 685 

1367: 

1368: %\bibitem[Otte \& Dixon(2006)]{ott06} Otte, B., \& Dixon, 

1369: %W.~V.~D.\ 2006, \apj, 647, 312 

1370: 

1371: %\bibitem[Panagia et al.(2005)]{pan05} Panagia, N., Fall, 

1372: %S.~M., Mobasher, B., Dickinson, M., Ferguson, H.~C., Giavalisco, M., Stern, 

1373: %D., \& Wiklind, T.\ 2005, \apjl, 633, L1

1374: 

1375: \bibitem[Peacock et al.(2000)]{peac00} Peacock, J.~A., et al.\ 

1376: 2000, \mnras, 318, 535 

1377: 

1378: \bibitem[Pellerin et al.(2002)]{pell02} Pellerin, A., et al.\ 

1379: 2002, \apjs, 143, 159 

1380: 

1381: %\bibitem[Pettini et al.(2002)]{pett02} Pettini, M., Rix, 

1382: %S.~A., Steidel, C.~C., Adelberger, K.~L., Hunt, M.~P., \& Shapley, A.~E.\ 

1383: %2002, \apj, 569, 742 

1384: 

1385: \bibitem[Robert et al.(2003)]{rob03} Robert, C., Pellerin, 

1386: A., Aloisi, A., Leitherer, C., Hoopes, C., \& Heckman, T.~M.\ 2003, \apjs, 

1387: 144, 21 

1388: 

1389: \bibitem[Robertson et al.(2005)]{rob05} Robertson, B., 

1390: Bullock, J.~S., Font, A.~S., Johnston, K.~V., \& Hernquist, L.\ 2005, \apj, 

1391: 632, 872 

1392: 

1393: %\bibitem[Rupke et al.(2005)]{rup05} Rupke, D.~S., Veilleux, S., \& Sanders, D.~B. 2005, \apjs,

1394: %160, 115

1395: 

1396: \bibitem[Sanders \& Mirabel (1996)]{sand96} Sanders, D.~B.~\&

1397: Mirabel, I.~F.\ 1996, \araa, 34, 749  

1398: 

1399: \bibitem[Sanders et al.(2003)]{sand03} Sanders, D.~B., 

1400: Mazzarella, J.~M., Kim, D.-C., Surace, J.~A., \& Soifer, B.~T.\ 2003, \aj, 

1401: 126, 1607 

1402: 

1403: %\bibitem[Sako et al.(2002)]{sak02} Sako, M., Kahn, S.~M., 

1404: %Paerels, F., Liedahl, D.~A., Watanabe, S., Nagase, F., \& Takahashi, T.\ 

1405: %2002, High Resolution X-ray Spectroscopy with XMM-Newton and Chandra

1406: 

1407: %\bibitem[Scannapieco et al.(2005)]{scann05} Scannapieco, C., 

1408: %Tissera, P.~B., White, S.~D.~M., \& Springel, V.\ 2005, \mnras, 907 

1409: 

1410: %\bibitem[Schild et al.(2004)]{schil04} Schild, H., et al.\ 

1411: %2004, \aap, 422, 177 

1412: 

1413: \bibitem[Schlegel et al.(1998)]{schleg98} Schlegel, D.~J., 

1414: Finkbeiner, D.~P., \& Davis, M.\ 1998, \apj, 500, 525 

1415: 

1416: 

1417: \bibitem[Schmitt et al.(2006)]{schmitt06} Schmitt, H.~R., 

1418: Calzetti, D., Armus, L., Giavalisco, M., Heckman, T.~M., Kennicutt, R.~C., 

1419: Jr., Leitherer, C., \& Meurer, G.~R.\ 2006, \apjs, 164, 52 

1420: 

1421: 

1422: \bibitem[Seibert et al.(2005)]{seib05} Seibert, M., et al.\ 

1423: 2005, \apjl, 619, L55

1424: 

1425: 

1426: \bibitem[Shapley et al.(2006)]{shap06} Shapley, A.~E., 

1427: Steidel, C.~C., Pettini, M., Adelberger, K.~L., \& Erb, D.~K.\ 2006, \apj, 

1428: 651, 688 

1429: 

1430: \bibitem[Shapley et al.(2003)]{shap03} Shapley, A.~E., 

1431: Steidel, C.~C., Pettini, M., \& Adelberger, K.~L.\ 2003, \apj, 588, 65 

1432: 

1433: \bibitem[Soifer, Boehmer, Neugebauer, \& Sanders(1989)]{soif89} 

1434: Soifer, B.~T., Boehmer, L., Neugebauer,

1435: G., \& Sanders, D.~B.\ 1989, \aj, 98, 766

1436: 

1437: \bibitem[Sommer-Larsen et al.(1999)]{som99} Sommer-Larsen, 

1438: J., Gelato, S., \& Vedel, H.\ 1999, \apj, 519, 501 

1439: 

1440: %\bibitem[Springel et al.(2005)]{spring05} Springel, V., et al.\ 

1441: %2005, \nat, 435, 629 

1442: 

1443: \bibitem[Steidel et al.(1999)]{steid99} Steidel, C.~C., 

1444: Adelberger, K.~L., Giavalisco, M., Dickinson, M., \& Pettini, M.\ 1999, 

1445: \apj, 519, 1 

1446: 

1447: %\bibitem[Steidel et al.(2001)]{steid01} Steidel, C.~C., 

1448: %Pettini, M., \& Adelberger, K.~L.\ 2001, \apj, 546, 665 

1449: 

1450: %\bibitem[Stiavelli et al.(2004)]{sti04} Stiavelli, M., Fall, 

1451: %S.~M., \& Panagia, N.\ 2004, \apjl, 610, L1 

1452: 

1453: \bibitem[Strickland et al.(2000)]{strick00b} Strickland, D.~K., 

1454: Heckman, T.~M., Weaver, K.~A., \& Dahlem, M.\ 2000, \aj, 120, 2965 

1455: 

1456: 

1457: \bibitem[Strickland \& Stevens(2000)]{strick00} Strickland, 

1458: D.~K., \& Stevens, I.~R.\ 2000, \mnras, 314, 511

1459: 

1460: \bibitem[Strickland et al.(2004)]{strick04b} Strickland, D.~K., 

1461: Heckman, T.~M., Colbert, E.~J.~M., Hoopes, C.~G., \& Weaver, K.~A.\ 2004, 

1462: \apj, 606, 829 

1463: 

1464: \bibitem[Strickland(2005)]{strick05} Strickland, D.~K.\ 2005, 

1465: Extra-Planar Gas, 331, 345 

1466: 

1467: %

1468: %\bibitem[Sutherland \& Dopita(1993)]{suth93} Sutherland, 

1469: %R.~S., \& Dopita, M.~A.\ 1993, \apjs, 88, 253 

1470: 

1471: \bibitem[Tamura et al.(2004)]{tam04} Tamura, T., Kaastra, 

1472: J.~S., den Herder, J.~W.~A., Bleeker, J.~A.~M., \& Peterson, J.~R.\ 2004, 

1473: \aap, 420, 135 

1474: 

1475: \bibitem[Tody \& Fitzpatrick(1996)]{tod96} Tody, D., \& 

1476: Fitzpatrick, M.\ 1996, Astronomical Data Analysis Software and Systems V, 

1477: 101, 322 

1478: 

1479: %\bibitem[Tremonti et al.(2004)]{trem04} Tremonti, C.~A., et 

1480: %al.\ 2004, \apj, 613, 898 

1481: 

1482: \bibitem[Vader et al.(1993)]{vad93} Vader, J.~P., Frogel, 

1483: J.~A., Terndrup, D.~M., \& Heisler, C.~A.\ 1993, \aj, 106, 1743 

1484: 

1485: \bibitem[V{\'a}zquez \& Leitherer(2005)]{vaz05} V{\'a}zquez, 

1486: G.~A., \& Leitherer, C.\ 2005, \apj, 621, 695 

1487: 

1488: 

1489: \bibitem[Veilleux et al.(2005)]{veill05} Veilleux, S., Cecil, 

1490: G., \& Bland-Hawthorn, J.\ 2005, \araa, 43, 769 

1491: 

1492: \bibitem[York et al.(2000)]{york00} York, D.~G., et al.\ 2000, 

1493: \aj, 120, 1579 

1494: 

1495: 

1496: 

1497: \end{thebibliography}

1498: 

1499: %%log(L_FUV)=10.39 L_sun, giving a FUV surface brightness of log(I_FUV)=8.89 L_sun/kpc2

1500: %% ($10^{10.3}\,\rm{L}_{\odot}$) and surface 

1501: %%brightness ($10^{9.5}\,\rm{L_{\odot}\,kpc^{-2}}$)

1502: 

1503: \clearpage

1504: 

1505: \begin{deluxetable}{lcccccccccc}

1506: \rotate 

1507: \tabletypesize{\scriptsize}

1508: \tablecolumns{11}

1509: \tablecaption{Comparison of \hoo\, versus \vv\label{prop}}

1510: \tablehead{

1511: \colhead{} & \multicolumn{2}{c}{Position} & \colhead{Distance} & 

1512: \colhead{$\rm{v_{sys}}$} &

1513: \colhead{Scale} & 

1514: \colhead{Z} &

1515: \colhead{$\rm{L_{FIR}}$\tablenotemark{a}} & \colhead{$\rm{L_{K}}$\tablenotemark{a}} &

1516: \colhead{$\rm{L_{FUV}}$\tablenotemark{b}}  &

1517: \colhead{$\rm{I_{FUV}}$\tablenotemark{c}} \\

1518: \colhead{} & \multicolumn{2}{c}{J2000} &  \colhead{Mpc} & \colhead{\kms} & \colhead{kpc arcsec$^{-1}$} & &

1519: \colhead{$\rm{L_\odot}$} & \colhead{$\rm{L_\odot}$} & \colhead{$\rm{L_\odot}$}  &  

1520: \colhead{$\rm{L_\odot\,kpc^{-2}}$} }

1521: \startdata

1522: \hoo & +00 36 52.5 & -33 33 19 & 88 & 6180 & 0.41 & 7.9\tablenotemark{c} & $6.7\times10^{10}$ & $4.5\times10^{9}$ & $2.0\times10^{10}$ & $2.8\times10^{9}$\tablenotemark{d}\\

1523: \vv & +01 07 47.1 & -17 30 24 & 86 & 5970\tablenotemark{e} & 0.40 & 8.6-8.7\tablenotemark{f} & $2.6\times10^{11}$ & $2.0\times10^{10}$ & $2.5\times10^{10}$ 

1524: & $7.8\times10^{8}$\tablenotemark{g}\\

1525: \enddata

1526: %\tablenotetext{a}{from NASA/IPAC extragalactic database (NED)}

1527: \tablenotetext{a}{IRAS \citep{soif89}  and 2MASS \citep{jarr03} results are transformed using methods of \citet{sand96} and \citet{carp01} respectively.}

1528: \tablenotetext{b}{Derived from the \fu\,continuum flux at 1150\,\AA}

1529: \tablenotetext{c}{Gas phase metallicity as log(O/H)+12 from \citet{berg02}.}

1530: \tablenotetext{d}{ACS observation}

1531: \tablenotetext{e}{\citet{grim06}}

1532: \tablenotetext{f}{Gas phase metallicity as log(O/H)+12 from \citet{kim95} using the transformations of \citet{char01}.}

1533: \tablenotetext{g}{GALEX observation}

1534: \end{deluxetable}

1535: 

1536: 

1537: 

1538: \begin{deluxetable}{ccclcccccc}

1539: \rotate

1540: \tabletypesize{\scriptsize}

1541: \tablecolumns{10}

1542: \tablewidth{0pc}

1543: \tablecaption{\hoo\, X-Ray Spectral Fit\tablenotemark{a}}

1544: \tablehead{

1545: \colhead{$\rm{N_H}$\tablenotemark{b}} & \colhead{kT}   & \colhead{$K$\tablenotemark{c}} & \colhead{$\alpha/\rm{Fe}$} & 

1546: \colhead{$N_{H}$} & \colhead{PL Norm\tablenotemark{d}}   

1547: & \colhead{$\Gamma$} & 

1548: \colhead{$\chi^2/\rm{DOF}$} & 

1549: \colhead{$\rm{L_{0.3-2.0\,keV,thermal}}$} & \colhead{$\rm{L_{2.0-8.0\,keV,powerlaw}}$}\\

1550:  \colhead{$10^{22}~\rm{cm^{-2}}$} & \colhead{keV}    & \colhead{} & \colhead{} &

1551: \colhead{$10^{22}~\rm{cm^{-2}}$}   & \colhead{}    & \colhead{} & \colhead{} &

1552: \colhead{\ergs} & \colhead{\ergs}

1553: \label{xspecdata}}

1554: \startdata

1555: $1.9\times10^{-2}$ & $0.68^{+0.05}_{-0.03}$ & $6.0^{+1.4}_{-4.1}\times10^{-5}$ & $4.0^{+1.1}_{-1.2}$ & 

1556: $0.5^{+0.3}_{-0.5}$ & $3.3^{+1.3}_{-1.6}\times10^{-5}$ & $1.7^{+0.3}_{-0.3}$ & 61/65 & $7.2\times10^{40}$ & $1.0\times10^{41}$\\\

1557: \enddata

1558: \tablenotetext{a}{Values derived from an xspec model of  wabs ( vmekal + zwabs ( powerlaw )) using the default abundance settings \citep{and89}.} 

1559: \tablenotetext{b}{Galactic $\rm{N_H}$ value fixed to COLDEN result}

1560: \tablenotetext{c}{Plasma model normalization in units of 

1561: $\frac{10^{-14}}{4\pi [D_A (1+z)]^2}\int n_e n_H dV$, where $D_A$ is the angular distance, and $n_H$ and $n_e$ are the hydrogen and electron number densities respectively.}

1562: \tablenotetext{d}{$\rm{photons~keV^{-1}\,cm^{-2}\, s^{-1}}$ at 1 keV}

1563: \end{deluxetable}

1564: 

1565: %

1566: %\rm{L_{0.3-2.0\,keV,thermal}\sim7.2\times10^{40}\,erg\,s^{-1}}$

1567: %for the soft thermal component and \linebreak

1568: %$\rm{L_{2.0-8.0\,keV,powerlaw}\sim1.0\times10^{41}\,erg\,s^{-1}}$

1569: 

1570: \clearpage

1571: 

1572: \begin{deluxetable}{lcccccc}

1573: \tabletypesize{\footnotesize}

1574: \tablecolumns{7}

1575: \tablewidth{0pc}

1576: \tablecaption{ISM Absorption Line Fit Data}

1577: \tablehead{

1578: \colhead{Ion} & \colhead{$\lambda_{0}$} & \colhead{$log(\lambda f N/N_H)$} &

1579:                  \colhead{Instrument} & \colhead{$\rm{W}_\lambda$} & 

1580:                  \colhead{$v_{c}$} & \colhead{FWHM} \\

1581:                   & \colhead{$\ang$} & & & \colhead{$\ang$} & \colhead{\kms} & 

1582:                   \colhead{\kms}\label{absdata}}

1583: \startdata

1584: Ly~$\beta$\tablenotemark{b} & 1025.722 & 1.91 & LiF 1A &  $2.3\pm0.1$ & $6055\pm9$ & $557\pm31$\\

1585:  & & & LiF 2B &  $2.3\pm0.1$ & $6051\pm9$ & $529\pm54$\\

1586: \ion{C}{3}\tablenotemark{b} & 977.02 & -0.61 & LiF 1A &  $0.9\pm0.2$ & 6081\tablenotemark{a} & $169\pm18$\\

1587:  & & & LiF 2B &  $0.7\pm1.4$ & $6133$\tablenotemark{a} & $87\pm14$\\

1588:  \ovi & 1031.926 & -1.13 & LiF 1A &  $0.4\pm0.1$ & $6042\pm16$ & $193\pm35$\\

1589:  & & & LiF 2B &  $0.4\pm0.1$ & $6051\pm15$ & $159\pm34$\\

1590:  \ion{C}{2} & 1036.337 & -1.37 & LiF 1A &  $1.4\pm0.1$ & $6061\pm6$ & $402\pm19$\\

1591:  & & & LiF 2B &  $1.2\pm0.1$ & $6075\pm8$ & $341\pm30$\\

1592: \ion{O}{1} & 988.733 & -1.56 & LiF 1A &  $0.7\pm0.1$ & $6092\pm13$ & $275\pm15$\\

1593:  & & & LiF 2B &  $0.7\pm0.1$ & $6079\pm23$ & $304\pm26$\\

1594: \ion{N}{2} & 1083.99 & -2.00 & LiF 1B &  $0.9\pm0.1$ & $6081\pm10$ & $352\pm28$\\

1595:  & & & LiF 2A &  $1.0\pm0.1$ & $6085\pm11$ & $381\pm33$\\

1596: \ion{N}{3} & 989.799 & -2.03 &  LiF 1A & $1.0\pm0.1$ & $6076\pm6$ & $275\pm137$\\

1597:  & & & LiF 2B &  $1.1\pm0.1$ & $6073\pm10$ & $304\pm152$\\

1598: \ion{N}{1} & 1134.415 & -2.09 & LiF 2A &  $0.2\pm0.1$ & $6074$\tablenotemark{a} & $103\pm36$\\

1599: \ion{O}{1} & 1039.23 & -2.29 & LiF 1A &   $0.3\pm0.1$ & $6110\pm5$ & $207\pm47$\\

1600:  & & & LiF 2B &  $0.3\pm0.1$ & $6135\pm13$ & $192\pm46$\\

1601: \ion{Fe}{2} & 1096.877 & -2.96 & LiF 1B &  $0.3\pm0.1$ & $6101\pm15$ & $197\pm72$\\

1602:  & & & LiF 2A &  $0.2\pm0.1$ & $6127\pm17$ & $143\pm40$\\

1603:  \ion{Si}{2} & 1020.699 & -2.989 & LiF 1A &  $0.1\pm0.1$ & $6071\pm37$ & $87\pm83$\\

1604:  & & & LiF 2B &  $0.2\pm0.1$ & $6071\pm12$ & $100\pm23$\\

1605:  \ion{Fe}{2} & 1121.975 & -3.15 & LiF 2A &  $0.3\pm0.2$ & $6088\pm68$ & $203\pm193$\\

1606: \ion{S}{3} & 1012.502 & -3.24 & LiF 1A &  $0.5\pm0.2$ & $6083\pm6$ & $225\pm108$\\

1607:  & & & LiF 2B &  $0.6\pm0.2$ & $6078\pm13$ & $290\pm75$\\

1608: \ion{Fe}{2} & 1125.448 & -3.41 & LiF 2A &  $0.2\pm0.1$ & $6062\pm17$ & $195\pm37$\\

1609: \ion{Ar}{1} & 1066.66 & -3.59 & LiF 2A &  $0.5\pm0.2$ & $6136\pm27$ & $332\pm82$\\

1610: \enddata

1611: \tablenotetext{a}{Errors are undetermined for this value.}

1612: \tablenotetext{b}{$\rm{Ly}$~$\beta$ and \ion{C}{3} are saturated at line center.}

1613: %\tablenotetext{c}{\ion{Fe}{2}~$\lambda$1144 line measurements are strongly affected by the second order airglow emission line \ion{He}{1}~$\lambda$584 

1614: %and an instrumental artifact that causes a break in the continuum}

1615: \end{deluxetable}

1616: 

1617: \clearpage

1618: \begin{deluxetable}{lccccc}

1619: \tabletypesize{\footnotesize}

1620: \tablecolumns{6}

1621: \tablewidth{0pc}

1622: \tablecaption{Stellar Photospheric Absorption Line Fit Data}

1623: \tablehead{

1624: \colhead{Ion} & \colhead{$\lambda_{0}$} &

1625:                  \colhead{Instrument} & \colhead{$\rm{W}_\lambda$} & 

1626:                  \colhead{$v_{c}$} & \colhead{FWHM} \\

1627:                   & \colhead{$\ang$} & & \colhead{$\ang$} & \colhead{\kms} & 

1628:                   \colhead{\kms}\label{phabsdata}}

1629: \startdata

1630: \ion{Si}{4} & 1122.487 & LiF 2A & $0.6\pm0.1$ & $6163\pm17$ & $222\pm29$\\

1631: \ion{Si}{4} & 1128.201 & LiF 2A & $0.3$\tablenotemark{a} & $6202$\tablenotemark{a} & $327$\tablenotemark{a}\\

1632: \ion{P}{5} & 1117.977 & LiF 2A & $0.3\pm0.1$ & $6165\pm24$ & $320\pm63$\tablenotemark{a}\\

1633: \ion{P}{5} & 1128.0 & LiF 2A & $0.1$\tablenotemark{a} & $6202$\tablenotemark{b} & $327$\tablenotemark{b}\\

1634: \enddata

1635: \tablenotetext{a}{Errors are undetermined for this value.}

1636: \tablenotetext{b}{Values of the velocity shift and FWHM of \ion{P}{5} $\lambda$1128 are tied to 

1637: the corresponding parameters in \ion{Si}{4} $\lambda$1128}

1638: \end{deluxetable}

1639: 

1640: \clearpage

1641: \begin{deluxetable}{lccccccccccc}

1642: \rotate 

1643: \tabletypesize{\scriptsize}

1644: \tablecolumns{12}

1645: \tablewidth{0pc}

1646: \tablecaption{Two Component Absorption Line Fits}

1647: \tablehead{

1648: & & & & & \multicolumn{3}{c}{Galaxy} & & \multicolumn{3}{c}{Outflow} \\

1649: \cline{6-8} \cline{10-12}

1650: \colhead{Ion} & \colhead{$\lambda_{0}$} & \colhead{$log(\lambda f N/N_H)$} &

1651:                  \colhead{Instrument} & \phantom{ } & \colhead{$\rm{W}_\lambda$} & 

1652:                  \colhead{$v_{c}$} & \colhead{FWHM} & \phantom{ } & \colhead{$\rm{W}_\lambda$} & 

1653:                  \colhead{$v_{c}$} & \colhead{FWHM} \\

1654:                   & \colhead{$\rm{\AA}$} & & & & \colhead{$\rm{\AA}$} & \colhead{\kms} & 

1655:                   \colhead{\kms} & & \colhead{$\rm{\AA}$} & \colhead{\kms} & 

1656:                   \colhead{\kms}\label{twoabsdata}}

1657: \startdata

1658: Ly~$\beta$\tablenotemark{a} & 1025.722 & 1.91 & LiF 1A &  & $2.22\pm0.07$ & $6060\pm7$ & $519\pm19$ & & $0.25\pm0.09$ & $5781\pm18$ & $124\pm40$\\

1659:  & & & LiF 2B & &$1.91\pm0.15$ & $6103\pm20$ & $467\pm31$ & & $0.56\pm0.16$ & $5827\pm21$ & $184\pm47$\\ 

1660: \ion{C}{3}\tablenotemark{a}  & 977.02 & -0.61 & LiF 1A &  & $0.85\pm0.18$ & $6084\pm6$ & $146\pm13$ & & $0.44\pm0.18$ & $5840\pm7$ & $60$\tablenotemark{b} \\

1661:  & & & LiF 2B & &$0.69$\tablenotemark{b} & $6133\pm4$ & $83\pm9$ & & $0.36$\tablenotemark{b}  & $5894\pm6$ & $53\pm25$\\ 

1662: \ion{C}{2} & 1036.337 & -1.374 & LiF 1A &  & $1.22\pm0.09$ & $6076\pm8$ & $356\pm28$ & & $0.11\pm0.05$ & $5855\pm27$ & $100$\tablenotemark{b} \\

1663:  & & & LiF 2B & &$1.18\pm0.10$ & $6084\pm12$ & $325\pm26$ & & $0.10\pm0.17$ & $5866\pm70$ & $104\pm164$\\ 

1664: \ion{N}{2} & 1083.99 & -2.002 & LiF 1B &  & $0.79\pm0.14$ & $6104\pm12$ & $310\pm50$ & & $0.13\pm0.05$ & $5935\pm11$ & $88\pm26$\\

1665:  & & & LiF 2A & & $1.05\pm0.20$ & $6114\pm20$ & $392\pm63$ & & $0.07\pm0.20$ & $5921\pm39$ & $123$\tablenotemark{b} \\

1666: \ion{N}{3}\tablenotemark{c} & 989.799 & -2.03 & LiF 1B &  & $0.81\pm0.04$ & $6099\pm11$ & $217\pm27$ & & $0.35\pm0.13$ & $5957\pm22$ & $150\pm40$\\

1667:  & & & LiF 2A & & $0.78\pm0.22$ & $6120\pm31$ & $268\pm41$ & & $0.41\pm0.17$ & $6022\pm17$ & $128\pm38$\\

1668: \enddata

1669: \tablenotetext{a}{The galaxy and/or outflow component of this line is saturated.}

1670: \tablenotetext{b}{Unable to determine errors on this value.}

1671: \tablenotetext{c}{\ion{N}{3}~$\lambda$990 is blended with \ion{O}{1}~$\lambda$989.  

1672: They are fit simultaneously with the tied values of the fwhm and relative velocity.}

1673: \end{deluxetable}

1674: 

1675: 

1676: \begin{deluxetable}{lcccccc}

1677: \tabletypesize{\footnotesize}

1678: \tablecolumns{7}

1679: \tablewidth{0pc}

1680: \tablecaption{Mass and Energy Outflow Rates}

1681: \tablehead{

1682: \colhead{Ion}  &  \colhead{$\rm{N_{Ion}}$\tablenotemark{a}} & 

1683: \colhead{Abundance\tablenotemark{b}} & \colhead{$\rm{N_H}$}   & 

1684: \colhead{$\frac{\Omega}{4\pi}\,\dot{\rm{M}}$\tablenotemark{c}} & 

1685: \colhead{$\frac{\Omega}{4\pi}\dot{\rm{E}}$\tablenotemark{c}}\\

1686: \colhead{} & \colhead{$10^{14}~\rm{cm^{-2}}$}   & \colhead{$\rm{\log(N_{Elem}/N_H)+12}$}    & \colhead{$10^{18}~\rm{cm^{-2}}$} & 

1687: \colhead{$\rm{M_\odot/yr}$} & \colhead{$\rm{10^{40}~erg/s}$}\label{outflow}}

1688: \startdata

1689: %\ion{O}{6} & 4.2 & 8.66 & $>$23 & $>$2.6 & $>$13 \\

1690: \ion{N}{2} & 1.4 & 7.78 & 2.3 & 0.1 & 0.2 \\

1691: \ion{N}{3} & 5.7 & 7.78 & 9.5 & 0.3 & 0.7 \\

1692: \ion{C}{2} & 3.6 & 8.39 & 1.5 & 0.1 & 0.1\\

1693: Total Cool Gas\tablenotemark{d}     & & & $>$11.8 & $>$0.4 & $>$0.9\\

1694: \cline{1-6} 

1695: Total Hot Gas\tablenotemark{e}       & & &  & $18f^{1/2}$ & $130f^{1/2}$\\ 

1696: \enddata

1697: \tablenotetext{a}{Based on outflow equivalent width and FWHM from Table \ref{twoabsdata}.}

1698: \tablenotetext{b}{\citet{asp05}}

1699: \tablenotetext{c}{Estimated mass and kinetic energy outflow rates (see text).} 

1700: \tablenotetext{d}{From \fu\, data using \ion{N}{2} and \ion{N}{3}. } 

1701: \tablenotetext{e}{From X-ray data.  Note that f is the volume filling factor of the hot gas.} 

1702: \end{deluxetable}

1703: 

1704: \clearpage

1705: 

1706: \begin{figure}

1707: \centering

1708: \leavevmode

1709: \includegraphics[width=3in]{f1.eps}

1710: \caption{

1711: \fu\, LWRS aperture overlaid on an HST/ACS 

1712: FUV (F140LP) continuum image of \hoo.    We have also 

1713: labeled the three knots (A-C) as in \citet{vad93}.  A fourth knot,  identified by \citet{kun03}

1714: is labeled D, although it is not visible in this image.

1715: \label{acs}}

1716: \end{figure}

1717: 

1718: \begin{figure}

1719: \centering

1720: \leavevmode

1721: \includegraphics[width=5in]{f2.eps}

1722: \caption{

1723: \ch\, false color (red 0.3-1.0 keV, green 1.0-2.0 keV, blue 2.0-8.0 keV) and \ha\, \citep{schmitt06} 

1724: images of \hoo.

1725: The two images have a similar morphology.  Knot B, likely the galaxy nucleus, is

1726: the dominant feature in both the X-ray and \ha.  The \ch\, image suggests

1727: a narrow band/disk of absorption running NE to SW through knot B.  This absorption feature

1728:  also appears in the ACS FUV image (Figure \ref{acs}).

1729: Knot D (SE of B), observed by \citet{kun03} in \lya\,emission, is only seen in the \ch\, image.  

1730: \label{chandrafalse}}

1731: \end{figure}

1732: 

1733: \begin{figure}

1734: \centering

1735: \leavevmode

1736: \includegraphics[width=3in,angle=-90]{f3.eps}

1737: \caption{

1738: \ch\, spectrum and model of \hoo.  The model fit parameters can be found in 

1739: Table \ref{xspecdata}.

1740: \label{chandraspec}}

1741: \end{figure}

1742: 

1743: 

1744: \begin{figure}

1745: \centering

1746: \leavevmode

1747: \includegraphics[width=2in,angle=-90]{f4.eps}

1748: \caption{

1749: Contour plot (1-3 $\sigma$) of the variance in $\chi^2$ as a function of

1750: \alphe\,and Fe abundances (relative to solar) from

1751: the {\it xspec} fit to the X-ray spectrum.  Although it is impossible to 

1752: determine the absolute abundances, 

1753: the \alphe\, abundance (dominated by oxygen) in the hot gas

1754: is significantly enhanced relative to the 20\% oxygen gas phase metallicity

1755: derived in \citet{berg02}.     

1756: \label{contour}}

1757: \end{figure}

1758: 

1759: 

1760: 

1761: \begin{figure}

1762: \centering

1763: \leavevmode

1764: \includegraphics[width=4in,angle=90]{f5.eps}

1765: \caption{

1766: SiC 2A, LiF 1A, and LiF 2A spectra of \hoo\, covering the majority

1767: of the \fu\, spectral range.  Prominent \hoo\,absorption lines, milky way

1768: lines, and airglow features (earth symbol) have been identified.

1769: \label{fullspectra}}

1770: \end{figure}

1771: 

1772: \begin{figure}

1773: \centering

1774: \leavevmode

1775: \includegraphics[width=5in,angle=90]{f6.eps}

1776: \caption{

1777: Expanded plots of five strong absorption features in the \fu\,spectra

1778: of \hoo.  All of the absorption lines except 

1779: \ion{O}{6} show signs of one strong, broad absorption line at $\sim6100$ \kms\,

1780: and another weaker absorption line at $\sim5900$ \kms.  

1781: Both are significantly blueshifted relative to

1782: the galaxy systemic velocity of $6180$ \kms.

1783: We have overlaid two component absorption line fits for the

1784: \ion{N}{2}, \ion{C}{3}, \ion{C}{2}, and $\rm{Ly\,\beta}$ lines and

1785: vertical lines at $5900$ and $6100$ \kms\, for visual comparison.  

1786: The absorption feature at $5900$ \kms\, is clearly visible in the spectra

1787: although there is some variability in the centroid velocity.

1788: The \ion{O}{6}~$\lambda$1032

1789: line has a P-Cygni profile as it has absorption on the blue wing and emission

1790: on the red side.  For

1791: \ion{O}{6} we have plotted a single absorption line although it does not

1792: well represent its non gaussian profile.  \ion{O}{6}~$\lambda$1038

1793: emission (and possibly absorption) is  observed in the \ion{C}{2}~$\lambda$1036 plot.  

1794: \label{absprofiles}}

1795: \end{figure}

1796: 

1797: 

1798: \begin{figure}

1799: \centering

1800: \leavevmode

1801: \includegraphics[width=3in,angle=90]{f7.eps}

1802: \caption{

1803: SiC 1B and SiC 2A night spectra (binned by 12 pixels=0.156\,\AA) of the Lyman continuum region.  Both spectra have

1804: been run through the standard \fu\, data pipeline.  The SiC 1B background

1805: has clearly been oversubtracted in this wavelength region resulting in a negative flux.

1806: \label{sic1b2aorig}}

1807: \end{figure}

1808: 

1809: \begin{figure}

1810: \centering

1811: \leavevmode

1812: \includegraphics[width=4in]{f8.eps}

1813: \caption{

1814: SiC 1B and SiC 2A image of the detectors in the Lyman continuum region during 

1815: orbital night.  The 

1816: point source extraction aperture has been overlayed for both detectors.  No clear

1817: emission can be seen in the Lyman continuum region on either detector.  

1818: Background subtraction of the SiC 1B spectra is complicated by its location near the lower  edge of 

1819: the detector.  Vertical structure in the background can also been seen on both detectors.

1820: \label{detector}}

1821: \end{figure}

1822: 

1823: 

1824: 

1825: 

1826: \begin{figure}

1827: \centering

1828: \leavevmode

1829: \includegraphics[width=4in,angle=90]{f9.eps}

1830: \caption{

1831: We have extracted the number of counts in 1\AA\, bins from a rectangular region

1832: enclosing the source aperture (black line) and background regions directly

1833: below (dotted) and directly above (dashed) the aperture.

1834: The background profiles are very similar to the source profile 

1835: below the Lyman limit.  At variance with \citet{berg06} there does not appear to be any excess counts

1836: in the source regions below the Lyman limit.  Note that several Lyman series absorption

1837: lines are visible above the Lyman limit.  These profiles are consistent with a visual

1838: analysis of the  2-D spectra shown in Figure \ref{detector}. 

1839: \label{radialprofile}}

1840: \end{figure}

1841: 

1842: \begin{figure}

1843: \centering

1844: \leavevmode

1845: \includegraphics[width=4in]{f10.eps}

1846: \caption{

1847: SiC 2A night image of the detector with the SiC 2A airglow regions identified by CalFUSE 3.1.8.  

1848: \label{sic2aairglow}}

1849: \end{figure}

1850: 

1851: 

1852: \begin{figure}

1853: \centering

1854: \leavevmode

1855: \includegraphics[width=3in,angle=90]{f11.eps}

1856: \caption{

1857: Total counts (spectra + background) from optimal extraction 

1858: (binned by 12 pixels=0.156\,\AA) in the SiC 2A night time observation as a function of wavelength.  The

1859: fitted background is shown as a dashed line.  This plot again shows that the background

1860: dominates any signal in the Lyman continuum region.

1861: \label{sic2acounts}}

1862: \end{figure}

1863: 

1864: 

1865: \begin{figure}

1866: \centering

1867: \leavevmode

1868: \includegraphics[width=3in,angle=90]{f12.eps}

1869: \caption{

1870: This plot shows the count rate in three wavelength regions of the LWRS aperture for the 

1871: SiC 2A detector as a function of earth limb angle.  

1872: The 920-930\,\AA~ bandpass, which represents the Lyman continuum emission region, 

1873: has the lowest count rate for all limb angles and appears to still be 

1874: falling at the highest angles.  The expected intrinsic count rate of the detector from 

1875: cosmic rays and $^{40}\rm{K}$ decays is $\sim 0.05$ counts $\rm{s^{-1}}$ which

1876:  is just below the count rate seen at the highest

1877: earth limb angles.  This suggests that any excess emission in the SiC 2A 

1878: spectra below the Lyman limit can be attributed to airglow.

1879: \label{earthlimb}}

1880: \end{figure}

1881: 

1882: 

1883: \begin{figure}

1884: \centering

1885: \leavevmode

1886: \includegraphics[width=3in,angle=90]{f13.eps}

1887: \caption{

1888: SiC 1B and SiC 2A night spectra (binned by 12 pixels=0.156\,\AA) of the Lyman continuum region.  The background subtraction

1889: of the SiC 1B spectrum has been adjusted to match a flux level similar to that in the

1890: SiC 2A spectrum.  Neither spectra show convincing evidence of Lyman continuum

1891: emission.  In particular it should be noted that apparent features in one spectra

1892: are not replicated in the other as expected in noisy, low S/N spectra.

1893: \label{sic1b2afix}}

1894: \end{figure}

1895: 

1896: 

1897: 

1898: 

1899: 

1900: 

1901: \begin{figure}

1902: \centering

1903: \columnwidth=.30\columnwidth

1904: \includegraphics[width=3.in]{f14a_color.eps}

1905: \includegraphics[width=3.in]{f14b_color.eps}

1906: \vspace{.2in}

1907: \caption{

1908: These plots are originally described in \citet{grim05}.  The left Figure shows a relationship

1909: between $\rm{L_{FIR}\,(SFR)}$ and the thermal X-ray luminosity (0.3-2.0 keV).

1910: On the right we have divided both the thermal X-ray 

1911: and the FIR luminosity by the $K$-band luminosity (a proxy for stellar mass).  The FIR luminosity

1912: of the dwarf starbursts is actually the sum of the their UV and FIR luminosities.

1913: There is a clear

1914: linear relation between the SFR per stellar mass and

1915: the thermal X-ray emission per stellar mass.  The IRAS and $K$-band

1916: data were obtained from {\it NED} and have been tranformed using

1917: the methods of \citet{sand96} and \citet{carp01} respectively.  Both \vv\, and \hoo\,

1918: follow the scaling

1919: relation defined by starbursts.

1920: Symbols indicate dwarf starbursts (orange squares), starbursts (blue stars), \ulg\, (red triangles), AGN \ulg\, (green circles), and local LBG analogues (purple crosses).

1921: \label{div_L_K}}

1922: \end{figure}

1923: 

1924: 

1925: \begin{figure}

1926: \centering

1927: \leavevmode

1928: \columnwidth=.30\columnwidth

1929: \includegraphics[width=3in]{f15a_color.eps}

1930: \hfil

1931: \includegraphics[width=3in]{f15b_color.eps}

1932: \caption{

1933: 90\% flux enclosed radii  in the 0.3-1.0 keV X-ray band vs $K$-band 

1934: and FIR luminosity (see \citet{grim05} for details).

1935: The size of the X-ray emitting region is correlated with the 

1936:  stellar mass ($K$-band) and SFR (FIR).  The FIR luminosity

1937: of the dwarf starbursts is actually the sum of the their UV and FIR luminosities.

1938: The LBG analogues (\vv\,and\,\hoo)

1939: fall on the low end of FIR relation due to their high SFR and compact size.  

1940: The spatial extent of the hot X-ray gas appears to be more strongly correlated

1941: with the mass of the host galaxy.

1942: Symbols indicate dwarf starbursts (orange squares), 

1943: starbursts (blue stars), \ulg\, (red triangles), AGN \ulg\, (green circles), and 

1944: local LBG analogues (purple crosses).

1945: \label{r90}}

1946: \end{figure}

1947: 

1948: 

1949: \begin{figure}

1950: \centering

1951: \includegraphics[width=3in]{f16_color.eps}

1952: \vspace{.2in}

1953: \caption{

1954: $\rm{F_{60\mu m}/F_{100\mu m}}$ vs  X-ray gas temperature.

1955: The $\rm{F_{60\mu m}/F_{100\mu m}}$ ratio is an indicator of

1956: dust temperature and hence the SFR per unit area. 

1957: There is a rough correspondence between the 

1958: gas and dust temperatures.

1959: While \vv\, falls near the principal relation, \hoo\,

1960: is an outlier with F60$\micron$/F100$\micron$=1.29.

1961:  Symbols indicate dwarf starbursts (orange squares), starbursts (blue stars), \ulg\, (red triangles), AGN \ulg\, (green circles), and local LBG analogues (purple crosses).

1962: \label{F60o100vkt}}

1963: \end{figure}

1964: 

1965: 

1966: 

1967: 

1968: 

1969: 

1970: \end{document}

1971: 

1972: 

1973: