0801.2380/ms.tex
1: %%
2: %% Beginning of file 'sample.tex'
3: %%
4: %% Modified 2004 January 9
5: %%
6: %% This is a sample manuscript marked up using the
7: %% AASTeX v5.x LaTeX 2e macros.
8: 
9: %% The first piece of markup in an AASTeX v5.x document
10: %% is the \documentclass command. LaTeX will ignore
11: %% any data that comes before this command.
12: 
13: %% The command below calls the preprint style
14: %% which will produce a one-column, single-spaced document.
15: %% Examples of commands for other substyles follow. Use
16: %% whichever is most appropriate for your purposes.
17: %%
18: %\documentclass[12pt,preprint]{aastex}
19: 
20: %% manuscript produces a one-column, double-spaced document:
21: 
22: %% \documentclass[manuscript]{aastex}
23: 
24: %% preprint2 produces a double-column, single-spaced document:
25: 
26: %% \documentclass[preprint2,longabstract]{aastex}
27: 
28: \documentclass[apjl]{emulateapj}
29: %\documentclass[12pt,preprint]{aastex}
30: %\usepackage{apjfonts}
31: 
32: \newcommand{\cmjj}{\mbox{${\rm cm^{-2}}$}}
33: \newcommand{\hI}{\mbox{${\rm H\ I}$}}
34: \newcommand{\lya}{\mbox{${\rm Ly}\alpha$}}
35: \newcommand{\lyb}{\mbox{${\rm Ly}\beta$}}
36: \providecommand{\kms}{\,\ensuremath{\rm{km\,s}^{-1}}}
37: \providecommand{\cm}{\,\ensuremath{\mbox{cm}^{-2}}}
38: 
39: \providecommand{\dndz}{\ensuremath{\mbox{d${\cal N}$/d$z$}}}
40: \providecommand{\dndznum}{$10.4\pm 2.2$}
41: \providecommand{\nabs}{27}
42: \providecommand{\OmegaOVI}{\ensuremath{\Omega_{\scriptOVI}\,h}}
43: \providecommand{\OmegaOVInum}{\ensuremath{(1.7 \pm 0.3) \times10^{-7}}}
44: \providecommand{\Omegab}{\ensuremath{\Omega_b}}
45: \providecommand{\OVI}{\ensuremath{\mbox{\ion{O}{6}}}}
46: \providecommand{\HI}{\ensuremath{\mbox{\ion{H}{1}}}}
47: \providecommand{\scriptHI}{\ensuremath{\mbox{\scription{H}{1}}}}
48: \providecommand{\HII}{\ensuremath{\mbox{\ion{H}{2}}}}
49: \providecommand{\Halpha}{\ensuremath{\mbox{H}\alpha}}
50: \providecommand{\CIV}{\ensuremath{\mbox{\ion{C}{4}}}}
51: \providecommand{\OI}{\ensuremath{\mbox{\ion{O}{1}}}}
52: \providecommand{\NV}{\ensuremath{\mbox{\ion{N}{5}}}}
53: \providecommand{\OVII}{\ensuremath{\mbox{\ion{O}{7}}}}
54: \providecommand{\Lya}{\ensuremath{\mbox{Ly\,}\alpha}}
55: \providecommand{\SiII}{\ensuremath{\mbox{\ion{Si}{2}}}}
56: \providecommand{\SII}{\ensuremath{\mbox{\ion{S}{2}}}}
57: \providecommand{\OI}{\ensuremath{\mbox{\ion{O}{1}}}}
58: \providecommand{\CII}{\ensuremath{\mbox{\ion{C}{2}}}}
59: \providecommand{\degree}{\ensuremath{^{\circ}}}
60: \providecommand{\deg}{^{\circ}}
61: \providecommand{\Msun}{\,\ensuremath{\mbox{M}_{\odot}}}
62: \providecommand{\kms}{\,\ensuremath{\rm{km\,s}^{-1}}}
63: \providecommand{\cm}{\,\ensuremath{\mbox{cm}^{-2}}}
64: \providecommand{\mA}{\,\ensuremath{\mbox{m\AA}}}
65: \providecommand{\K}{\,\ensuremath{\mbox{K}}}
66: \providecommand{\JyBeam}{\,\ensuremath{\mbox{Jy\,Beam$^{-1}$}}}
67: \providecommand{\kpc}{\,\ensuremath{\mbox{kpc}}}
68: \providecommand{\Mpc}{\,\ensuremath{\mbox{Mpc}}}
69: \providecommand\arcmin{\mbox{$^\prime$}}%
70: \providecommand\arcsec{\mbox{$^{\prime\prime}$}}%
71: \newcommand{\llambda}{\ensuremath{\lambda \lambda}}
72: \providecommand{\tabOVI}{\ensuremath{\mbox{\tabion{O}{6}}}}
73: \providecommand{\tabHI}{\ensuremath{\mbox{\tabion{H}{1}}}}
74: \providecommand{\scriptOVI}{\ensuremath{\mbox{\scription{O}{6}}}}
75: \providecommand\scription[2]{\scriptsize#1$\;${\scriptsize\uppercase\expandafter{\romannumeral #2}}\relax}%
76: \providecommand{\etal}{\ensuremath{\mbox{et~al.}}}
77: 
78: 
79: 
80: 
81: \shorttitle{A STIS Survey for OVI Absorbers}
82: \shortauthors{Thom \& Chen}
83: 
84: \begin{document}
85: 
86: \slugcomment{Submitted to the Astrophysical Journal}
87: 
88: \title{A STIS Survey for O\,VI Absorption Systems at $0.12 < z\lesssim
89: 0.5$ I.: The Statistical Properties of Ionized Gas$^{1}$}
90: 
91: \author{Christopher Thom and Hsiao-Wen Chen}
92: \affil{Dept.\ of Astronomy \& Astrophysics, 5640 S.\ Ellis Ave, Chicago, IL, 60637, U.S.A. \\ {\tt cthom,hchen@oddjob.uchicago.edu}}
93: 
94: \altaffiltext{1}{Based in part on observations with the NASA/ESA Hubble Space
95: Telescope, obtained at the Space Telescope Science Institute, which is operated
96: by the Association of Universities for Research in Astronomy, Inc., under NASA
97: contract NAS5--26555.}
98: 
99: \begin{abstract}
100: 
101:   We have conducted a systematic survey for intervening O\,VI absorbers in available echelle spectra
102:   of 16 QSOs at $z_{\rm QSO}=0.17-0.57$.  These spectra were obtained using HST/STIS with the E140M
103:   grating.  Our search uncovered a total of \nabs\ foreground \OVI\ absorbers with rest-frame
104:   absorption equivalent width $W_r(1031)\gtrsim 25\,\mA$.  Ten of these QSOs exhibit strong O\,VI
105:   absorbers in their vicinity.  Our O\,VI survey does not require the known presence of \lya, and
106:   the echelle resolution allows us to identify the O\,VI absorption doublet based on their common
107:   line centroid and known flux ratio.  We estimate the total redshift survey path, $\Delta\,z$,
108:   using a series of Monte-Carlo simulations, and find that $\Delta\,z=1.66, 2.18$, and 2.42 for
109:   absorbers of strength $W_r=30, 50$ and $80$ m\AA, respectively, leading to a number density of
110:   $d\,{\cal N}(W\ge 50\,\mA)/dz = 6.7\pm 1.7 $ and $d\,{\cal N}(W\ge 30\,\mA)/dz = 10.4 \pm 2.2$.
111:   In contrast, we also measure $d\,{\cal N}/dz=27\pm 9$ for O\,VI absorbers of $W_r>50\ \mA$ at
112:   $|\Delta\,v|< 5000$ \kms\ from the background QSOs.  Using the random sample of O\,VI absorbers
113:   with well characterized survey completeness, we estimate a mean cosmological mass density of the
114:   O\,VI gas $\Omega({\rm O}^{5+})\,h = \OmegaOVInum$.  In addition, we show that $<5$\% of O\,VI
115:   absorbers originate in underdense regions that do not show a significant trace of H\,I.
116:   Furthermore, we show that the neutral gas column $N(\hI)$ associated with these O\,VI absorbers
117:   spans nearly five orders of magnitude, and shows moderate correlation with $N({\rm O\,VI})$.
118:   Finally, while the number density of O\,VI absorbers varies substantially from one sightline to
119:   another, it also appears to be inversely correlated with the number density of \hI\ absorbers
120:   along individual lines of sight.
121: 
122: \end{abstract}
123: 
124: \keywords{cosmology: observations---intergalactic medium---quasars: absorption lines}
125: 
126: \section{Introduction}
127: 
128: Uncovering the missing baryons in the present-day universe is one of the central issues in
129: observational cosmology, because an accurate accounting of the baryon budget at different locations
130: sets an important constraint in modeling galaxy formation and evolution.  Over the past few years,
131: measurements of the baryonic mass density from independent methods have approached a consistent
132: value with high precision, $\Omega_b\,h^2 = 0.020 \pm 0.002$ \citep{burles-etal-01, spergel-etal-03,
133:   wang-etal-03, omeara-etal-06}.  But while $> 95$\% of all baryons can be accounted for in the
134: \lya\ forest at redshift $z\sim 3$ \citep{rauch-etal-97}, only $1/3$ have been identified at
135: redshift $z=0$ in known components, such as stars and neutral gas in galaxies or hot plasma in
136: galaxy clusters \citep[e.g.][]{persic-salucci-92, fukugita-etal-98}.
137: 
138: Cosmological simulations indicate that $\approx 30-40$\% of all baryons at the present epoch reside
139: in a diffuse warm-hot intergalactic medium (WHIM), with temperatures in the range $10^5 < T < 10^7$ K,
140: rather than in virialized halos where galaxies or galaxy groups lie
141: \citep{cen-ostriker-99-missing-baryons,dave-etal-01-WHIM-baryons}.  The simulations also show that
142: this WHIM is shock-heated as tenuous gas accretes onto large-scale filamentary structures and
143: remains hot owing to the low gas density and inefficient cooling.  Locating the WHIM in order to
144: verify these models is difficult, because the temperature of the gas is too low to be easily
145: detected with the current generation of x-ray satellites and is too hot to be observed from the
146: ground using optical-IR facilities.  Simple ionization modeling of hot gas at $T\sim 3\times 10^5$ K
147: and metallicity 0.1 solar has shown that the gaseous clouds would imprint prominent absorption
148: features produced by ionic transitions such as O\,VI $\lambda\lambda$ 1031, 1037 and Ne\,VIII
149: $\lambda\lambda$ 770, 780 in the spectra of background QSOs \citep[e.g.][]{verner-etal-94,
150:   mulchaey-etal-96}.  Surveys for the WHIM based on the presence of these ions at $z<1$ therefore
151: require a high-resolution UV spectrograph in space.
152: 
153: \citet{tripp-etal-00-H1821+643} searched for intervening O\,VI absorbers along the sightline toward
154: H1821$+$643 ($z_{\rm QSO}=0.297$) in echelle spectra ($R\equiv\lambda/\Delta\lambda\approx 43,000$)
155: obtained using the Space Telescope Imaging Spectrograph (STIS) on board the Hubble Space Telescope
156: (HST).  The search indicated that intervening O\,VI absorption systems found in QSO spectra may
157: indeed contain a significant fraction of baryons that had been missed by surveys based on the direct
158: detection of photons emitted by known objects.  Subsequent searches along individual sightlines have
159: been carried out toward PG0953$+$415 at $z_{\rm QSO}=0.239$ \citep{savage-etal-02-PG0953+415},
160: PG1259$+$593 at $z_{\rm QSO}=0.478$ \citep{richter-etal-04-PG1259+593}; PG1116$+$215 at $z_{\rm
161:   QSO}=0.1765$ \citep{sembach-etal-04-PG1116+215}, PKS0405$-$123 at $z_{\rm QSO}=0.573$
162: \citep{prochaska-etal-04-PKS0405-12}, and HE0226$-$4110 at $z_{\rm QSO}=0.495$
163: \citep{lehner-etal-06-HE0226-4110}.  The measured line density of intervening O\,VI absorbers with
164: rest-frame absorption equivalent width of the 1031 transition $W_r \ge 50$ m\AA\ ranges from $d{\cal
165:   N}({\rm O\,VI})/dz=11$ to 36\footnote{\citet{tripp-etal-00-H1821+643}
166:   reported $d{\cal N}({\rm O\,VI})/dz\sim 48$ based on four O\,VI absorbers found with $W_r > 30$
167:   m\AA\ toward H1821$+$643.  Three of the four absorbers have $W_r > 50$ m\AA.  We therefore scale
168:   $d{\cal N}({\rm O\,VI})/dz$ accordingly and work out $d{\cal N}({\rm O\,VI})/dz=36$ for O\,VI
169:   absorbers stronger than $W_r=50$ m\AA.} per line of sight at $z_{\rm O\,VI}=[0,0.5]$.
170: 
171: The large scatter in the observed number density of O\,VI absorbers suggests a significant variation
172: in the spatial distribution of O$^{5+}$ ions across different sightlines.  It also underscores a
173: potential systematic bias in evaluating the cosmological mass density of warm-hot gas based on
174: selective quasar sightlines.  The first blind search of O\,VI absorbers was carried out at $z_{\rm
175:   O\,VI}=[0.5,2.1]$ toward 11 background quasars by \citet{burles-tytler-96-OVI-Omega_b} using
176: medium resolution ($R\approx 1300$) and $S/N\approx 20$ Faint Object Spectrograph spectra.  They
177: reported $d{\cal N}({\rm O\,VI})/dz=1.0\pm 0.6$ per line of sight at a mean redshift $\langle z_{\rm
178:   O\,VI}\rangle=0.9$ with $W_r \ge 210$ m\AA.  More recently,
179: \citet{danforth-etal-05-OVI-baryon-census} conducted a search of O\,VI absorption features in {\it
180:   known} \lya\ absorbers at $z_{\rm abs}<0.15$, using Far Ultraviolet Spectroscopic Explorer (FUSE)
181: spectra ($R\approx 15,000$) of 31 active galactic nuclei (AGN).  Their search yielded $d{\cal
182:   N}({\rm O\,VI})/dz=9\pm 2$ and $17\pm 3$ per line of sight with $W_r \ge 50$ and $W_r\ge 30$ m\AA,
183: respectively, at $z_{\rm O\,VI} < 0.15$, which is lower than measurements from individual lines of
184: sight discussed earlier.
185: 
186: Despite the large scatter in the number density of intervening O\,VI absorbers, these measurements
187: imply that at least $\sim 10\%$ of the total baryons reside in the gaseous clouds probed by the
188: O\,VI transition.  Assuming a mean metallicity of 1/10 solar and an ionization fraction of 20\% for
189: the O\,VI absorbers, Tripp \etal\ (2000) derived a lower limit to the cosmological mass density in
190: the ionized gas of $\Omega_b({\rm O\,VI})\gtrsim 0.003\,h^{-1}$ and Danforth \& Shull (2005)
191: derived $\Omega_b({\rm O\,VI})\approx 0.002\,h^{-1}$.  This general inference of O\,VI absorbers
192: being a substantial reservoir of missing baryons is supported by numerical simulations that result
193: in a mean metallicity of $1/10$ solar for the WHIM \citep{cen-etal-01, fang-bryan-01-H1821+643_IGM}.
194: However, additional uncertainties arise due to the unknown ionization state and metallicity of the
195: gas.
196: 
197: We have conducted a systematic survey of intervening O\,VI absorbers in available echelle spectra
198: obtained using HST/STIS with the E140M grating.  A search of the HST data archive yielded
199: high-resolution ($R\approx 45,000$) spectra of 16 distant quasars at $z_{\rm QSO}=0.17-0.57$ with a
200: mean $S/N > 7$ per resolution element.  These quasars have been selected by different authors for
201: different studies and have bright UV fluxes that are suitable for STIS echelle observations.
202: Therefore, they represent a random, flux-limited sample of quasar spectra for an unbiased search of
203: intervening absorbers.  The entire STIS echelle spectroscopic sample of quasars summarized in Table
204: 1 defines the random sightlines along which our blind search of intervening O\,VI absorbers was
205: conducted.
206: 
207: The primary objectives of the blind O\,VI survey are: (1) to obtain a statistically representative
208: sample of O\,VI absorbers along random lines of sight for an accurate measurement of the
209: cosmological mass density of warm-hot gas at low redshift; and (2) to study the incidence of O\,VI
210: absorbers along individual sightlines for quantifying the spatial variation of O\,VI absorbing gas.
211: These O\,VI absorbers identified along random sightlines also form a uniform sample for studying the
212: physical nature of tenuous gas probed by the presence of O$^{5+}$ and for investigating their
213: large-scale galactic environment.  Here we focus our discussion on the statistical properties of
214: O\,VI absorbers.  Studies of the gas properties and discussions of individual absorbers are given in
215: an accompanying paper (Thom \& Chen, 2008; hereafter paper~II), and studies of the surrounding
216: galaxies will be presented in future papers.
217: 
218: This paper is organized as follows.  In Section 2, we summarize the available quasar spectra and the
219: data quality.  In Section 3, we describe our approach for identifying an O\,VI $\lambda\lambda$
220: 1031, 1037 absorption doublet with no prior knowledge of the presence of H\,I or other ionic
221: transitions.  A Monte-Carlo analysis was performed for evaluating the completeness of the survey.
222: In Section 4, we present a statistical sample of O\,VI absorbers identified along the sightlines
223: toward the 16 background quasars.  We measure different statistical quantities of the ionized gas.
224: In Section 5, we compare our search results with others and discuss the implications in Section 6.
225: We adopt a $\Lambda$ cosmology, $\Omega_{\rm M}=0.3$ and $\Omega_\Lambda = 0.7$, with a
226: dimensionless Hubble constant $h = H_0/(100 \ {\rm km} \ {\rm s}^{-1}\ {\rm Mpc}^{-1})$ throughout
227: the paper.
228: 
229: 
230: \section{The STIS Echelle Spectra}
231: \label{sec: observations}
232: 
233: \begin{deluxetable*}{p{1.5in}rrrcr}
234: %\tabletypesize{\scriptsize}
235: \tablecaption{Summary of the STIS Echelle Spectra}
236: \tablewidth{0pt}
237: %\tablecolumns{6}
238: \tablehead{\colhead{QSO} & \colhead{$z_{\rm QSO}$} & \colhead{$z_{\rm min}$} & 
239: \colhead{$z_{\rm max}$\tablenotemark{a}} & \colhead{$t_{\rm exp}$} & \colhead{PID}  \\
240: \colhead{(1)} & \colhead{(2)} & \colhead{(3)} & \colhead{(4)} & \colhead{(5)} & \colhead{(6)}}
241: \startdata
242: HE\,0226-4110 \dotfill  &  0.4950 &    0.1154 &    0.4707 &   43772 &   9184  \\
243: PKS\,0312-77  \dotfill  &  0.2230 &    0.1241 &    0.1999 &   37908 &   8651  \\
244: PKS\,0405-12  \dotfill  &  0.5726 &    0.1241 &    0.5478 &   27208 &   7576  \\
245: HS\,0624+6907 \dotfill  &  0.3700 &    0.1222 &    0.3464 &   61950 &   9184  \\
246: PG\,0953+415  \dotfill  &  0.2390 &    0.1144 &    0.2163 &   24478 &   7747  \\
247: 3C\,249.1     \dotfill  &  0.3115 &    0.1222 &    0.2885 &   68776 &   9184  \\
248: PG\,1116+215  \dotfill  &  0.1765 &    0.1144 &    0.1536 &   39836 &   8165/8097  \\
249: PG\,1216+069  \dotfill  &  0.3313 &    0.1309 &    0.3078 &   69804 &   9184  \\
250: 3C\,273       \dotfill  &  0.1580 &    0.1144 &    0.1417 &   18671 &   8017  \\
251: PG\,1259+593  \dotfill  &  0.4778 &    0.1144 &    0.4533 &   95760 &   8695  \\
252: PKS\,1302-102 \dotfill  &  0.2784 &    0.1183 &    0.2558 &   22119 &   8306  \\
253: PG\,1444+407  \dotfill  &  0.2673 &    0.1222 &    0.2442 &   48624 &   9184  \\
254: 3C\,351.0     \dotfill  &  0.3719 &    0.1309 &    0.3483 &   73198 &   8015  \\
255: PHL\,1811     \dotfill  &  0.1917 &    0.1144 &    0.1690 &   33919 &   9418  \\
256: H\,1821+643   \dotfill  &  0.2970 &    0.1144 &    0.2741 &   50932 &   8165  \\
257: Ton\,28       \dotfill  &  0.3297 &    0.1231 &    0.3059 &   48401 &   9184  \\
258: \enddata
259: \tablenotetext{a}{The maximum redshift is defined for O\,VI absorbers
260: at velocity separation $> 5000$ \kms\ from the background QSO, but the
261: line search is conducted through the emission redshift of the QSO.}
262: \end{deluxetable*}
263: 
264: The goal of our project is to conduct a blind survey of O\,VI $\lambda\lambda$ 1031, 1037 absorption
265: doublet features along random lines of sight toward background quasars for establishing a
266: statistically representative sample of the O\,VI absorbers at $z<0.5$.  The survey therefore
267: requires high-quality UV spectra of distant quasars.  In addition, we focused our search on
268: available echelle quality spectra in order to uncover the majority of O\,VI absorbing gas.  Both
269: numerical simulations presented by \citet{chen-etal-03} and the recent FUSE survey by
270: \citet{danforth-etal-05-OVI-baryon-census} show that the bulk ($>60$\%) of O\,VI absorbers have
271: rest-frame absorption equivalent width of the 1031 transition $W_r=30-100$ m\AA.  An echelle
272: spectrum with resolution element $\delta\,v\approx 7$ \kms\ at moderate $S/N$ ($\approx 5$) per
273: resolution element is therefore necessary for detecting these weak lines.  This criterion excludes
274: Faint Object Spectrograph data (typically $\delta\,v\approx 230-270$ \kms) from our search.
275: Finally, the throughput of STIS declines steeply at $\lambda < 1180$ \AA, corresponding to a minimum
276: $z_{\rm O\,VI}\approx 0.14$, and we therefore include in our sample only those sightlines toward
277: quasars at $z_{\rm QSO}\gtrsim 0.16$ for the search of intervening O\,VI absorbers.
278: 
279: Our search in the HST data archive yielded 16 quasars observed with the STIS E140M grating.  A
280: summary of the STIS observations is presented in Table 1, where we list in columns (1)--(6), the QSO
281: field, $z_{\rm QSO}$, minimum and maximum redshifts over which the sample of random O\,VI absorbers
282: is established ($z_{\rm min}$, $z_{\rm max}$), total exposure time, and the program ID.  We note
283: that $z_{\rm max}$ is defined for O\,VI absorbers at velocity separation $> 5000$ \kms\ from the
284: background QSO, but the line search is conducted through the emission redshift of the QSO.
285: Individual flux-calibrated spectra were processed using the latest calibration files
286: \citep{aloisi-etal-05} and retrieved from the MAST server.  For each QSO, individual echelle orders
287: were continuum normalized and co-added to form a single, stacked spectrum using our own software.
288: The continuum was determined using a low-order polynomial fit to spectral regions that are free of
289: strong absorption features.  The combined echelle spectra of 16 QSOs have on average $S/N>7$ per
290: pixel over the majority of the spectral regions covered by the data.  Together the total redshift
291: survey path length covered by the echelle sample is $\Delta\,z\approx 2.73$ over $z=[0.12,0.5]$,
292: comparable to the total redshift path length at $z<0.15$, $\Delta\,z=2.2$, carried out by
293: \citet{danforth-etal-05-OVI-baryon-census} toward 31 sightlines in FUSE spectra.  Given the varying
294: $S/N$ versus wavelength along and between individual sightlines, we determine the $\Delta\,z$ versus
295: the O\,VI 1031 absorption equivalent width threshold $W_0$ in \S\ 3.2.
296: 
297: \section{A Statistical Sample of O\,VI Absorption Systems}
298: 
299: \subsection{The Search for O\,VI Doublet Features}
300: 
301: To establish a statistically representative sample of O\,VI absorbers, we carry out a blind survey
302: of O\,VI doublet features along each sightline that does not require known presence of hydrogen
303: \lya\ or other ionic transitions.  Before conducting the doublet search in each QSO spectrum, we
304: first exclude spectral regions of known absorption features from the Milky Way.  Next, we search in
305: the literature for known intervening absorbers, and further exclude spectral regions that contain
306: these known features from the subsequent doublet search. For those sightlines not well studied, we
307: performed this search ourselves.  For the doublet search, we first smooth each stacked echelle
308: spectrum using the Hanning function.  Next, we step through the smoothed spectrum and search for
309: deviant pixels that show negative fluctuation from unity by $> 1.5\ \sigma_c$, where $\sigma_c$ is
310: the 1-$\sigma$ error per smoothed pixel.  We then determine whether a deviant pixel is a candidate
311: absorption line based on a single gaussian-profile fit to the negative fluctuation.  We restrict the
312: width of the model profile to be no less than the spectral resolution element and no larger than 300
313: \kms.  The last criterion is justified based on the line width distribution of known O\,VI absorbers
314: \citep[]{heckman-etal-02, danforth-etal-05-OVI-baryon-census}.  The deviant pixel is discarded from
315: further consideration if the absorption equivalent width $W$ based on the best-fit gaussian profile
316: has $< 2\ \sigma$ significance.
317: 
318: Next, we treat each candidate absorption feature as the O\,VI 1031 line and determine the
319: corresponding redshift.  Then we examine whether the spectrum is consistent with the presence of
320: both members of the O\,VI doublet based on a simultaneous doublet profile fit at the expected
321: locations of the redshifted O\,VI 1031 and O\,VI 1037 lines.  At this stage, we restrict both
322: members to share the same line width and have a fiducial line ratio of $2:1$, given the ratio of
323: their respective oscillator strengths.  The product of the doublet fitting procedure is a list of
324: candidate O\,VI absorption doublets and their rest-frame absorption equivalent widths, $W_r(1031)$
325: and $W_r(1037)$.  We visually examine every candidate and accept doublets that satisfy both of the
326: following two criteria.  First, the O\,VI 1031 line has $> 3\,\sigma$ significance.  Second, the
327: line ratio $R_{\rm O\,VI}\equiv W_r(1031)/W_r(1037)$ falls between $1-\sigma_{R_{\rm O\,VI}}$ and
328: $2+ 2\,\sigma_{R_{\rm O\,VI}}$, where $\sigma_{R_{\rm O\,VI}}$ is the associated measurement
329: uncertainty in $R_{\rm O\,VI}$.  Because the O\,VI 1031 transition is required to have at least
330: 3-$\sigma$ significance, the presence of both members together warrants a higher significance for
331: the detection of each O\,VI doublet.
332: 
333: The entire sample of uncovered O\,VI absorbers include both foreground absorbers and absorbers that
334: are associated with the host of the background QSOs.  In the following discussion, we define a
335: ``random sample'' that consists of absorbers with $|\Delta\,v|> 5000$ \kms\ from the background QSOs
336: and an ``associated sample'' that are within $|\Delta\,v|< 5000$ \kms\ of the QSOs, where the
337: ionization background is expected to be significantly different from the meta-galactic radiation
338: field \citep[e.g.][]{weymann-etal-81}.  The results of our search are summarized in Table 2.  For
339: each QSO line of sight, we report in columns (2) through (4) the total number of random absorbers,
340: the minimum absorption equivalent width of the uncovered absorbers on that sightline, $W_r^{\rm
341:   min}$, and the presence/absence of absorbers in the vicinity of the background QSO.
342: 
343: \begin{figure*}
344: \epsscale{1.0}
345: \plotone{f1.eps}
346: \caption{Examples of \scriptOVI\ doublets.  For each doublet, the
347:   $\lambda 1031$ member is plotted in the top panel and the associated
348:   $\lambda 1037$ line at the bottom.  The corresponding 1-$\sigma$
349:   error array is shown in orange.  We also present a best-fit Voigt
350:   profile of each absorber in blue.  The spectra have been binned by a
351:   factor of 2 for display only.  In panels (a) and (b), we present new
352:   O\,VI absorbers uncovered at $z=0.2028$ and $z=0.3166$ in our search
353:   routine, along the sightlines toward PKS0312$-$77 and 3C351,
354:   respectively.  In panel (c), we present a previously known absorber
355:   at $z=0.24532$ toward H\,1821$+$643 \citep{tripp-etal-00-H1821+643}
356:   that has also been recovered in our blind search.  Finally, we show
357:   in panel (d) an absorber previously reported by these same authors
358:   at $z=0.1214$ toward H\,1821$+$643 based on additional FUSE spectra
359:   of the QSO.  It is clear that the quality of available STIS spectra
360:   does not allow us to recover this absorber.  Our random sample
361:   therefore does not include this system. }
362: \label{fig: absorber_plots}
363: 
364: \end{figure*}
365: 
366: 
367: Figure 1 presents three examples of the O\,VI absorption doublet uncovered in our blind search
368: (panels a--c), together with an example of a previously reported absorber that is not recovered in our
369: search (panel d).  While our search routine has uncovered \nabs\ absorbers toward 16 random lines of
370: sight, four of 16 systems previously reported toward seven lines of sight by other authors are not
371: confirmed in our survey.  We discuss these systems in more detail in \S\ 5 and paper~II.  As a
372: final step, we identify additional transitions associated with each fiducial O\,VI absorber.  We
373: perform a Voigt profile analysis to the \OVI\ doublets, as well as $\mbox{Ly}\,\alpha, \beta$ (where
374: available) using the VPFIT\footnote{http://www.ast.cam.ac.uk/~rfc/vpfit.html} software package. In
375: cases where absorbers have a complex structure, we attempt a multiple-component model profile.  Note
376: that we consider absorption components with velocity separation $\Delta\,v \le 200$ \kms\ as a
377: single absorber in the statistical sample.
378: 
379: \subsection{Completeness of the Search}
380: 
381: \begin{figure*}
382:   \epsscale{1.0} 
383:   \plotone{f2.eps}
384:   \caption{The 99.7\% completeness limit in the absorption equivalent
385:     width, $W_0$, of the O\,VI 1031 transition versus redshift, for
386:     all 16 sightlines in the STIS echelle sample.  At each redshift,
387:     the spread in $W_0$ represents varying spectral quality across
388:     different lines of sight.  For absorbers at $z_i$ with
389:     $W_r>W_0(z_i)$, they are expected to be detected at greater than
390:     3-$\sigma$ confidence level.  The plot shows that the STIS echelle
391:     sample as a whole is most sensitive to $W_r \sim 30$ m\AA\
392:     absorbers at $z\approx 0.2-0.5$. while at $z<0.2$ the majority of
393:     the spectra are sensitive to absorbers of $W_r>50$ m\AA.}
394:   \label{fig: ew}
395: \end{figure*}
396: 
397: To characterize the statistical properties of the O\,VI absorbers uncovered in our blind doublet
398: search, it is necessary to first estimate the completeness of the survey.  This is particularly
399: important for the STIS data because (1) the $S/N$ varies substantially across the entire wavelength
400: range in each echelle spectrum and (2) individual QSO spectra were obtained from different programs
401: and span a wide range in their mean noise levels.  We perform a suite of Monte-Carlo simulations to
402: quantify the completeness of our blind doublet search.  A specific goal of the Monte-Carlo
403: simulations is to estimate the total redshift path length $\Delta\,z$ covered by the survey for a
404: given absorption equivalent width threshold in the O\,VI 1031 transition, $W_*$, over which
405: absorbers of $W_r\ge W_*$ can be recovered.  Exploring the completeness versus $W_*$ allows us to
406: maximize the survey efficiency, given the already small spectroscopic quasar sample.
407: 
408: To carry out the simulations, we first mask regions along a single sightline $k$ that are blocked by
409: strong Galactic or extra-galactic absorption (e.g.\ Galactic O\,I\,1302, or Galactic and
410: extra-galactic \Lya).  Note that a strong absorption line at $\lambda_x$ will render us insensitive
411: to doublet features in {\it two} regions in redshift space ($\lambda_x/1031.927 - 1.0$ and
412: $\lambda_x/1037.616 - 1.0$), since it may mask either member of the \OVI\ doublet.  Next, we step
413: through the redshift space defined by $z_{\rm min}$ and $z_{\rm QSO}$ of the sightline defined in
414: Table 1.  At each redshift $z_i$, we adopt an intrinsic line width of $b=10$ \kms\ (corresponding to
415: a gas temperature of $T\sim 10^5$ K), generate a model absorption line of varying strength $W_{in}$,
416: redshift the model to $z_i$, and perturb the smooth model within the random noise characterized by
417: the 1-$\sigma$ error array of the original echelle spectrum.  Next, we apply the line search program
418: and evaluate whether the synthetic absorption feature can be recovered.  We repeat the procedure
419: 1,000 times for each input line strength and determine the 3-$\sigma$ detection threshold at
420: $W_0^k(z_i)$, where the recovery rate of the synthetic absorption features is 99.7\%.  Finally, we
421: repeat the procedure for all 16 QSO spectra in the echelle sample.
422: 
423: The result of the Monte-Carlo simulations is a sensitivity curve $W_0^k(z)$ for every QSO sightline
424: $k$ that represents the 99.7\% completeness of our blind O\,VI survey.  These curves together allow
425: us to calculate the total survey path length versus $W_0$ for the entire sample of O\,VI absorbers
426: from our search.  For absorbers of $W_r>W_0$, they are expected to be detected at greater than
427: 3-$\sigma$ confidence level.  We present in Figure 2 the collective $W_0(z)$ curve from all 16 QSO
428: sightlines together.  At each redshift, the spread in $W_0$ represents varying spectral quality in
429: the echelle spectra of different QSOs.  The plot shows that our blind search is sensitive to $W_r
430: \sim 30$ m\AA\ absorbers at $z\approx 0.2-0.5$, while at $z<0.2$ the majority of the spectra are
431: sensitive to absorbers of $W_r>50$ m\AA.
432: 
433: To calculate the total survey path for our blind search, we first define a redshift path density
434: $g(W_r,z)$ at $z_i$ as
435: \begin{equation}
436: g(W_r,z_i)=\sum_k\,H(z_i-z_{\rm min}^k)\,H(z_{\rm max}^k)\,H(W_r-W_0^k(z_i)),
437: \end{equation}
438: where $H$ is the Heaviside step function, $[z_{\rm min}^k, z_{\rm max}^k]$ are the minimum and
439: maximum redshifts observed for the $k$th QSO in Table 1, and the sum extends over all QSO sightlines
440: in the echelle sample.  The total redshift path length covered by the survey for a given equivalent
441: width threshold $W_r$ is therefore defined by
442: \begin{equation}
443: \Delta\,z\,(W_r)\equiv g(W_r)=\int g(W_r,z)\,dz.
444: \end{equation} 
445: We evaluate equations (1) and (2) based on a redshift interval of $\delta\,z=0.0001$.  The results
446: are shown in panel (a) of Figure 3 for three different equivalent width thresholds, $W_r=30, 50, 80$
447: m\AA.  The plot illustrates that, as stronger absorber are considered, the total survey path length
448: increases; and that weaker absorbers require higher signal-to-noise data.  Several strong dips are
449: immediately apparent at redshifts $z \approx 0.18, 0.21, 0.22, 0.26, 0.29 $, which correspond to
450: Galactic absorption from \Lya\,1215, \SII\,1250, \SII\,1259, \OI\,1302, \CII\,1334 respectively.
451: 
452: \begin{figure*}
453:   \epsscale{1.0} 
454:   \plotone{f3.eps}
455:   \caption{Summary of the results from our blind search of O\,VI
456:     absorbers.  In Panel (a), we present the total redshift path
457:     covered by the entire sample of 16 sightlines for different
458:     absorption equivalent width thresholds $W_*$.  Based on a series
459:     of Monte-Carlo simulations, we derive a total redshift path of
460:     $\Delta\,z=1.66, 2.18$, and 2.42 for $W_r=30, 50, 80$ m\AA,
461:     respectively.  In panel (b), we show for each individual QSO
462:     sightline the total redshift path allowed by the STIS echelle
463:     spectrum (black solid lines) and the redshift path that is
464:     sufficiently sensitive for detecting a $W_r=30$ m\AA\ absorber at
465:     $> 99.7$\% confidence level (blue lines).  In panel (c), we
466:     present the number density of $W_r\ge 30\,\mA$ O\,VI absorbers per
467:     sightline evaluated based on individual lines of sight, together
468:     with their associated errors.  The errors are estimated according
469:     to Equation (4).  We see a significant variation across the 16
470:     sightlines that cannot be easily accounted for by random errors.}
471:   \label{fig: gz}
472: \end{figure*}
473: 
474: In panel (b) of Figure 3, we show the redshift path along individual sightlines covered by our
475: survey (solid black lines).  Masked out regions due to Galactic absorption features (e.g.\ the \lya\
476: 1215 transition) or other ionic transitions of known intervening absorbers appear as gaps in the
477: path.  Blue segments indicate the spectral regions that are sensitive to $W_r=30$ m\AA.  The
478: tick-marks in panel (b) indicate the location of random O\,VI absorbers identified in our survey.
479: We summarize the completeness of our O\,VI search in columns (5) and (6) of Table 2, where we list
480: the total redshift path sensitive to $W_r=30$ and 50 m\AA\ for each line of sight.
481: 
482: \section{Statistical Properties of the O\,VI Absorbers}
483: \label{sec: results}
484: 
485: %\begin{deluxetable*}{llllllll}
486: \begin{deluxetable*}{p{1.3in}crccccr}
487: \tabletypesize{\scriptsize}
488: %\rotate
489: \tablecaption{\label{tab: target_summary}Summary of the Blind O\,VI Survey and the Survey Limits}
490: \tablewidth{0pt}
491: %\tablecolumns{8}
492: \tablehead{ & \colhead{} & \colhead{$W_r^{\rm min}$}  & \colhead{Associated} & 
493: \multicolumn{2}{c}{$\Delta\,z$ ($W\ge W_*$)} & & \colhead{\dndz} \\
494: \cline{5-6} 
495: \cline{8-8} 
496: \colhead{QSO} & \colhead{$N_{\mbox{abs}}^{\rm random}$} & \colhead{$(\mA)$} &  
497: \colhead{Absorbers} & \colhead{$(W_* = 50\,\mA)$} & 
498: \colhead{$(W_* = 30\,\mA)$} & & \colhead{$(W_* = 30\,\mA)$} \\
499: \colhead{(1)} & \colhead{(2)} & \colhead{(3)} & \colhead{(4)} & \colhead{(5)} & 
500: \colhead{(6)} & & \colhead{(7)} }
501: \startdata
502: HE\,0226$-$4110\tablenotemark{1,2,3} \dotfill & 4 &  43 & Y & 0.3055 & 0.2593 & & $13\pm 7$ \\
503: PKS\,0312$-$77                       \dotfill & 1 & 655 & N & 0.0375 & 0.0014 & & $18\pm 18$ \\
504: PKS\,0405$-$12\tablenotemark{4,5}    \dotfill & 4 &  30 & N & 0.3542 & 0.2780 & & $12\pm 6$ \\
505: HS\,0624$+$6907\tablenotemark{6}     \dotfill & 2 &  27 & Y & 0.1681 & 0.1458 & & $6\pm 6$ \\
506: PG\,0953$+$415\tablenotemark{7,8}    \dotfill & 1 & 121 & Y & 0.0698 & 0.0434 & & $11\pm 11$ \\
507: 3C\,249.1                            \dotfill & 1 &  72 & Y & 0.1206 & 0.0826 & & $8\pm 8$ \\
508: PG\,1116$+$215\tablenotemark{9,10}   \dotfill & 1 &  76 & Y & 0.0307 & 0.0053 & & $28\pm 28$ \\
509: PG\,1216$+$069                       \dotfill & 1 &  26 & N & 0.1339 & 0.1047 & & $0\pm 10$ \\
510: 3C\,273\tablenotemark{11}            \dotfill & 1 &  22 & N & 0.0271 & 0.0195 & & $0\pm 51$ \\
511: PG\,1259$+$593\tablenotemark{12}     \dotfill & 2 &  77 & N & 0.2851 & 0.2680 & & $7\pm 5$ \\
512: PKS\,1302$-$102                      \dotfill & 2 &  30 & N & 0.0992 & 0.0596 & & $26\pm 19$ \\
513: PG\,1444$+$407                       \dotfill & 0 & ... & Y & 0.0927 & 0.0478 & & $0\pm 21$ \\
514: 3C\,351.0\tablenotemark{13}          \dotfill & 1 & 232 & Y & 0.1575 & 0.1313 & & $6\pm 6$ \\
515: PHL\,1811\tablenotemark{14}          \dotfill & 1 &  64 & Y & 0.0388 & 0.0133 & & $23\pm 23$ \\
516: H\,1821$+$643\tablenotemark{15}      \dotfill & 4 &  30 & Y & 0.1226 & 0.0999 & & $34\pm 17$ \\
517: Ton\,28                              \dotfill & 1 &  26 & Y & 0.1396 & 0.0991 & & $0 \pm 10$\\
518: \hline 
519: total (16 sightlines)                         &27 &     & 10&  2.18  &  1.66  & & $10\pm 2$       \\
520: \enddata
521: \tablecomments{See also: (1) \citet{lehner-etal-06-HE0226-4110}; (2) \citet{savage-etal-05-HE0226-4110-NeVIII}; (3) \citet{ganguly-etal-06-HE0226-4110_qso_host_lines}; (4) \citet{chen-prochaska-00-PKS0405-12_z0.167}; (5) \citet{prochaska-etal-04-PKS0405-12}; (6) \citet{aracil-etal-06a-metals}; (7) \citet{tripp-savage-00-PG0953+415_z0.1423}; (8) \citet{savage-etal-02-PG0953+415}; (9) \citet{sembach-etal-04-PG1116+215}; (10) \citet{richter-etal-06-BLAs}; (11) \citet{heap-etal-02}; (12) \citet{richter-etal-04-PG1259+593}; (13) \citet{yuan-etal-02-3C351-qso-lines}; (14) \citet{jenkins-etal-03}; (15) \citet{tripp-etal-00-H1821+643}.}
522: %\end{deluxetable*}
523: \end{deluxetable*}
524: 
525: We detect a total of \nabs\ random \OVI\ doublets along 16 QSO lines of sight over a redshift range
526: from $z_{\rm O\,VI}=0.12$ to $z_{\rm O\,VI}=0.495$ with a mean redshift of $\langle z\rangle =
527: 0.25$.  Furthermore, we uncover associated absorbers toward nine QSOs at $|\Delta\,v|< 5000$ \kms\
528: from the background QSOs.  The random O\,VI absorbers form a statistically representative sample for
529: constraining their global mean properties.  The associated O\,VI absorbers form a uniform sample for
530: studying the gaseous halos of distant QSO hosts.  Here we consider the statistical properties of the
531: two samples separately.  Studies of the physical nature of individual absorbers will be presented in
532: a separate paper (Thom \& Chen 2008; Paper II).
533: 
534: \subsection{The Incidence of Random O\,VI Absorbers}
535: 
536: The Monte-Carlo simulations presented in \S\ 3.2 allow us to estimate the total redshift path length
537: as a function of absorption equivalent width, from which we can then derive an accurate estimate of
538: the number density of O\,VI absorbers per line of sight, using individual sightlines separately or
539: the entire QSO echelle sample collectively.  For each absorber of rest-frame O\,VI 1031 absorption
540: equivalent width $W_r^i>W_*$ identified in our search, the contribution to the total number density
541: per line of sight $d{\cal N}/dz$ is $1/g(W_r^i)$.  Over a statistical sample, we can therefore
542: derive for a given equivalent width limit $W_*$,
543: \begin{equation}
544: d\,{\cal N}(W \ge W_*)/dz = \sum_{i}\,\frac{1}{\Delta\,z (W_r^i)}\,\times\,H(W_r^i-W_*)
545: \end{equation}
546: where the sum extends over all systems found in a random sample and $H$ is the Heaviside step
547: function.  Note that $\Delta\,z\,(W_r^i)$ varies with $W_r^i$.  A stronger absorber may be detected
548: in lower $S/N$ spectra, allowing a larger survey path length for a given QSO echelle sample.
549: Equation (3) is therefore not the total number of absorbers, $N_{\rm abs}^{\rm random}$, divided by
550: a fixed path length, $(\Delta\,z)$.  The variance of $\dndz$ is evaluated according to
551: \begin{equation}
552: \sigma_{d{\cal N}/dz}^2 = \sum_{i}\,\frac{1}{[\Delta\,z (W_r^i)]^2}\,\times\,H(W_r^i-W_*).
553: \end{equation}
554: 
555: We first evaluate \dndz\ separately for individual lines of sight, adopting a common threshold for
556: all QSO spectra.  The results for $W_*=30$ \mA\ are summarized in column (7) of Table 2, and
557: presented in panel (c) of Fig.~\ref{fig: gz}.  Despite the relatively short path length per line of
558: sight, it is immediately apparent from Fig.~\ref{fig: gz} that there is a substantial variation in
559: \dndz\ from one line-of-sight to another.  In contrast to the line-of-sight toward H\,1821$+$643,
560: along which four O\,VI absorbers are found over $\Delta\,z\lesssim 0.1$ \citep{tripp-etal-00-H1821+643},
561: we note that the sightlines toward HS\,0624$+$6907, PG1259$+$593, and 3C\,351.0 allow longer survey
562: paths but exhibit fewer absorbers (between 1 and 2).  The large scatter in \dndz\ between random
563: sightlines underscores the necessity of a statistically significant sample of O\,VI absorbers for an
564: unbiased estimate of their incidence.
565: 
566: Combining all 16 QSO lines of sight together, we derive $d\,{\cal N}(W\ge 50\,\mA)/dz = 6 \pm 3$
567: over $\Delta\,z\,(W_*=50\,\mA)=2.18$, and $d\,{\cal N}(W\ge 30\,\mA)/dz = 9 \pm 4$ over
568: $\Delta\,z\,(W_*=30\,\mA)=1.66$.  In \S\ 5 and panel (b) of Figure 5, we  show that our number
569: density measurements can be explained well by recent numerical simulations that do not require super
570: starburst outflow, but account for non-equilibrium ionization conditions \citep{cen-fang-06}.
571: 
572: 
573: \subsection{The Cosmological Mass Density of Ionized Gas}
574: 
575: A particularly interesting quantity to determine, based on a statistically representative O\,VI
576: absorber sample, is the cosmological mean mass density contained in the O$^{5+}$ ions, $\Omega_{{\rm
577:     O}^{5+}}$.  In principle, we can then derive the cosmological mean mass density of warm baryons,
578: $\Omega_{\rm b}$, probed by the O\,VI transition according to
579: 
580: \begin{equation}
581: \Omega_{\rm b}=\frac{1}{f}\,\frac{\mu}{Z\times ({\rm O}/{\rm H})_\odot}\,\frac{m_{\rm H}}{m_{\rm O}}\,\Omega_{{\rm O}^{5+}},
582: \end{equation}
583: where $f\equiv N({\rm O}^{5+})/N({\rm O})$ is the fiducial ionization fraction of the O\,VI
584: absorbing gas, $Z\equiv ({\rm O}/{\rm H})/({\rm O}/{\rm H})_\odot$ represents its mean metallicity,
585: $({\rm O}/{\rm H})_\odot$ is the solar oxygen to hydrogen ratio by number, $\mu$ is the mean atomic
586: mass of the gas ($\mu=1.3$ for gas of solar composition), and $m_{\rm H}$ and $m_{\rm O}$ are the
587: atomic weight of hydrogen and oxygen, respectively.  In the following calculation, we adopt $({\rm
588:   O}/{\rm H})_\odot=4.57\times 10^{-4}$ (Asplund \etal\ 2004).
589: 
590: Following \citet{lanzetta-etal-01}, we calculate $\Omega_{{\rm O}^{5+}}$ according to
591: \begin{eqnarray}
592: \Omega_{{\rm O}^{5+}} &=& \frac{H_0\,m_{\rm O}}{c\,\rho_c} \int N({\rm O\,VI})\,f(N({\rm O\,VI}))\,d\,N({\rm O\,VI}) \\
593: &=& \frac{H_0\,m_{\rm O}}{c\,\rho_c} \sum_i\,\frac{N_i({\rm O\,VI})}{\Delta\,X_i}
594: \end{eqnarray}
595: where $\rho_c$ is the present-day critical density, $f(N({\rm O\,VI}))$ is the frequency distribution
596: function defined as the number of O\,VI absorbers per column density interval per unit absorption
597: path-length, $N_i$ is the corresponding O\,VI absorption column density for absorber $i$ with
598: rest-frame absorption equivalent width $W_r^i$ and $\Delta\,X_i$ is the co-moving absorption path
599: length covered by the entire echelle sample at the absorption strength $W_r^i$.  The co-moving
600: absorption path length $\Delta\,X_i$ for absorber $i$ is related to the total redshift path length
601: $\Delta\,z$ evaluated from Equation (2) at $W_r^i$ according to
602: \begin{equation}
603: \frac{d\,X}{d\,z} = \frac{(1+z)^2}{[\Omega_M\,(1+z)^3 - (\Omega_M +
604: \Omega_{\Lambda} - 1)(1+z)^2 + \Omega_{\Lambda}]^{1/2}}.
605: \end{equation}
606: The corresponding 1-$\sigma$ error is evaluated according to
607: \begin{equation}
608: \sigma_{\Omega_{{\rm O}^{5+}}} = \frac{H_0\,m_{\rm O}}{c\,\rho_c} \left[\sum_i\,\frac{\sigma_{N_i({\rm O\,VI})}^2}{(\Delta\,X_i)^2}\right]^{1/2}.
609: \end{equation}
610: For each uncovered absorber of $W_r^i$, we perform a Voigt profile fit to determine the
611: corresponding $N_i$.
612: 
613: We present the frequency distribution function determined based on the random O\,VI absorber sample
614: in panel (a) of Figure 5.  The bin size is selected to contain roughly the same number of absorbers
615: per bin.  Error bars represent 1-$\sigma$ Poisson fluctuations.  For the sample of 23 random O\,VI
616: absorbers with $W_r > 30\mA$, we therefore derive $\Omega({\rm O}^{5+})\,h = \OmegaOVInum$.
617: Assuming a $Z=1/10\,Z_\odot$ mean metallicity and $f=0.2$ ionization fraction for the O\,VI gas, we
618: further derive $\Omega_{\rm b}h\approx 0.0014$.  Bearing in mind the large uncertainties in the
619: metallicity and ionization fraction of the O$^{5+}$ bearing gas, our derived mean mass density of
620: ionized gas is less than 10\% of the total baryon mass density (e.g.\ Burles \etal\ 2001).
621: 
622: \subsection{Associated O\,VI Absorbers}
623: 
624: \begin{figure*}
625:   \epsscale{1.0} 
626:   \plotone{f4.ps}
627:   \caption{Absorption profiles of associated O\,VI absorbers
628:   identified in our STIS echelle sample.}
629:   \label{fig: qso}
630: \end{figure*}
631: 
632: We have excluded spectral regions within $|\Delta\,v|=0-5000$ \kms\ from the background QSOs for
633: establishing the statistical sample of random foreground absorbers in the previous discussion, due
634: to potential contamination features from the QSO host environment.  However, the blind O\,VI search
635: has been carried out all the way through the emission redshifts of these QSOs.  The O\,VI absorbers
636: found in the proximity of QSOs form a sample of associated absorbers that offers important
637: constraints on the halo gas in QSO hosts (Chelouche \etal\ 2007, in prep).
638: 
639: Ten of the 16 QSOs in our STIS echelle sample exhibit strong O\,VI absorbers $|\Delta\,v|\le 5000$
640: \kms\ blue-shifted from their emission redshifts (indicated in Table 2).  We present the absorption
641: profiles of these O\,VI absorbers in Figure 4.  The rest-frame absorption equivalent width of the
642: 1031 member ranges from $W_r=60\ \mA$ (the absorber at $\Delta\,v=-480$ \kms\ from Ton\,28) to
643: $W_r=3$ \AA\ (the multi-component absorber at $\Delta\,v\approx -1600$ \kms\ from 3C\,351.0).  The
644: multi-component absorber found in the vicinity of 3C\,351.0 has complex velocity profiles, extending
645: from $\Delta\,v\approx 0$ \kms\ to $\Delta\,v\approx -2200$ \kms.  It exhibits evidence of partial
646: covering of the O$^{5+}$ ions and is understood to originate in the outflow gas from the background
647: QSO \citep{yuan-etal-02-3C351-qso-lines}.
648: 
649: Excluding the complex sightline toward 3C351.0, the total redshift path length defined within
650: $|\Delta\,v|=0-5000$ \kms\ blue-ward of the emission redshifts of the remaining 15 QSOs in our
651: echelle sample is $\Delta\,z=0.33$.  At the equivalent width limit of our search $W_*=50\ \mA$, we
652: found nine O\,VI absorbers.  We measure $d\,{\cal N}/dz=27\pm 9$ for O\,VI absorbers of
653: $W_r(1031)>50\ \mA$ in the vicinity of background QSOs.  This is significantly higher than the
654: incidence of random absorbers in the foreground.
655: 
656: \section{Comparisons with Previous Surveys}
657: 
658: Our search in the STIS echelle QSO sample has uncovered a total of \nabs\ random \OVI\ absorbers of
659: $W_r\gtrsim 25\,\mA$ along 16 lines of sight, from which we derive the incidence of random O\,VI
660: absorbers, $d\,{\cal N}(W\ge 30\,\mA)/dz = $ \dndznum\ at $\langle z\rangle = 0.25$.  This is roughly
661: half of what is mentioned in Tripp \etal\ (2006), who reported a total of 44 O\,VI absorbers along
662: 16 lines of sight and $d\,{\cal N}(W\ge 30\,\mA)/dz = 23 \pm 4$ at $z<0.5$.  An independent survey
663: of O\,VI absorbers in known \lya\ absorbers by \citet{danforth-etal-05-OVI-baryon-census} yields an
664: absorber frequency of $d\,{\cal N}(W\ge 30\,\mA)/dz = 17 \pm 3$ or $d\,{\cal N}(W\ge 50\,\mA)/dz = 9
665: \pm 2$ at $z<0.15$.  We present these different measurements in Figure 5, along with recent
666: numerical simulation results that include galactic wind feedback and non-equilibrium ionization
667: conditions \citep{cen-fang-06}.  It is interesting to note that models with starburst feedback are
668: not required by our measurements.
669: 
670: The results of our blind search are consistent with those of
671: \citet{danforth-etal-05-OVI-baryon-census} from a FUSE survey at lower redshift that had been
672: carried out in known \lya\ absorbers.  The agreement is qualitatively consistent with the small
673: fraction of O\,VI absorbers that do not have an associated \lya\ transition.  The discrepancy
674: between our results and those of \citet{tripp-etal-06-conf} based on a similar set of QSO spectral
675: sample is most puzzling.  In the following, we discuss possible reasons for this discrepancy.
676: 
677: It is possible that the discrepancy is due to different QSO spectra covered by different surveys.
678: Our random sample has been established based exclusively on available STIS echelle data.  The
679: sightline along H\,1821$+$643 has the highest incidence of O\,VI absorbers of all 16 sightlines in
680: our echelle sample (See Table 2 and the discussion in \S\ 4.1).  In addition to the four O\,VI
681: absorbers originally reported in \citet{tripp-etal-00-H1821+643} and confirmed in our blind search,
682: \citet{tripp-etal-01-H1821+643_z0.1212} reported an additional O\,VI absorber at $z=0.1212$ in a
683: targeted search in higher quality FUSE spectra.  At $z=0.1212$, the O\,VI $\lambda\,1031$ line is
684: redshifted to $\lambda_{\rm obs}\approx 1157$ \AA, where the sensitivity of STIS echelle declines
685: steeply.  The lack of sensitivity at the blue end of the echelle sample is accounted for by the
686: completeness study of our random sample discussed in \S\ 3.2.  This $z=0.1212$ absorber found in a
687: targeted search after including additional FUSE spectra does not satisfy the criterion for
688: establishing a statistically unbiased sample, and has therefore been excluded in our random sample.
689: 
690: In addition, there are O\,VI absorbers reported in the literature but not confirmed by our search.
691: Along the same redshift path-length covered by our survey, 16 foreground O\,VI systems have been
692: reported in the literature by other authors toward seven QSOs (HE\,0226$-$4110, PKS\,0405$-$123,
693: PG\,0953$+$415, PG\,1116$+$215, PG1259$+$593, H\,1821$+$643, and 3C\,273; See Table 2 for
694: corresponding references), four are not confirmed in our survey.  An example is shown in panel (d)
695: of Figure 1, where we show the spectral region of an O\,VI doublet along the sightline toward
696: PG\,1259$+$593 reported by \citet{richter-etal-04-PG1259+593}.  The putative O\,VI 1031 member is
697: blended with the Si\,III 1206 transition of a \lya\ absorber at $z=0.0461$ (the feature at $v=-30$
698: \kms) and the C\,III transition of a \lya\ absorbers at $z=0.2924$ (the feature at $v=+100$ \kms).
699: Our search routine did not accept this absorber due to discrepant line strengths.  Our random sample
700: does not include this possible absorber.  Along the same sightline,
701: \citet{richter-etal-04-PG1259+593} also reported the presence of an O\,VI absorber at $z=0.3198$,
702: which is not confirmed by our routine due to discrepant absorption profiles and an inconsistent line
703: ratio.
704: 
705: Along the sightline toward HE\,0226$-$4110, \citet{lehner-etal-06-HE0226-4110} reported five O\,VI
706: absorbers with $W_r(1031)>50\,\mA$.  We cannot confirm the presence of their identification for a
707: $z=0.4266$ absorber.  In particular, the putative feature identified as the O\,VI 1037 member is not
708: present in our own reduction of the echelle spectra.  Inspections of individual exposures showed
709: that the broad feature are predominant in two of the 12 exposures, but the discrepancy may be due to
710: different treatment in the hot pixels in echelle spectra (T.\ Tripp, private communication).
711: Additional spectra obtained using the Cosmic Origin Spectrograph (COS) or STIS will be valuable for
712: confirming the presence of the doublet.  Consequently, this absorber is not included in our random
713: sample.
714: 
715: \section{Discussion}
716: \label{sec: discussion}
717: 
718: \subsection{The Ionization State of the O\,VI Gas}
719: 
720: 
721: \begin{figure}[t]
722:  \epsscale{1.0}
723:  \plotone{f5.eps}
724:  \caption{(a) The frequency distribution function of O\,VI absorbers
725:    from our random sample.  The bin size is chosen for each bin to
726:    contain roughly the same number of absorbers.  Error bars represents
727:    1-$\sigma$ Poisson fluctuations.  (b) Observed incidence of O\,VI
728:    absorbers from different studies. Star points are from this work;
729:    open circles are from \citet{danforth-etal-05-OVI-baryon-census} ;
730:    and open squares are from \citet{savage-etal-02-PG0953+415,
731:    prochaska-etal-04-PKS0405-12, tripp-etal-06-conf, cooksey-etal-08}.
732:    Note that we have slightly offset the data points in the horizontal
733:    direction for clarification.  The curves are recent WHIM
734:    predictions based on numerical simulations \citep{cen-fang-06} that
735:    account for non-equilibrium ionization conditions.  The top curve
736:    further includes Galactic wind feedback.}
737:  \label{fig: dndz}
738: \end{figure}
739: 
740: Various studies of individual O\,VI absorbers that consider both kinematic signatures of the
741: absorption profiles and abundance ratios between different ions have concluded that a large fraction
742: of these absorbers can be explained by a simple photo-ionization model
743: \citep[e.g.][]{tripp-etal-00-H1821+643, prochaska-etal-04-PKS0405-12, sembach-etal-04-PG1116+215,
744:   lehner-etal-06-HE0226-4110}.  These results are in contradiction to the expectation of a WHIM
745: origin, under which the bulk of the O$^{5+}$ ions originate in the shocks created as the gas
746: accretes onto the large filamentary structures.  However, it is often difficult to rule out a
747: collisional ionization scenario for these absorbers.  Additional uncertainties arise, if O\,VI
748: absorbers arise in post-shock gas that is undergoing radiative cooling and a simple ionization
749: equilibrium assumption does not apply (Gnat \& Sternberg 2007).
750: 
751: The comparison of the absorption column densities in Figure 6 between \lya\ and O\,VI transitions
752: for the O\,VI selected sample (solid points) demonstrates that the $N(\hI)$ associated with these
753: O\,VI absorbers spans nearly five orders of magnitude.  In addition, stronger O\,VI transitions have
754: on average stronger associated \lya\ transitions.  The null hypothesis in which \lya\ and O\,VI
755: column densities are randomly distributed among each other can be ruled out at $>97$\% confidence
756: level based on a generalized Kendall test that considers the non-detection of \lya\ for the O\,VI
757: absorber at $z=0.32639$ toward HE\,0226$-$4110.  We note that the STIS echelle sample is sensitive
758: to O\,VI absorbers of $W_r>80$ \mA\ over the entire redshift path-length allowed by the data, and
759: O\,VI absorbers of $W_r>50$ \mA\ over $\approx 80$\% of the sightlines.  The echelle spectra
760: therefore allow us to uncover O\,VI absorbers of $\log\,N({\rm O\,VI}) > 13.5$ over the majority of
761: the sightlines.  The correlation is therefore unlikely due to a selection bias.  Including
762: measurements from known O\,VI absorbers at low redshift (open squares and crosses), we find that the
763: correlation remains at $>80$\% significance level.
764: 
765: The broad range of $N(\hI)$ spanned by these O\,VI absorbers also makes it difficult to apply a
766: simple collisional ionization equilibrium model for explaining their origin.  Furthermore, the
767: correlation between $N({\rm O\,VI})$ and $N(\hI)$ indicates that the O\,VI absorbers appears to
768: track closely H\,I gas in overdense regions, while the opposite does not apply
769: \citep[e.g.][]{danforth-etal-05-OVI-baryon-census}.  This trend can be understood under a simple
770: photo-ionization scenario, and the slope in $N({\rm O\,VI})$ versus $N(\hI)$ implies a declining
771: ionization parameter (defined as the number of ionizing photons per gas particle) with increasing
772: gas column density (see Figure 31 of Paper II).
773: 
774: \subsection{O\,VI Absorbers without Associated H\,I}
775: 
776: A unique feature of our blind search is that the finding of O\,VI doublet does not rely on known
777: presence of \lya\ transitions.  The sample therefore allows us to constrain the fraction of O\,VI
778: absorbers that do not show significant traces of H\,I.  These absorbers may represent an extreme
779: case for outflow gas that has been ejected into underdense regions by strong galactic winds
780: \citep[e.g.][]{kawata-rauch-07}.
781: 
782: In Figure 6, we find that 26 of the 27 O\,VI absorbers in the random sample show associated H\,I
783: absorption features.  The only one possible \lya-free system is found to have $N({\rm
784:   O\,VI})=13.6\pm 0.1$ at $z_{\rm O\,VI}=0.3264$ toward HE\,0226$-$4110 (the arrow in Figure 6).
785: The comparison indicates that $<5$\% of O\,VI absorbers originate in underdense regions of
786: $N(\hI)\lesssim 10^{13}$ \cmjj\ that are heavily metal-enriched by starburst outflows
787: 
788: \subsection{Large-scale Variation in Absorber Number Density}
789: 
790: \begin{figure*}
791: \epsscale{0.6}
792: \plotone{f6.eps}
793: \caption{Comparison between $N({\rm O\,VI})$ and $N(\hI)$ for the 27
794:   O\,VI absorbers identified in our blind search (solid points.  All
795:   absorbers but one have an associated \lya\ absorption line.  We have
796:   also included published absorbers at low redshifts.  Open squares
797:   are $z<0.12$ absorbers presented in
798:   \citet{savage-etal-02-PG0953+415} for PG\,0953$+$415, in
799:   \citet{sembach-etal-04-PG1116+215} for PG\,1116$+$215, in
800:   \citet{richter-etal-04-PG1259+593} fo PG\,1259$+$593, in
801:   \citet{prochaska-etal-04-PKS0405-12} for PKS\,0405$-$123, and in
802:   \citet{cooksey-etal-08} for PKS\,1302$-$102.  Crosses indicate
803:   measurements from a FUSE survey by
804:   \citet{danforth-etal-05-OVI-baryon-census}.}
805:   \label{fig: ncomp}
806: \end{figure*}
807: 
808: Cosmological simulations also show that O\,VI absorbers originating in the WHIM reside in regions of
809: over-density $\delta\equiv \rho/\langle\rho\rangle=10-100$ at $z\sim 0$
810: \citep{dave-etal-01-WHIM-baryons, cen-etal-01, fang-bryan-01-H1821+643_IGM}.  This is similar to
811: what is found in simulation studies of \lya\ forest absorbers, according to which by $z=0$ the
812: majority of \lya\ absorbers are associated with $\delta = 10-100$.  It is therefore interesting to
813: examine the relation between \lya\ and O\,VI absorbers.
814: 
815: Of the 16 QSO sightlines in our echelle sample, seven have been studied extensively for the
816: line-of-sight properties of \lya\ absorbers \citep{sembach-etal-04-PG1116+215, williger-etal-06,
817:   lehner-etal-06-HE0226-4110}.  We present in Figure 7 a comparison of the observed number densities
818: between \lya\ absorbers and O\,VI absorbers measured along individual sightlines.  These include
819: H\,1821$+$643, HE\,0226$-$4110, HS\,0624$+$6907, PG\,0953$+$415, PG\,1116$+$215, PG\,1259$+$593, and
820: PKS\,0405$-$123.  The statistics of \lya\ absorbers are taken from the compilation of
821: \citet{lehner-etal-06-HE0226-4110}.  We have included the corrected \lya\ absorber catalog for the
822: PKS\,0405$-$123 sightline.  The error bars represent 1-$\sigma$ uncertainties.
823: 
824: Considering all seven sightlines together, we find a mean number density of $\langle\,d\,{\cal
825:   N}/dz\rangle=66$ for the \lya\ absorbers with an r.m.s.\ dispersion of 15.  In contrast, we find
826: $\langle\,d\,{\cal N}/dz\rangle=12$ for the O\,VI absorbers with an r.m.s.\ dispersion of 10.
827: Examining the number density variation from one sightline to another, we find that the line of sight
828: toward H\,1821$+$643 with the highest incidence of O\,VI absorbers has the lowest incidence of \lya\
829: absorbers.  Figure 7 shows a possible inverse correlation between the number density of \hI\
830: absorbers and the number density of O\,VI absorbers.  While \lya\ absorbers appear to be relatively
831: more abundant than O\,VI along each sightline, the number density of O\,VI absorbers do not scale
832: proportionally with the number density of \lya\ absorbers along individual lines of sight.
833: 
834: This anti-correlation between $[d{\cal N}/dz]_{\rm O\,VI}$ and $[d{\cal N}/dz]_{\rm \lya}$ may be
835: understood, if those sightlines with a higher incidence of O\,VI absorbers probe a larger fraction
836: of more highly ionized regions.  However, we note the result from \S\ 6.1, where we find nearly
837: every one of the O\,VI absorbers in our statistical sample shows an associated \lya\ transition.  On
838: the other hand, if O\,VI absorbers originate preferentially in underdense regions, then a higher
839: incidence of O\,VI implies a substantial fraction of the sightline probe underdense regions and
840: hence a lower incidence of \lya\ absorbers.  This is expected if heavy elements from starburst
841: outflows are ejected primarily into underdense regions \citep[e.g.][]{kawata-rauch-07}.  Further
842: insights can be obtained from a detailed examination of the large-scale galaxy environment of the
843: O\,VI absorbers.  A larger sample of QSO echelle spectra will also be valuable for a more
844: quantitative study to confirm this result.
845: 
846: \begin{figure*}
847: \epsscale{0.6}
848: \plotone{f7.ps}
849: \caption{Comparison of the line density variation from one line of
850:   sight to another for hydrogen \lya\ absorbers and O\,VI absorbers.
851:   The number density of \lya\ absorbers measured for individual
852:   sightlines are taken from Lehner \etal\ (2007) for \lya\ absorbers
853:   of rest-frame equivalent width $W_r(\lya) > 90 \mA$ and the number
854:   density of O\,VI absorbers measured for individual sightlines are
855:   from the blind search for O\,VI absorbers of rest-frame equivalent
856:   width $W_r({\rm O\,VI}) > 30 \mA$ presented in this paper.  We find a
857:   possible inverse correlation between the number density of \hI\
858:   absorbers and the number density of O\,VI absorbers.}
859:   \label{fig: lden}
860: \end{figure*}
861: 
862: \section{Summary}
863: 
864: We have conducted a systematic survey of intervening O\,VI absorbers in available echelle spectra
865: obtained using HST/STIS with the E140M grating.  We have collected echelle spectra of 16 distant
866: QSOs from the HST data archive.  These UV bright quasars at $z_{\rm QSO}=0.17-0.57$ have been
867: selected for HST/STIS observations by different authors for different studies, and represent a
868: random, flux-limited sample of QSO echelle spectra for an unbiased search of intervening absorbers.
869: The primary objectives of our blind O\,VI survey are: (1) to obtain a statistically representative
870: sample of O\,VI absorbers along random lines of sight for an accurate measurement of the
871: cosmological mass density of warm-hot gas at low redshift; and (2) to study the incidence of O\,VI
872: absorbers along individual sightlines for quantifying the spatial variation of O\,VI absorbing gas.
873: 
874: Our search in the STIS echelle QSO sample has uncovered a total of \nabs\ random ($|\Delta\,v|>5000$
875: \kms\ from the background QSO) \OVI\ absorbers with rest-frame absorption equivalent width of the
876: 1031 member $W_r\gtrsim 25\,\mA$ along 16 lines of sight.  Ten of these QSOs also exhibit strong
877: O\,VI absorbers at $|\Delta\,v|<5000$ \kms\ blue-shifted from the QSO emission redshifts.  Our O\,VI
878: survey differs from previous studies in two fundamental ways.  First, we do not require prior
879: knowledge of the presence of other transitions such as \lya.  The echelle resolution allows us to
880: identify the O\,VI absorption doublet based on their common line centroid and known flux ratio.
881: Second, the survey completeness is well understood and characterized based on a series of
882: Monte-Carlo simulations.  We obtain an equivalent width threshold curve for every QSO sightline.
883: Together, these curves allow us to estimate the total redshift survey path as a function of
884: equivalent width threshold, thereby maximizing the survey efficiency at the limit allowed by the
885: echelle spectra.
886: 
887: Based on the results of the Monte-Carlo simulations, we derive a total redshift path for random
888: absorbers of $\Delta\,z=1.66, 2.18$, and 2.42 for $W_r=30, 50$ and $80$ m\AA, respectively.  The total
889: path-length surveyed in our sample is comparable to the FUSE survey carried out at $z<0.15$ by
890: Danforth \& Shull (2005), and is the largest at $z=0.12-0.5$.  The estimated total redshift path
891: leads to a number density of $d\,{\cal N}(W\ge 50\,\mA)/dz = 6.7 \pm 1.7$ and $d\,{\cal N}(W\ge
892: 30\,\mA)/dz = $ \dndznum.  Recent numerical simulations that allow non-equilibrium ionization
893: conditions explain the observations very well.  In contrast, we also measure $d\,{\cal N}/dz=27\pm
894: 9$ for O\,VI absorbers of $W_r(1031)>50\ \mA$ in the vicinity of background QSOs.  For the sample of
895: 27 random O\,VI absorbers, we further derive a mean cosmological mass density of the O\,VI ions
896: $\Omega({\rm O}^{5+})\,h = \OmegaOVInum$ that leads to a mean cosmological mass density of warm
897: baryons $\Omega_{\rm b}h\approx 0.0014$.
898: 
899: Using the statistical sample of O\,VI absorbers, we show that $<5$\% of O\,VI absorbers originate in
900: underdense regions that do not show significant trace of H\,I.  In addition, a comparison of the
901: absorption column densities between \lya\ and O\,VI transitions for the O\,VI selected sample shows
902: that the $N(\hI)$ associated with these O\,VI absorbers spans nearly five orders of magnitude,
903: making it difficult to apply a simple collisional ionization equilibrium model for explaining their
904: origin.  Furthermore, stronger O\,VI transitions have on average stronger associated \lya\
905: transitions, indicating that the O\,VI absorbers appear to track closely H\,I gas in overdense
906: regions, while the opposite does not apply.  At the same time, comparisons of individual
907: line-of-sight properties show a moderate inverse correlation between the number density of \hI\
908: absorbers and the number density of O\,VI absorbers.  While nearly all O\,VI absorbers have an
909: associated \lya\ transitions to $N(\hI)\gtrsim 10^{13}$ \cmjj, the presence of abundant O\,VI
910: absorbers implies that a larger fraction of the gas along the line of sight becomes more ionized and
911: therefore reducing the line density of \lya\ absorbers.  A larger sample of QSO echelle spectra is
912: necessary for a more detailed study.
913: 
914: 
915: 
916: 
917: \acknowledgments
918: 
919: We thank N.\ Gnedin, J.~X.~Prochaska, and J.\ Tinker for valuable discussions.  This research has
920: made use of the NASA/IPAC Extragalactic Database (NED) which is operated by the Jet Propulsion
921: Laboratory, California Institute of Technology, under contract with the National Aeronautics and
922: Space Administration.  C.~T. and H.-W.C.\ acknowledges support from NASA grant NNG06GC36G.
923: H.-W.C. acknowledges partial support from an NSF grant AST-0607510.
924: 
925: 
926: %\bibliography{ovi-statistics}
927: %\bibliographystyle{apj}
928: \begin{thebibliography}{41}
929: \expandafter\ifx\csname natexlab\endcsname\relax\def\natexlab#1{#1}\fi
930: 
931: \bibitem[{{Aloisi} et~al.(2005){Aloisi}, {van der Marel}, {Mack}, {Leitherer},
932:   {Sirianni}, \& {Tosi}}]{aloisi-etal-05}
933: {Aloisi}, A., {van der Marel}, R.~P., {Mack}, J., {Leitherer}, C., {Sirianni},
934:   M., \& {Tosi}, M. 2005, \apjl, 631, L45
935: 
936: \bibitem[{{Aracil} et~al.(2006){Aracil}, {Tripp}, {Bowen}, {Prochaska}, {Chen},
937:   \& {Frye}}]{aracil-etal-06a-metals}
938: {Aracil}, B., {Tripp}, T.~M., {Bowen}, D.~V., {Prochaska}, J.~X., {Chen},
939:   H.-W., \& {Frye}, B.~L. 2006, \mnras, 367, 139
940: 
941: \bibitem[{{Asplund} et~al. (2004){Asplund}, {Grevesse}, {Sauval}, {Allende-Prieto}, {KiselmanAsplund}}]{asplund-etal-04} {Asplund}, M., {Grevesse}, N., {Sauval}, A. J., {Allende-Prieto}, C., \& {Kiselman}, D. 2004, A\&A, 417, 751
942: 
943: \bibitem[{{Burles} et~al.(2001){Burles}, {Nollett}, \&
944:   {Turner}}]{burles-etal-01}
945: {Burles}, S., {Nollett}, K.~M., \& {Turner}, M.~S. 2001, \apjl, 552, L1
946: 
947: \bibitem[{{Burles} \& {Tytler}(1996)}]{burles-tytler-96-OVI-Omega_b}
948: {Burles}, S. \& {Tytler}, D. 1996, \apj, 460, 584
949: 
950: \bibitem[{{Cen} \& {Fang}(2006)}]{cen-fang-06}
951: {Cen}, R. \& {Fang}, T. 2006, \apj, 650, 573
952: 
953: \bibitem[{{Cen} \& {Ostriker}(1999)}]{cen-ostriker-99-missing-baryons}
954: {Cen}, R. \& {Ostriker}, J.~P. 1999, \apj, 514, 1
955: 
956: \bibitem[{{Cen} et~al.(2001){Cen}, {Tripp}, {Ostriker}, \&
957:   {Jenkins}}]{cen-etal-01}
958: {Cen}, R., {Tripp}, T.~M., {Ostriker}, J.~P., \& {Jenkins}, E.~B. 2001, \apjl,
959:   559, L5
960: 
961: \bibitem[{{Chen} \& {Prochaska}(2000)}]{chen-prochaska-00-PKS0405-12_z0.167}
962: {Chen}, H.-W. \& {Prochaska}, J.~X. 2000, \apjl, 543, L9
963: 
964: \bibitem[{{Chen} et~al.(2003){Chen}, {Weinberg}, {Katz}, \&
965:   {Dav{\'e}}}]{chen-etal-03}
966: {Chen}, X., {Weinberg}, D.~H., {Katz}, N., \& {Dav{\'e}}, R. 2003, \apj, 594,
967:   42
968: 
969: \bibitem[{{Cooksey} et~al.(2008){Cooksey}, {Prochaska}, {Chen}, {Mulchaey}, \&
970:   {Weiner}}]{cooksey-etal-08}
971: {Cooksey}, K.~L., {Prochaska}, J.~X., {Chen}, H.-W., {Mulchaey}, J.~S., \&
972:   {Weiner}, B.~J. 2008, \apj, 676, 262
973: 
974: \bibitem[{{Danforth} \& {Shull}(2005)}]{danforth-etal-05-OVI-baryon-census}
975: {Danforth}, C.~W. \& {Shull}, J.~M. 2005, \apj, 624, 555
976: 
977: \bibitem[{{Dav{\'e}} et~al.(2001){Dav{\'e}},
978:   et~al.}]{dave-etal-01-WHIM-baryons}
979: {Dav{\'e}}, R., et~al. 2001, \apj, 552, 473
980: 
981: \bibitem[{{Fang} \& {Bryan}(2001)}]{fang-bryan-01-H1821+643_IGM}
982: {Fang}, T. \& {Bryan}, G.~L. 2001, \apjl, 561, L31
983: 
984: \bibitem[{{Fukugita} et~al.(1998){Fukugita}, {Hogan}, \&
985:   {Peebles}}]{fukugita-etal-98}
986: {Fukugita}, M., {Hogan}, C.~J., \& {Peebles}, P.~J.~E. 1998, \apj, 503, 518
987: 
988: \bibitem[{{Ganguly} et~al.(2006){Ganguly}, {Sembach}, {Tripp}, {Savage}, \&
989:   {Wakker}}]{ganguly-etal-06-HE0226-4110_qso_host_lines}
990: {Ganguly}, R., {Sembach}, K.~R., {Tripp}, T.~M., {Savage}, B.~D., \& {Wakker},
991:   B.~P. 2006, \apj, 645, 868
992: 
993: \bibitem[{{Heap} et~al.(2002){Heap}, {Williger}, {Dav{\'e}}, {Weymann},
994:   {Jenkins}, \& {Tripp}}]{heap-etal-02}
995: {Heap}, S.~R., {Williger}, G.~M., {Dav{\'e}}, R., {Weymann}, R.~J., {Jenkins},
996:   E.~B., \& {Tripp}, T.~M., \emph{Extragalactic Gas at Low Redshift}, edited by
997:   {Mulchaey}, J.~S. \& {Stocke}, J. 2002, vol. 254 of \emph{Astronomical
998:   Society of the Pacific Conference Series},  63--+
999: 
1000: \bibitem[{{Heckman} et~al.(2002){Heckman}, {Norman}, {Strickland}, \&
1001:   {Sembach}}]{heckman-etal-02}
1002: {Heckman}, T.~M., {Norman}, C.~A., {Strickland}, D.~K., \& {Sembach}, K.~R.
1003:   2002, \apj, 577, 691
1004: 
1005: \bibitem[{{Jenkins} et~al.(2003){Jenkins}, {Bowen}, {Tripp}, {Sembach},
1006:   {Leighly}, {Halpern}, \& {Lauroesch}}]{jenkins-etal-03}
1007: {Jenkins}, E.~B., {Bowen}, D.~V., {Tripp}, T.~M., {Sembach}, K.~R., {Leighly},
1008:   K.~M., {Halpern}, J.~P., \& {Lauroesch}, J.~T. 2003, \aj, 125, 2824
1009: 
1010: \bibitem[{{Kawata} \& {Rauch}(2007)}]{kawata-rauch-07}
1011: {Kawata}, D. \& {Rauch}, M. 2007, ArXiv e-prints, 704
1012: 
1013: \bibitem[{{Lanzetta} et~al.(1991){Lanzetta}, {McMahon}, {Wolfe}, {Turnshek},
1014:   {Hazard}, \& {Lu}}]{lanzetta-etal-01}
1015: {Lanzetta}, K.~M., {McMahon}, R.~G., {Wolfe}, A.~M., {Turnshek}, D.~A.,
1016:   {Hazard}, C., \& {Lu}, L. 1991, \apjs, 77, 1
1017: 
1018: \bibitem[{{Lehner} et~al.(2006){Lehner}, {Savage}, {Wakker}, {Sembach}, \&
1019:   {Tripp}}]{lehner-etal-06-HE0226-4110}
1020: {Lehner}, N., {Savage}, B.~D., {Wakker}, B.~P., {Sembach}, K.~R., \& {Tripp},
1021:   T.~M. 2006, \apjs, 164, 1
1022: 
1023: \bibitem[{{Mulchaey} et~al.(1996){Mulchaey}, {Mushotzky}, {Burstein}, \&
1024:   {Davis}}]{mulchaey-etal-96}
1025: {Mulchaey}, J.~S., {Mushotzky}, R.~F., {Burstein}, D., \& {Davis}, D.~S. 1996,
1026:   \apjl, 456, L5+
1027: 
1028: \bibitem[{{O'Meara} et~al.(2006){O'Meara}, {Burles}, {Prochaska}, {Prochter},
1029:   {Bernstein}, \& {Burgess}}]{omeara-etal-06}
1030: {O'Meara}, J.~M., {Burles}, S., {Prochaska}, J.~X., {Prochter}, G.~E.,
1031:   {Bernstein}, R.~A., \& {Burgess}, K.~M. 2006, \apjl, 649, L61
1032: 
1033: \bibitem[{{Persic} \& {Salucci}(1992)}]{persic-salucci-92}
1034: {Persic}, M. \& {Salucci}, P. 1992, \mnras, 258, 14P
1035: 
1036: \bibitem[{{Prochaska} et~al.(2004){Prochaska}, {Chen}, {Howk}, {Weiner}, \&
1037:   {Mulchaey}}]{prochaska-etal-04-PKS0405-12}
1038: {Prochaska}, J.~X., {Chen}, H.-W., {Howk}, J.~C., {Weiner}, B.~J., \&
1039:   {Mulchaey}, J. 2004, \apj, 617, 718
1040: 
1041: \bibitem[{{Rauch} et~al.(1997){Rauch}, et~al.}]{rauch-etal-97}
1042: {Rauch}, M., et~al. 1997, \apj, 489, 7
1043: 
1044: \bibitem[{{Richter} et~al.(2006){Richter}, {Savage}, {Sembach}, \&
1045:   {Tripp}}]{richter-etal-06-BLAs}
1046: {Richter}, P., {Savage}, B.~D., {Sembach}, K.~R., \& {Tripp}, T.~M. 2006, \aap,
1047:   445, 827
1048: 
1049: \bibitem[{{Richter} et~al.(2004){Richter}, {Savage}, {Tripp}, \&
1050:   {Sembach}}]{richter-etal-04-PG1259+593}
1051: {Richter}, P., {Savage}, B.~D., {Tripp}, T.~M., \& {Sembach}, K.~R. 2004,
1052:   \apjs, 153, 165
1053: 
1054: \bibitem[{{Savage} et~al.(2005){Savage}, {Lehner}, {Wakker}, {Sembach}, \&
1055:   {Tripp}}]{savage-etal-05-HE0226-4110-NeVIII}
1056: {Savage}, B.~D., {Lehner}, N., {Wakker}, B.~P., {Sembach}, K.~R., \& {Tripp},
1057:   T.~M. 2005, \apj, 626, 776
1058: 
1059: \bibitem[{{Savage} et~al.(2002){Savage}, {Sembach}, {Tripp}, \&
1060:   {Richter}}]{savage-etal-02-PG0953+415}
1061: {Savage}, B.~D., {Sembach}, K.~R., {Tripp}, T.~M., \& {Richter}, P. 2002, \apj,
1062:   564, 631
1063: 
1064: \bibitem[{{Sembach} et~al.(2004){Sembach}, {Tripp}, {Savage}, \&
1065:   {Richter}}]{sembach-etal-04-PG1116+215}
1066: {Sembach}, K.~R., {Tripp}, T.~M., {Savage}, B.~D., \& {Richter}, P. 2004,
1067:   \apjs, 155, 351
1068: 
1069: \bibitem[{{Spergel} et~al.(2003){Spergel}, et~al.}]{spergel-etal-03}
1070: {Spergel}, D.~N., et~al. 2003, \apjs, 148, 175
1071: 
1072: \bibitem[{{Thom} \& {Chen}(2008)}]{thom-chen-08-OVI-II}
1073: {Thom}, C. \& {Chen}, H.-W. 2008, \apjs, submitted (paper~II)
1074: 
1075: \bibitem[{{Tripp} et~al.(2006){Tripp}, {Bowen}, {Sembach}, {Jenkins}, {Savage},
1076:   \& {Richter}}]{tripp-etal-06-conf}
1077: {Tripp}, T.~M., {Bowen}, D.~V., {Sembach}, K.~R., {Jenkins}, E.~B., {Savage},
1078:   B.~D., \& {Richter}, P., \emph{Astrophysics in the Far Ultraviolet: Five
1079:   Years of Discovery with FUSE}, edited by {Sonneborn}, G., {Moos}, H.~W., \&
1080:   {Andersson}, B.-G. 2006, vol. 348 of \emph{Astronomical Society of the
1081:   Pacific Conference Series},  341--+
1082: 
1083: \bibitem[{{Tripp} et~al.(2001){Tripp}, {Giroux}, {Stocke}, {Tumlinson}, \&
1084:   {Oegerle}}]{tripp-etal-01-H1821+643_z0.1212}
1085: {Tripp}, T.~M., {Giroux}, M.~L., {Stocke}, J.~T., {Tumlinson}, J., \&
1086:   {Oegerle}, W.~R. 2001, \apj, 563, 724
1087: 
1088: \bibitem[{{Tripp} \& {Savage}(2000)}]{tripp-savage-00-PG0953+415_z0.1423}
1089: {Tripp}, T.~M. \& {Savage}, B.~D. 2000, \apj, 542, 42
1090: 
1091: \bibitem[{{Tripp} et~al.(2000){Tripp}, {Savage}, \&
1092:   {Jenkins}}]{tripp-etal-00-H1821+643}
1093: {Tripp}, T.~M., {Savage}, B.~D., \& {Jenkins}, E.~B. 2000, \apjl, 534, L1
1094: 
1095: \bibitem[{{Verner} et~al.(1994){Verner}, {Tytler}, \&
1096:   {Barthel}}]{verner-etal-94}
1097: {Verner}, D.~A., {Tytler}, D., \& {Barthel}, P.~D. 1994, \apj, 430, 186
1098: 
1099: \bibitem[{{Wang} et~al.(2003){Wang}, {Tegmark}, {Jain}, \&
1100:   {Zaldarriaga}}]{wang-etal-03}
1101: {Wang}, X., {Tegmark}, M., {Jain}, B., \& {Zaldarriaga}, M. 2003, \prd, 68,
1102:   123001
1103: 
1104: \bibitem[{{Weymann} et~al.(1981){Weymann}, {Carswell}, \&
1105:   {Smith}}]{weymann-etal-81}
1106: {Weymann}, R.~J., {Carswell}, R.~F., \& {Smith}, M.~G. 1981, \araa, 19, 41
1107: 
1108: \bibitem[{{Williger} et~al.(2006){Williger}, {Heap}, {Weymann}, {Dav{\'e}},
1109:   {Ellingson}, {Carswell}, {Tripp}, \& {Jenkins}}]{williger-etal-06}
1110: {Williger}, G.~M., {Heap}, S.~R., {Weymann}, R.~J., {Dav{\'e}}, R.,
1111:   {Ellingson}, E., {Carswell}, R.~F., {Tripp}, T.~M., \& {Jenkins}, E.~B. 2006,
1112:   \apj, 636, 631
1113: 
1114: \bibitem[{{Yuan} et~al.(2002){Yuan}, {Green}, {Brotherton}, {Tripp}, {Kaiser},
1115:   \& {Kriss}}]{yuan-etal-02-3C351-qso-lines}
1116: {Yuan}, Q., {Green}, R.~F., {Brotherton}, M., {Tripp}, T.~M., {Kaiser}, M.~E.,
1117:   \& {Kriss}, G.~A. 2002, \apj, 575, 687
1118: 
1119: \end{thebibliography}
1120: 
1121: 
1122: 
1123: \end{document}
1124: