1: %\documentclass[a4paper]{aa}
2: \documentclass[letterpaper]{aa}
3: \usepackage{natbib}\usepackage{graphicx}
4: \bibpunct{(}{)}{;}{a}{}{,} % refs A&A style
5:
6: \newcommand{\kms}{km\,s$^{-1}$} \newcommand{\sqcm}{cm$^{-2}$}
7: \newcommand{\hi}{\ion{H}{i}} \newcommand{\hw}{\ion{H}{ii}}
8: \newcommand{\os}{\ion{O}{vi}} \newcommand{\cf}{\ion{C}{iv}}
9: \newcommand{\sif}{\ion{Si}{iv}} \newcommand{\nf}{\ion{N}{v}}
10: \newcommand{\sit}{\ion{Si}{iii}}\newcommand{\ct}{\ion{C}{iii}}
11: \newcommand{\siw}{\ion{Si}{ii}} \newcommand{\cw}{\ion{C}{ii}}
12: \newcommand{\alth}{\ion{Al}{iii}} \newcommand{\few}{\ion{Fe}{ii}}
13: \newcommand{\ssix}{\ion{S}{vi}} \newcommand{\sfo}{\ion{S}{iv}}
14: \newcommand{\lya}{Lyman-$\alpha$}
15: \newcommand{\zgrb}{$z_{\rm GRB}$}
16:
17: \begin{document}
18:
19: \title{High-ion absorption in seven GRB host galaxies at
20: $z$=2--4\thanks{Based on observations taken under Programme IDs
21: 070.A-0599, 070.D-0523, 075.A-0385, 075.A-0603, 077.D-0661, and
22: 080.D-0526 with the Ultraviolet and Visual
23: Echelle Spectrograph (UVES) on the Very Large Telescope (VLT) Unit 2
24: (Kueyen) at Paranal, Chile, operated by ESO.}}
25: \subtitle{Evidence for both circumburst plasma and outflowing
26: interstellar gas}
27:
28: \author{Andrew J. Fox\inst{1} \and C\'edric Ledoux\inst{1}
29: \and Paul M. Vreeswijk\inst{2} \and Alain Smette\inst{1}
30: \and Andreas O. Jaunsen\inst{3}}
31: \institute{European Southern Observatory, Alonso de C\'ordova
32: 3107, Casilla 19001, Vitacura, Santiago, Chile; afox@eso.org \and
33: Dark Cosmology Centre, Niels Bohr Institute,
34: U. of Copenhagen, Juliane Maries Vej 30, 2100 Copenhagen, Denmark
35: \and Institute of Theoretical Astrophysics, University of Oslo, PO
36: Box 1029 Blindern, 0315 Oslo, Norway}
37:
38: \date{Accepted 16 September 2008, Received 15 September 2008, in
39: original form 29 May 2008}
40:
41: \authorrunning{Fox et al.}
42: \titlerunning{High ions in GRB host galaxies}
43:
44: \abstract{}
45: %Aims
46: {We use VLT/UVES high-resolution optical spectroscopy of
47: seven GRB afterglows (at \zgrb\ between 2.20 and 3.97) to investigate
48: circumburst and interstellar plasma in the host galaxies of the bursts.}
49: %Methods
50: {GRBs 050730, 050820, 050922C, 060607, 071031, and 080310 were each
51: detected by the \emph{Swift} satellite.
52: With the minimum possible time delay (as short as eight minutes),
53: follow-up optical spectroscopy at 6.0~\kms\ resolution
54: began with UVES. We present Voigt-profile component fits and
55: analysis of the high-ion absorption detected in these spectra and in
56: the pre-\emph{Swift} UVES spectrum of GRB~021004, at velocities
57: between \zgrb--5\,000~\kms\ and \zgrb.}
58: %Results
59: {
60: We identify several distinct categories of high-ion absorption at
61: velocities close to \zgrb:
62: (i) Strong high-ion absorption at \zgrb\ itself is always seen in
63: \os, \cf, and \sif, usually (in six of seven cases) in \nf,
64: and occasionally in \sfo\ and \ssix.
65: Three of the cases show log\,$N$(\nf)$>$14 in a single-component,
66: suggesting a circumburst origin, but we cannot rule out an
67: interstellar origin. Indeed, using the
68: non-detection of \sfo$^*$ at \zgrb\ toward GRB~050730 together with a UV
69: photo-excitation model, we place a lower limit of
70: 400~pc on the distance of the \sfo-bearing gas from the GRB.
71: (ii) Complex, multi-component \cf\ and \sif\ profiles
72: extending over 100--400~\kms\ around \zgrb\ are observed
73: in each spectrum; these velocity fields are similar to those measured
74: in \cf\ in damped \lya\ systems at similar redshifts, suggesting a
75: galactic origin.
76: (iii) Asymmetric, blueshifted, absorption-line wings covering
77: 65--140~\kms\ are seen in the \cf, \sif, and \os\
78: profiles in four of the seven GRB afterglow spectra.
79: The wing kinematics together with the ``Galactic'' \cf/\sif\ ratios
80: measured in two cases suggest that the wings
81: trace outflowing interstellar gas in the GRB host galaxies.
82: (iv) High-velocity (HV; 500--5\,000~\kms\ relative to \zgrb)
83: components are detected in six of the seven spectra; these components
84: are not necessarily a single homogeneous population, and many of them
85: may arise in unrelated foreground galaxies. However, in the cases of
86: GRBs~071031 and 080310, the ionization properties of the HV components
87: (very high \cf/\sif\ ratios and absence of neutral-phase absorption
88: in \siw\ or \cw) are suggestive of a circumburst origin; models of
89: Wolf-Rayet winds from the GRB progenitors can explain both
90: the kinematics and ionization level of these HV components.
91: }{}
92:
93: \keywords{gamma rays: bursts - galaxies: halos - galaxies:
94: high-redshift - galaxies: ISM - stars: Wolf-Rayet}
95:
96: \maketitle
97: \section{Introduction}
98: For many years the only backlights available for studying the
99: distant Universe through absorption-line spectroscopy were AGN.
100: Since the discovery that gamma-ray bursts (GRBs) are extragalactic
101: \citep{vP97, Me97}, GRBs have also been employed as backlights, with the
102: advantages of extremely high luminosity and power-law continua, but the
103: disadvantage of rapid temporal fading, implying that rapid-response
104: observations are necessary for their study.
105: Not only do analyses of GRB afterglow spectra allow intervening systems to
106: be studied, they also enable the properties of the interstellar gas in
107: the host galaxy of the GRB to be investigated, particularly its
108: kinematics, chemical abundances, dust content, molecular gas content,
109: and ionized gas content
110: \citep{Ca03, Kl04, Fi05, Sv03, Sv04, Pe06, Fy06, Pr06, Pr07a}.
111: Many GRB spectra show damped \lya\ (DLA) absorption at \zgrb,
112: i.e. they show a neutral gas column density
113: log\,$N$(\hi)$>$20.3 \citep{Je01, Hj03,
114: Ja04, Ja06, Vr04, St05, Be06, Wa06, Pr07b},
115: typically with higher \hi\ column densities than are seen in DLAs
116: toward QSOs \citep[QSO-DLAs, reviewed by][]{Wo05}.
117: The higher $N$(\hi) in GRB-DLAs may be
118: related to their intense star-formation activity \citep{Bl02}, or to
119: the GRB sight-lines passing preferentially through the inner
120: regions of the host galaxies \citep{Fy08, Pr08a}.
121: Photometric studies of GRB host galaxies have shown them to be
122: star-forming, dwarf galaxies \citep{Ch04, Wi07, Th07, Th08a}, with no
123: evidence that the hosts are peculiar \citep{Sv08}.
124:
125: In this paper we use rapid-response mode spectra of seven
126: high-redshift long-duration GRB afterglows to investigate high-ion
127: absorption (in \os, \nf, \cf, \sif, \sfo, and \ssix) in the ISM of the
128: host galaxy, as well as in the circumburst medium immediately
129: surrounding the GRB. Observations of ionized interstellar gas in
130: galactic environments provide important constraints on several
131: physical processes, including star formation and subsequent feedback,
132: galactic winds, interactions with satellite galaxies, and accretion.
133: In the pre-\emph{Swift} era, studying high ions in galaxy
134: environments at high redshifts was complicated by the difficulty of
135: associating individual absorbers with individual galaxies.
136: There are several \emph{indirect} ways to explore high-ion absorption
137: in high-redshift galaxies.
138: Firstly, outflowing \cf\ and \sif\ is seen in the composite spectrum of
139: high-redshift Lyman-break galaxies \citep[LBGs;][]{Sh03}.
140: Secondly, there is a statistical correlation between the redshifts of
141: \cf\ absorbers seen in QSO spectra and the redshifts of galaxies near
142: the QSO lines-of-sight \citep{Ad05}.
143: And thirdly, high-ion absorption is detected in all
144: DLAs \citep{WP00, Fo07a, Fo07b, Le08}, which trace galactic structures.
145: However, because of the faintness of high-redshift galaxies,
146: only in a single published case \citep[the lensed LBG cB58;][]{Pe00, Pe02}
147: are interstellar high-ion absorption lines seen at moderate-to-high
148: resolution in the spectra of \emph{individual} galaxies at $z\!>\!2$.
149:
150: Now, with the ability to be on-target within a few minutes of a
151: \emph{Swift} GRB trigger, a new means to directly probe the
152: interstellar (as well as circumburst) medium in individual high-redshift
153: galaxies is available \citep[see][]{Fi05, DE07, Ch07}.
154: High-resolution optical spectroscopy of GRB afterglows
155: now routinely produces spectra with high-enough signal-to-noise to
156: uncover the velocity sub-structure in the absorbing gas.
157: Evidence has even been found for time-variation in the strength of
158: absorption in fine-structure lines at \zgrb\ \citep{DZ06, Vr07}.
159: In this paper, we present a survey of high-ion absorption in the
160: highest-quality GRB optical afterglow spectra taken to date.
161: We pay particular attention to the high-ion kinematics, since velocity
162: measurements can constrain the presence of gaseous outflows, either
163: from the GRB progenitor or from the host galaxy.
164: We structure this paper as follows.
165: In \S2 we discuss the GRB observations, data processing, and our line
166: profile fitting procedures. In \S3 we describe the observational
167: properties of the high-ion absorption at velocities near \zgrb\ in
168: each afterglow spectrum. We discuss four distinct observational
169: categories of high-ion absorption in the GRB afterglow spectra in \S4,
170: and we present a summary in \S5.
171:
172: \section{GRB observations}
173: We formed our sample of seven GRB afterglow spectra by selecting
174: all GRBs at $z\!>\!2$ with follow-up high-resolution
175: UVES\footnote{UVES is described in \citet{De00}.} optical spectroscopy
176: available as of March 2008.
177: Lower redshift GRBs were not considered since no information on the
178: high ions (particularly \os) is available in these cases.
179: With the exception of GRB~021004, which was detected by the
180: \emph{High-Energy Transient Explorer (HETE-2)} satellite \citep{Ri03},
181: each of the GRBs in our sample was detected by
182: the Burst Alert Telescope (BAT) on-board
183: NASA's \emph{Swift} satellite \citep{Ge04}.
184: The seven GRBs are 021004, 050730, 050820, 050922C, 060607, 071031,
185: and 080310, named according to the UT date of their detection (yymmdd).
186: Following each trigger, \emph{Swift} began follow-up
187: observations with the XRT and UVOT instruments, to determine an
188: error circle of $\approx$2.5\arcsec\ around the source.
189:
190: After the \emph{Swift} localization was distributed through the GCN
191: (GRB Coordinates Network), rapid-response mode (RRM) UVES observations
192: began on each GRB afterglow with the minimum possible time delay,
193: provided the target was observable from Paranal. Three of the seven GRB
194: afterglows in our sample were observed in this mode.
195: RRM observations involve the immediate interruption of any current UVES
196: exposure in order to slew to and acquire the new target. No human
197: intervention is required, except for alignment of the target on the
198: spectrograph slit. On two occasions, a sub-ten minute response was
199: achieved. These observations demonstrate the power of the RRM concept,
200: and the success of its implementation on ESO telescopes.
201: The four afterglow spectra that were not taken in the RRM mode are
202: GRBs~021004, 050820, 050730 and 050922C, which either occurred
203: before the RRM was implemented or were unobservable from Paranal at
204: the time of the trigger. In each of these cases, the UVES
205: observations began at the earliest possible time in the form of a
206: traditional Target-of-Opportunity (ToO) observation.
207: A log of the observations, together with a summary of
208: the basic properties of the GRBs, is given in Table 1.
209:
210: The spectra were taken with 2$\times$2 binning and a 1.0\arcsec\ slit,
211: providing $R\approx50\,000$ (FWHM=6.0~\kms) spectra in
212: $\approx$2~\kms\ pixels.\footnote{Though the nominal spectral resolution of
213: UVES in this mode is 43\,000, we find that a higher resolution of
214: $\approx$50\,000 is achieved in practice, due to variations in the seeing
215: conditions.}
216: The UVES data were reduced with a customized version of the {\sc MIDAS}
217: reduction pipeline \citep{Ba00}, with the wavelength scale
218: subsequently corrected into the vacuum heliocentric frame.
219: For the latest GRB in our sample (GRB~080310), we also reduced the
220: data independently with version 3.4.5 of the CPL (Common Pipeline Library)
221: pipeline\footnote{See manual at
222: ftp://ftp.eso.org/pub/dfs/pipelines/uves/.}.
223: A detailed comparison of line profiles found the results from the two
224: pipeline versions to be essentially identical.
225: Individual exposures were combined with a least-squares weighting
226: to produce a combined spectrum.
227: In the case of GRB~071031, we also maintain and analyze the
228: individual exposures to look for time-series variation (see \S3.6.1). %*
229: Continua were fit locally to each absorption line of interest.
230: Before displaying and fitting the \os, \ct, \nf, and \ssix\ profiles,
231: which fall in low S/N regions of the spectrum, we perform a further
232: rebinning, generally by five pixels, but occasionally by other
233: factors, as described in the captions to Figures 1 to 4. These
234: differing rebinning factors are driven by the desire to maximize
235: the signal-to-noise ratio and to investigate the significance of the
236: high-ion detections, even if some resolution is compromised.
237:
238: \begin{table*}
239: \begin{minipage}[t]{18cm}
240: \caption{GRB sample and log of UVES observations}
241: \begin{tabular}{lcccc ccccc c}
242: \hline\hline
243: GRB & UT(Trig)\footnote{UT of trigger by the BAT instrument on-board
244: \emph{Swift}. Exception: GRB~020104, detected by \emph{WXM}
245: on-board \emph{HETE-2}.} & UT(UVES)\footnote{UT of start of first UVES
246: exposure (after target acquisition).} &
247: $\delta t$\footnote{Time delay between satellite trigger and start
248: of UVES exposure.} & $t_{\rm total}$\footnote{Total UVES exposure
249: time over all setups.} & ESO ID & \zgrb\ & log\,$N_{\rm
250: \hi}$\footnote{\hi\ column density and metallicity of GRB-DLA taken from
251: reference 8, except 071031 and 080310 values, from this %*
252: paper.} & [Z/H]$^e$ & GCN\footnote{We list here the GCN first
253: reporting the burst, and the GCN first reporting the UVES spectrum.}
254: & Science\footnote{We list references relevant to the study of the
255: host galaxy ISM.}\\
256: (yymmdd) & & (hh:mm) & (hh:mm) & (s) & & & & & Refs. & Refs.\\
257: \hline
258: 021004 & 12:06:13 & 01:37 & 13:31 & 7200 & 070.A-0599\footnote{Also 070.D-0523.}
259: & 2.3290 & 19.00 & 0.0 & 1,2 & 3,4,5,6,7,8\\% 01:37:37
260: 050730 & 19:58:23 & 00:07 & 04:09 & 6000 & 075.A-0603 & 3.9686 & 22.15 & $-$2.3 & 9,10 & 8,11,12,13,14,15,16\\
261: 050820 & 06:34:53 & 07:08 & 00:34 & 6043 & 075.A-0385 & 2.6147 & 21.00 & $-$0.6 & 17,18 & 8,13,15,16\\
262: 050922C & 19:55:50 & 23:42 & 03:47 & 6000 & 075.A-0603 & 2.1990 & 21.55 & $-$2.0 & 19,20 & 8,16,21\\
263: 060607 & 05:12:13 & 05:20 & 00:08 & 11980 & 077.D-0661 & 3.0749 & 16.80 & 0.0 & 22,23 & 8\\%05:20:06
264: 071031 & 01:06:36 & 01:15\footnote{Due to a difficult target acquisition,
265: the first fully aligned exposure of GRB~071031 began at 01:25.}& 00:09 & 9480 & 080.D-0526 & 2.6922 & 22.15 & $-$1.7 & 24,25 & ...\\%01:15:29
266: 080310 & 08:37:58 & 08:51 & 00:13 & 4680 & 080.D-0526 & 2.4274 & 18.80 & $-$1.4 & 26,27 & ...\\%08:50:46
267: \hline
268: \end{tabular} %*
269: \end{minipage}
270: REFERENCES: 1 \citet{Sh02}; 2 \citet{Sv02};
271: 3 \citet{Mo02}; 4 \citet{Mi03}; 5 \citet{Sc03}; 6 \citet{Fi05}; 7 \citet{La06}; 8 \citet{Pr08b};
272: 9 \citet{Ho05}; 10 \citet{DE05a};
273: 11 \citet{St05}; 12 \citet{Ch05}; 13 \citet{Ch07}; 14 \citet{DE07}; 15 \citet{Pr07a}; 16 \citet{Pr07b};
274: 17 \citet{Pa05}; 18 \citet{Le05};
275: 19 \citet{No05}; 20 \citet{DE05b}; 21 \citet{Ja06};
276: 22 \citet{Zi06}; 23 \citet{Le06};
277: 24 \citet{St07}; 25 \citet{Le07};
278: 26 \citet{Cu08}; 27 \citet{Vr08}.
279: \end{table*}
280:
281: We searched in each spectrum for high-ion absorption over the velocity
282: range \zgrb--5\,000~\kms\ to \zgrb+2\,000~\kms, where \zgrb\ is defined by
283: the position of strongest absorption in the low-ionization lines.
284: This wide range was chosen because of the possibility of detecting
285: high-velocity outflow features driven by the burst progenitor.
286: We make a distinction between low-velocity (LV) absorption components
287: seen within several hundred~\kms\ of \zgrb, which are seen in all cases, and
288: high-velocity (HV) components at 500-5\,000~\kms\ from \zgrb, which
289: are seen in 6/7 sight lines, and which are clearly separated from the LV
290: absorption. When absorption was found, we used the
291: VPFIT software package\footnote{Available at
292: http://www.ast.cam.ac.uk/$\sim$rfc/vpfit.html.} to fit the high-ion
293: components in each GRB spectrum with a series of Voigt profiles.
294: Each high ion was fit independently, even species with similar ionization
295: potentials. The VPFIT software accounts for instrumental resolution, and
296: returns the redshift, line width, and column density ($z$, $b$, and
297: log\,$N$) for each component together with the associated errors.
298: We convert the component redshift to velocity, using \zgrb\ as the
299: zero point. By summing over the component column densities, and adding
300: the component errors in quadrature, we form the total column density
301: and its error in each absorber.
302: In cases where the high-ion absorption is saturated, the
303: errors on the total column density are large.
304: The fit solutions are non-unique, since the number of components to be
305: fit has to be specified manually. This number is often larger than
306: ten, due to the complexity of the absorption-line profiles.
307: Nonetheless, our fitting technique has been tested by comparing the
308: total model column densities with the values from direct apparent
309: optical depth integrations \citep{Fo07a}. The successful outcome of
310: these tests supports the reliability of the fit results.
311: The details of the fits are given in Tables 2 to 8, for
312: GRBs 021004, 050730, 050820, 050922C, 060607, 071031, and 080310,
313: respectively. Atomic data (rest wavelengths and $f$-values)
314: were taken from \citet{Mo03}.
315:
316: \input{tab2.tex}
317: \input{tab3.tex}
318: \input{tab4.tex}
319: \input{tab5.tex}
320: \input{tab6.tex}
321: \input{tab7.tex}
322: \input{tab8.tex}
323:
324: \section{Description of the high-ion absorption}
325: High-ion absorption within several hundred \kms\ of \zgrb\ is found in
326: all seven GRB spectra in our sample.
327: The detected lines are \cf\ and \sif\ (seen in 7/7 cases),
328: \nf\ (6/7), \os\ (6/7), \sfo\ (5/7), and \ssix\ (2/7).
329: The high-ion absorption-line profiles in the velocity range within
330: 200~\kms\ of \zgrb\ are shown for each of our seven GRB afterglow
331: spectra in Figures 1 to 4. In addition to the high ions, we include on
332: each figure absorption lines that trace the neutral gas, typically
333: \siw\ $\lambda$1260.422, \cw\ $\lambda$1334.532, or
334: \ion{Fe}{ii} $\lambda$1608.451.
335:
336: In cases where we do not report a detection for \os, \sfo, and \ssix\
337: (e.g. \os\ toward GRB~060607), it is because the profiles are blended, not
338: because there are actual non-detections in these ions.
339: The only clear non-detection at \zgrb\ in any of the high ions mentioned above
340: is in \nf\ toward GRB~060607. In two cases our data cover the wavelength of
341: \ion{Ne}{viii} $\lambda\lambda$770.409, 780.324 at \zgrb, but
342: the doublet lies below the Lyman limit where no flux is transmitted.
343: \ion{P}{v} $\lambda\lambda$1117.977, 1128.008 is not detected in any of
344: the seven spectra. \ct\ $\lambda$977.020 is detected in one case
345: (GRB~071031), but in the other sight lines the S/N is too low at the
346: observed wavelength (far into the blue) for the \ct\ data to be useful.
347: Absorption in \sfo$^*$~$\lambda$1072.973 is not observed in any of the
348: seven spectra. Finally, with the exception of GRB~060607 with its
349: unusually low $N$(\hi), \sit\ $\lambda$1206.500 lies in the strong
350: damping wings of \lya, so no useful information on this ion at \zgrb\
351: is available.
352:
353: \begin{figure}
354: \includegraphics[width=8.0cm]{f1.eps}
355: \caption{Normalized high-ion and low-ion absorption-line profiles
356: on a velocity scale relative to \zgrb\ for GRB~021004.
357: Voigt profile fits are shown in blue (solid line), with the center of each
358: component in the fit marked with a red dash.
359: Dark shading denotes the strong absorption components at \zgrb.
360: All data have been rebinned by four pixels (eight in the case of
361: \os). We do not fit the \cf\ or
362: \sif\ doublets because of saturation. Absorption components at higher
363: velocity relative to \zgrb\ are shown in Figure 5.}
364: \end{figure}
365:
366: \begin{figure*}
367: \includegraphics[width=18cm]{f2.eps}
368: \caption{Normalized high-ion and low-ion absorption-line profiles on a velocity
369: scale relative to \zgrb\ for GRBs 050730 and 050820.
370: Voigt profile fits are shown in blue (solid line), with the center of each
371: component in the fit marked with a red dash.
372: Dark shading denotes the regions of strong absorption at \zgrb\
373: itself, and light shading denotes the negative-velocity wings
374: (GRB~050730 only). The \os, \nf, and \ssix\ data have been
375: rebinned by five pixels. Absorption components at higher
376: velocity relative to \zgrb\ are shown in Figure 5.}
377: \end{figure*}
378:
379: \begin{figure*}
380: \includegraphics[width=18cm]{f3.eps}
381: \caption{Normalized high-ion and low-ion absorption-line profiles
382: on a velocity scale relative to \zgrb\ for GRBs 050922C and 060607.
383: Voigt profile fits are shown in blue (solid line), with the center of each
384: component in the fit marked with a red dash.
385: Dark shading denotes the regions of strong absorption,
386: and light shading denotes the negative-velocity wing
387: (GRB~060607 only).
388: The \sfo\ and \ssix\ profiles are blended in each of these sight
389: lines. The \os\ and \nf\ data have been rebinned
390: by eight pixels for GRB~050922C, and the \nf\ data by five pixels
391: for GRB~060607. Absorption components at higher
392: velocity relative to \zgrb\ are shown in Figure 5.}
393: \end{figure*}
394:
395: \begin{figure*}
396: \includegraphics[width=18cm]{f4.eps}
397: \caption{Normalized high-ion and low-ion absorption-line profiles on a velocity
398: scale relative to \zgrb\ for GRBs 071031 and 080310. Voigt profile
399: fits are shown in blue (solid line), with the center of each
400: component in the fit marked with a red dash. Dark shading denotes
401: the regions of strong absorption, and light shading denotes the
402: negative-velocity wings. The \os\ and \nf\ data have been rebinned
403: by four pixels for GRB 071031 and by five pixels for GRB 080310. For
404: \ssix, the rebin factors are five and ten, respectively. Absorption
405: components at higher velocity relative to \zgrb\ are shown in Figure 5.}
406: \end{figure*}
407:
408: \begin{figure*}
409: \includegraphics[width=18cm]{f5a.eps}
410: \includegraphics[width=18cm]{f5b.eps}
411: \caption{Seven high-velocity (500--5\,000~\kms\ from \zgrb)
412: high-ion absorbers are seen in the seven GRB afterglow spectra in
413: our sample; these HV absorbers are shown here
414: (GRB~021004 shows two HV absorbers, GRB~050922C
415: shows none, and the other five GRB spectra show one each).
416: Voigt profile fits are only shown (with blue solid lines) for cases
417: where we judge that reliable column densities can be extracted.
418: The properties of these HV absorbers are diverse, and it is unlikely
419: that they represent a single population. Several of these HV
420: absorbers are seen in both neutral-phase and high-ion-phase
421: absorption; others (e.g. the HV components
422: toward GRB~071031 and GRB~080310) are seen only in high-ion transitions.}
423: \end{figure*}
424:
425: The detection rate of \nf\ in our sample is identical to that reported by
426: \citet{Pr08b}, who find \nf\ at \zgrb\ in 6/7 GRB afterglow spectra,
427: but since five systems are common to both samples, this is not surprising.
428: \nf\ has also been detected at \zgrb\ in the afterglow spectrum of
429: GRB~060206 \citep{Th08b}.
430: Other than the \os\ absorption in the GRB~050730 spectrum, which has been
431: published by \citet{DE07}, no reports of \os\ at \zgrb\ in GRB
432: afterglow spectra exist, and no previous reports of \sfo\ or \ssix\
433: exist either. Indeed, very few reports of interstellar \ssix\ absorption exist
434: anywhere in the high-$z$ \citep{CC99, Lv03} or low-$z$ \citep{Sa05b}
435: Universe. Interstellar \sfo\ absorption-line detections have
436: (to our knowledge) only been reported in quasar-intrinsic systems
437: \citep{Ar99, Ga06}. Finally, our non-detections of absorption in \sfo$^*$
438: $\lambda$1072.973 contrast with detections of this line
439: in the winds of OB-type stars \citep{Ma03, Le03}.
440:
441: The absorption features observed at and near \zgrb\
442: can be classified into various types.
443: First, strong absorption is seen exactly at \zgrb\ in the form of
444: saturated \os, \cf, and \sif\ lines, usually with accompanying \nf,
445: and (where the data are unblended) also in \sfo\ and \ssix.
446: Second, complex multi-component \cf\ and \sif\ profiles with total velocity
447: widths of 100--400~\kms\ are seen in all seven GRB afterglow spectra.
448: Third, 4/7 GRB afterglow spectra show asymmetric negative-velocity
449: absorption-line wings extending over velocity regions of
450: 65--140~\kms. These wings are seen clearest in \cf\ and \sif, but are
451: also present in \os.
452: Fourth, 6/7 cases show HV components in \cf, and occasionally also in
453: neutral-phase lines such as \siw\ and \cw.
454: Whereas the first three of these categories are visible
455: in Figures 1 to 4, we present a separate plot (Figure 5) showing the
456: HV high-ion absorbers detected in our sample. We return to a
457: discussion of these categories of absorption in \S4,
458: after describing the details of the high-ion absorption in each spectrum.
459: Unless otherwise stated, solar abundances were taken from \citet{Lo03}.
460:
461: \subsection{GRB~021004 \zgrb=2.3290}
462: Extremely strong \cf\ absorption is seen at \zgrb, with a fully
463: saturated trough spreading over 300~\kms.
464: \sif\ is also present and less saturated than \cf, but we cannot
465: discern the true component structure in \sif\ because of low S/N.
466: A component centered at 0~\kms\ is seen in \os, \nf, and
467: \sfo, with absorption also detected in \alth, \siw, \cw, and \few.
468: \sfo$^*$ $\lambda$1072.973 is not detected at \zgrb; we measure
469: log\,$N$(\sfo$^*$)$<$14.19 (3$\sigma$) over the velocity range $-$60 to
470: 60~\kms, and log\,$N$(\sfo)=15.30$\pm$0.29,
471: so that $N$(\sfo$^*$)/$N$(\sfo)$<$0.08. %*
472: Two HV absorbers are observed: one centered at $-$2\,900~\kms\ relative to
473: \zgrb, the other centered at $\approx\!-$600~\kms. The first is seen
474: in \sif, \cf, \nf, \siw, and \cw. The second is seen in \cf, \sif, and
475: \siw. These two HV absorption-line complexes are shown in separate
476: columns in Figure 5. The \os\ data are too blended in the HV
477: components to be of use. A full study of the UVES spectrum of
478: GRB~021004 is presented by \citet{Fi05}, and possible origins of the HV
479: components (e.g. a series of shells around the progenitor star)
480: are discussed in \citet{Mo02}, \citet{Mi03}, \citet{Sc03}, and \citet{La06}.
481: The velocity difference between the absorber at \zgrb\ and the HV absorber near
482: $-$500~\kms\ is equal to the velocity difference between the two lines
483: of the \cf\ doublet, i.e. line-locking is observed in \cf\
484: \citep[first noticed by][]{Mo02}\footnote{Line-locking is understood as
485: occurring when a cloud is accelerated by radiation pressure in a
486: given doublet line until the wavelength shifts into the shadow of
487: the other doublet line cast by a separate cloud \citep{Vi99}.} .
488:
489: \subsection{GRB~050730 \zgrb=3.9686}
490: High-ion absorption is seen in \os, \nf, \cf, \sif, \sfo, and \ssix\
491: centered at \zgrb. All of these lines except \nf\ are saturated.
492: The \nf\ profile shows two components separated by 27$\pm$5~\kms, each
493: with fairly narrow $b$-values (7$\pm$3 and 14$\pm$8~\kms); %*
494: we fit the \ssix\ absorption with one component with $b$=17$\pm$2~\kms. %*
495: \sfo$^*$ $\lambda$1072.973 is not detected at \zgrb; we measure
496: log\,$N$(\sfo$^*$)$<$13.70 (3$\sigma$) over the velocity range $-$25 to
497: 20~\kms, in which we measure log\,$N$(\sfo)=14.61$\pm$0.15,
498: so that $N$(\sfo$^*$)/$N$(\sfo)$<$0.12. %*
499: Note that this limit applies only to the strongest component of absorption.
500: The \cf, \sif, \os, and \cw\ profiles show a clear negative-velocity
501: absorption-line wing extending from $-$70 to $-$145~\kms.
502: Finally, a HV absorber at $-$1550~\kms\ is
503: seen with at least five \sif\ and \cf\ components spread over
504: $\approx$200~\kms\ \citep[see also][]{Ch07}.
505: This HV absorption-line complex is shown in Figure 5. %*
506: This GRB shows an extremely high \hi\ column of log\,$N$(\hi)=22.15,
507: and is the highest redshift GRB in our sample.
508: A full study of the neutral and excited-state lines at \zgrb\
509: in the UVES spectrum of GRB~050730 is presented by \citet{DE07}.
510:
511: \subsection{GRB~050820 \zgrb=2.6147}
512: \nf\ shows a single narrow component at $-$109~\kms\ %*
513: relative to \zgrb\ with $b$=9$\pm$3~\kms.
514: This \nf-\zgrb\ offset is the largest among our seven spectra,
515: and has been noticed before \citep{Pr08b}, but the formal significance
516: of the $\lambda$1238.821 and $\lambda$1242.804 detections is
517: only 3.1$\sigma$ and 2.0$\sigma$ respectively. %*
518: The \cf\ and \sif\ profiles exhibit at least ten components
519: spread over 400~\kms\ in velocity width, with a \cf/\sif\ ratio that
520: changes dramatically between components. \cf\ $\lambda$1550.781 is
521: blended so we only display \cf\ $\lambda$1548.204. A similar velocity
522: structure is observed at lower S/N in \alth\ $\lambda$1862.789.
523: \ssix\ $\lambda$933.378 shows three components aligned
524: with three components seen in \cf, strongly
525: suggesting the genuine presence of \ssix.
526: However, \ssix\ $\lambda$944.523 is affected by low S/N and possible
527: blending, so we treat the \ssix\ detection as marginal.
528: The \os\ profiles are highly blended and/or saturated, except for the
529: range $-$140 and $-$200~\kms\ in \os\ $\lambda$1031.926, where the profile
530: closely follows the shape of the \cf\ and \sif\ profile. Therefore
531: while we are not able to measure the \os\ absorption in this case, we
532: know that it is present.
533: \sfo\ $\lambda$1062.662 shows a profile that broadly follows the \sif.
534: The \sfo$^*$ $\lambda$1072.973 profile is difficult to interpret.
535: An absorption feature at 15~\kms does not correspond to any component
536: in any other high ion, so we identify it as a blend. However, we
537: cannot rule out the presence of genuine \sfo$^*$ absorption between
538: $-$50 and 0~\kms.
539: Finally a two-component HV absorber at $-$3\,660~\kms\ relative to \zgrb\ is
540: seen in \cf, but with no detectable \sif, \cw, or \siw\ (see Figure 5).
541:
542: \subsection{GRB~050922C \zgrb=2.1990}
543: A single, strong component in \os, \nf, \cf\ and \sif\ is detected at
544: 65~\kms\ relative to \zgrb. \alth\ is also detected in
545: this component in both $\lambda$1854.716 and $\lambda$1862.789.
546: Our best-fit to the \os\ line has $b$=24$\pm$2~\kms; for the
547: \nf\ line, we find $b$=16$\pm$3~\kms.
548: \cf\ and \sif\ show a complex series of at least
549: six components spread over $\approx$280~\kms, %*
550: but with no evidence for a negative-velocity wing.
551: The \sfo\ and \ssix\ data are blended and not shown on Figure 3. %*
552: This GRB shows a high log\,$N$(\hi) value of 21.60, and a low
553: metallicity of $-$2.0 \citep{Pr08b}.
554: This is the only GRB afterglow spectrum in our sample in which no HV
555: \cf\ components are detected.
556:
557: \subsection{GRB~060607 \zgrb=3.0749}
558: The high ions in the GRB~060607 spectrum behave differently than in the
559: other afterglow spectra. The \cf, \sif, and \sit\ profiles show a
560: four-component absorber centered at
561: \zgrb\ and covering only $\approx$100~\kms\ of total width.
562: However, \nf\ is not detected at \zgrb; we derive a 3$\sigma$ upper limit
563: of log\,$N$(\nf)$<$12.70 in the velocity range $-$140 to 50~\kms,
564: where \cf\ absorption is seen. %*
565: The \os, \sfo, and \ssix\ data are blended near \zgrb\ and are of no use.
566: \alth\ shows a non-detection.
567: The \cf\ profiles also show a weak (low optical depth) wing extending from
568: $-$60 to $-$120~\kms. %*
569: This GRB has an extremely low \hi\ column density,
570: log\,$N$(\hi)=16.8 \citep{Pr08b}, 3.5 orders of magnitude lower than
571: the DLA columns seen in five of the other six cases.
572: A HV absorber seen in \cf\ and \sif\ is
573: centered at $-$1\,850~\kms\ relative to \zgrb, with nearby low-ion
574: absorption present in \siw\ and \cw (Figure 5).
575:
576: We note that a sub-DLA at $z$=2.9372 is detected in the GRB~060607
577: spectrum with log\,$N$(\hi)=19.50$\pm$0.05
578: \citep[first noticed by][]{Le06}. This sub-DLA is separated
579: from the nominal GRB redshift of 3.0749 by $\Delta z$=0.1377, or
580: 10\,300~\kms, and shows strong, multi-component
581: absorption in \os, \cf, and \sif. We considered the
582: possibility that the host galaxy of GRB~060607 is actually located
583: at this redshift ($z$=2.9372), which would imply its
584: ISM properties are more similar to those of the host galaxy ISM
585: of the other bursts in our sample. However, this hypothesis is
586: unlikely to be correct since
587: (a) \lya\ forest lines are detected up to the redshift of 3.0749,
588: suggesting that \zgrb\ really is this high, and
589: (b) this scenario would require the absorber observed at $z$=3.0749 to be
590: in front of the host galaxy yet moving toward it at over
591: 10\,000~\kms, which is not easy to explain dynamically.
592: We conclude that the sub-DLA at 2.9372 traces a foreground galaxy
593: unrelated to the GRB.
594:
595: \subsection{GRB~071031 \zgrb=2.6922}
596: A strong component in \os, \nf, \cf, and \sif\ is detected at
597: \zgrb. The least-saturated high-ion line, \nf, shows
598: a narrow line width of $b$(\nf)=11$\pm$2~\kms, implying
599: log\,$T\!<\!5.08$ for the gas containing the N$^{+4}$ ions, regardless
600: of ionization mechanism. %*
601: No information is available on the line widths of \cf\ and \sif\ at
602: \zgrb\ because the absorption in these lines is saturated.
603: Whereas \ssix\ $\lambda$933.378 is blended, \ssix\ $\lambda$944.523
604: appears to show absorption at \zgrb, with a profile that matches
605: that of \os. However, because of the low S/N of the data at these
606: wavelengths, we do not attempt to derive a \ssix\ column density.
607: We measure log\,$N$(\sfo$^*$)$<$13.87 (3$\sigma$) over the velocity range
608: 0 to 50~\kms, but complete saturation (and possible blending) prevents
609: a reliable measurement of log\,$N$(\sfo) in this velocity range.
610: Absorption in several fine-structure lines of Fe, Ni, O, Si, and C is
611: detected at \zgrb\ in the GRB~071031 afterglow spectrum: in Figure 4
612: we show a line (\few$^{*****}$~$\lambda$1559.085) from the fifth
613: excited level ($^4$F$_{9/2}$) of \few. % (wavelength from NIST).
614: We derive an \hi\ column density log\,$N$(\hi)=22.15$\pm$0.05 and a
615: metallicity [Zn/H] of $-$1.72$\pm$0.06, where zinc is chosen since it
616: is typically undepleted onto interstellar dust grains.
617:
618: A notable feature of the \cf, \sif, \sfo, and (at lower S/N)
619: the \os\ profiles is the wing extending blueward from 0~\kms\ down
620: to $-$160~\kms. A component at $-$115~\kms\ is superimposed on the wing.
621: The \cf\ and \sif\ profiles follow one another closely, suggesting
622: these two ions form in the same physical regions of gas.
623: However, the slope d$N_{\rm a}(v)$/d$v$ is shallower for \os\ than for
624: \cf\ and \sif\ in the range $-$90 to $-$50~\kms\ (see \S4.3). %*.
625:
626: Three strong components of HV high-ion absorption are seen
627: (Figure 5), at velocities of $-$560, $-$510, and $-$370 \kms.
628: The $-$560 and $-$510~\kms\ components are seen in
629: \cf\ $\lambda$1548.204, \os\ $\lambda$1031.926,
630: \os\ $\lambda$1037.617, and \ct\ $\lambda$977.020, but are blended in \cf\
631: $\lambda$1550.781 and absent in \sif.
632: The $-$370~\kms\ component is detected in \ct, \cf, and \sif,
633: but the \os\ profiles are fully saturated at this velocity.
634: The \nf\ data show no significant detection in the HV gas,
635: though the S/N near the \nf\ lines is too low to place a strong
636: constraint on $N$(\nf).
637: Neutral-phase absorption is not seen in the three HV
638: components, even in the sensitive (high $f$-value) lines
639: \ion{O}{i} $\lambda$1302.169, \siw\ $\lambda$1260.422, and \cw\
640: $\lambda$1334.532, with the single exception of a detection of weak
641: \cw\ at $-$370~\kms. The $N$(\cf)/$N$(\sif) ratios change dramatically
642: between the various components, from 4.8$\pm$1.0 at $-$110~\kms, to %*
643: 40$\pm$9 at $-$370~\kms, %*
644: $>$140 (3$\sigma$) at $-$510~\kms\ and %*
645: $>$110 (3$\sigma$) at $-$560~\kms. %*
646:
647: \subsubsection{Time-series observations of GRB~071031}
648:
649: \begin{figure}
650: \includegraphics[width=9cm]{f6.eps}
651: \caption{Time-series observations of \cf\ and \sif\ absorption in the
652: GRB~071031 afterglow spectra. Each row shows a different observing
653: epoch, with the UT of the start of each
654: exposure annotated with the exposure duration on the right panel.
655: These data have been binned by four pixels to improve the S/N.
656: No evidence for time-variation in the high-ion profiles is seen.}
657: \end{figure}
658:
659: The combined UVES spectrum of GRB~071031 (shown in Figure 4) %*
660: was formed by co-adding four useful individual exposures, with
661: exposure times of 300, 600, 1200, and 2400~s (the sequence of
662: increasing exposure times was chosen to compensate for the
663: fading of the afterglow). Analysis of these
664: individual frames allows us to look for time-variation in the high-ion
665: absorption. We look for variation in the \cf\ and \sif\ profiles, since these
666: lines fall in parts of the spectrum where the S/N is highest
667: (a \nf\ time-series is unavailable).
668: We focus on the HV components, since the high-ion absorption
669: at \zgrb\ is saturated, making time-variation at that redshift
670: difficult to detect. The time-series spectra are shown in Figure 6.
671: No evidence for any evolution in the \cf\ or \sif\ profiles is
672: seen. The lack of variation can be quantified by measuring the
673: equivalent width at each observing epoch. The total equivalent width
674: and 1$\sigma$ error measurements in the range $-$600 to $-$330~\kms\
675: (over three HV components) are 266$\pm$88, 327$\pm$49, 233$\pm$36,
676: and 306$\pm$30~m\AA\ in the four exposures respectively, %*
677: consistent with no variation in the absorption with time.
678: Thus the behavior of the high-ion lines in these HV absorbers is in
679: contrast to the time-variation in low-ion and fine-structure
680: absorption lines at \zgrb\ reported by \citet{DZ06} and \citet{Vr07}.
681:
682: \subsection{GRB~080310 \zgrb=2.4274}
683: Strong high-ion absorption is seen in \os, \nf, \cf, \sfo, and \sif\
684: centered within 50~\kms\ of \zgrb.
685: The \nf\ profile shows two components separated by
686: 50$\pm$3~\kms\ and with $b$-values of 18$\pm$3 and 19$\pm$3~\kms. %*
687: The \ssix\ profiles are suggestive of absorption at \zgrb, but with
688: very low S/N and an inconsistency between the absorption in
689: $\lambda$933.378 and in $\lambda$944.523. For this reason, we do not
690: attempt to derive a \ssix\ column density. Similarly the \sfo\
691: $\lambda$1062.662 profile is too noisy to measure a reliable column, but
692: we can measure a lower limit of log\,$N$(\sfo)$>$14.77 over the
693: velocity range $-$70 to 30~\kms. Combining this with the constraint
694: log\,$N$(\sfo$^*$)$<$14.38 (3$\sigma$) from the non-detection of
695: \sfo$^*$~$\lambda$1072.973, we find $N$(\sfo$^*$)/$N$(\sfo)$<$0.41. %*
696: The \cf, \sif, and \os\ profiles show a clear negative-velocity
697: absorption-line wing extending from $-$120 to $-$260~\kms. %*
698: This is the most extended absorption-line wing in our sample.
699: Finally, a HV absorber at $-$1400~\kms\ is seen (see Figure 5)
700: with two \cf\ components, one of which is very narrow
701: ($b$=4$\pm$2~\kms). %*
702: For this GRB-DLA we derive an \hi\ column density
703: log\,$N$(\hi)=18.80$\pm$0.10 and a metallicity
704: [O/H] of $-$1.39$\pm$0.10. No ionization correction was applied since
705: a charge-exchange reaction closely links \ion{O}{i} and \hi, implying
706: [\ion{O}{i}/\hi]$\approx$[O/H] \citep{FS71}.
707:
708: \section{Discussion}
709: The six high-ion absorption lines reported in this paper
710: at velocities near \zgrb\ are
711: \sif, \sfo, \cf, \ssix, \nf, and \os, which trace the
712: Si$^{+3}$, S$^{+3}$, C$^{+3}$, S$^{+5}$, N$^{+4}$, and O$^{+5}$
713: ions, requiring energies of 33.5, 34.8, 47.9, 72.6, 77.5, and 113.9~eV
714: for their creation, respectively \citep{Mo03}.
715: A galactic ionizing spectrum, dominated by the integrated radiation
716: from O and B stars, drops strongly above 54~eV,
717: the \ion{He}{ii} ionization edge \citep{BH86}. So whereas the ions
718: Si$^{+3}$, S$^{+3}$, and C$^{+3}$ can be photoionized in the ISM by starlight,
719: the ions S$^{+5}$, N$^{+4}$, and O$^{+5}$ cannot.
720: We expect this to remain true in GRB host galaxies, even though
721: they have been shown to harbor significant numbers of
722: Wolf-Rayet stars \citep{Ha06}, since Wolf-Rayet spectra also show a
723: strong break at 54\,eV \citep{Cr07}. Instead, the detection of the
724: three ions S$^{+5}$, N$^{+4}$, and O$^{+5}$ implies the presence of
725: either a hard, non-stellar radiation source or hot, collisionally
726: ionized gas.
727:
728: In the seven GRB afterglow spectra in our sample, the
729: median and standard deviation of log\,$N$(\os) at \zgrb\ is 15.12$\pm$0.54.
730: For comparison, the median log\,$N$(\os) measured in 100 sight lines passing
731: through the Milky Way halo is 14.38 \citep{Wa03, Sa03}, rising to
732: 14.80 when including the contribution from high-velocity clouds \citep{Se03}.
733: Turning to \nf, we report a median log\,$N$(\nf)=14.03$\pm$0.49 in the
734: seven GRB spectra, whereas the median log\,$N$(\nf) in 32 Galactic
735: halo sight lines is 13.45 \citep{IS04}.
736: High-ion absorption in \nf\ and \os\ has also been observed in
737: $z$=2--3 DLA galaxies, with a median log\,$N$(\os)=14.77 \citep{Fo07a},
738: and 76 per cent of DLAs showing log\,$N$(\nf)$<$13.50 (Fox et
739: al. 2008, in preparation). %*
740: Thus the column densities of \os\ and \nf\ seen at \zgrb\ are, in the
741: mean, higher than those seen in the halo of the Milky Way and those in DLAs.
742: The highly ionized Galactic ISM
743: is a complex environment, with many physical
744: processes contributing to the production of the high ions,
745: including conductive interfaces, turbulent mixing layers, shocks,
746: expanding supernova remnants, and galactic fountain flows
747: \citep{Sa03, Zs03, IS04, Bo08}, and there is no reason why the
748: ISM in the host galaxies of GRBs should be any less complex.
749:
750: With this in mind, we now discuss the various categories of high-ion
751: absorption seen near \zgrb. Throughout this section, we refer the
752: reader to Table 9, which summarizes the high-ion to high-ion column
753: density ratios measured in the GRB afterglow spectra, and compares
754: them to the ratios measured in various other astrophysical environments.
755:
756: \begin{table}
757: \begin{minipage}[t]{\columnwidth} %*
758: \caption{High-ion column density ratios}
759: \begin{tabular}{lccc}
760: \hline\hline
761: GRB\footnote{We report the observed column density ratios
762: in different velocity regions: in the strong components seen at
763: \zgrb, in the blueshifted wings, and in the HV
764: absorbers. All limits are 3$\sigma$. In cases marked $^*$,
765: saturation, blending, or low S/N prevents
766: the ratio from being measured reliably.} &
767: $\frac{N({\rm \cf})}{N({\rm \os)}}$ &
768: $\frac{N({\rm \cf})}{N({\rm \sif)}}$ &
769: $\frac{N({\rm \nf})}{N({\rm \os)}}$\\
770: \hline
771: \bf{At $z$(GRB)}\\
772: 050922C & $^*$ & $^*$ & 0.21$\pm$0.04\\ %VPF
773: 060607 & $^*$ & 7.6$\pm$1.2 & $^*$ \\ %VPF
774: \hline
775: \bf{Blueshifted Wings}\\
776: 050730 & 0.40$\pm$0.25 & 3.7$\pm$0.9 & $<$0.09 \\ %AOD-TAB10
777: 060607 & $^*$ & $>$13 & $^*$ \\ %AOD-TAB10
778: 071031 & 1.6$\pm$0.4 & 3.2$\pm$0.9 & $<$0.11 \\ %AOD-TAB10
779: 080310 & 0.55$\pm$0.11 & 25$\pm$4 & $<$0.05 \\ %AOD-TAB10
780: \hline
781: \bf{HV components}\\
782: 021004 $-$2900~\kms\ & $^*$ & 5.1$\pm$3.1 & $^*$\\ %VPF-SiIV/VPF-CIV
783: 050730 $-$1550~\kms\ & $^*$ & 3.7$\pm$0.4 & $^*$\\ %VPF-SiIV/VPF-CIV
784: 050820 $-$3660~\kms\ & $^*$ & $>$24 & $^*$\\ %AOD-SiIV/VPF-CIV
785: 060607 $-$1850~\kms\ & $^*$ & 3.2$\pm$0.5 & $^*$\\ %VPF-SiIV/VPF-CIV
786: 071031 $-$560~\kms\ & 0.19$\pm$0.06 & $>$110 & $<$0.06\\ %VPF-OVI-CIV/AOD-NV-SiIV
787: 071031 $-$510~\kms\ & 0.35$\pm$0.10 & $>$140 & $<$0.07\\ %VPF-OVI-CIV/AOD-NV-SiIV
788: 071031 $-$370~\kms\ & $^*$ & 40$\pm$9 & $^*$\\ %VPF-SiIV/VPF-CIV
789: 080310 $-$1400~\kms\ & $^*$ & $>$36 & $^*$\\ %AOD-SiIV/VPF-CIV
790: \hline
791: Galactic Halo\footnote{Mean$\pm$1$\sigma$ in 16 Milky Way halo sight lines
792: \citep{Zs03}.}
793: & 0.60$\pm$0.47 & 3.46$\pm$1.09 & 0.12$\pm$0.07\\
794: LMC\footnote{Range observed in LMC gas (not including \hw\
795: regions) in four sight lines \citep{LH07}.}
796: & $<$0.09--0.59 & 1.5--2.5 & $<$0.33\\ %$<$0.10 to
797: DLAs, $z$=2--3\footnote{Mean$\pm$1$\sigma$ in
798: 10 DLAs \citep{Fo07a}.}
799: & 2.1$\pm$0.7 & 5.4$\pm$2.2 & ... \\
800: IGM, $z$=2--4 & 0.35\footnote{Median value among four IGM absorbers
801: measured by \citet{Si06}.}
802: & 16$\pm$4\footnote{Median$\pm$1$\sigma$
803: in 188 IGM absorbers \citep{Bo03}.} & $<$0.23\footnote{Measured in
804: composite QSO spectrum by \citet{LS93}.}\\
805: \hline
806: \end{tabular}
807: \end{minipage}
808: \end{table}
809:
810: \subsection{Strong absorption at \zgrb}
811: We always detect strong high-ion absorption
812: exactly at \zgrb\footnote{In two cases the velocity centroid of the
813: strongest high-ion absorption component differs from the nominal GRB redshift
814: (measured from the low-ion lines) by more than 10~\kms: for
815: GRB~050922C the offset is 65~\kms, and for GRB~071031 the offset is
816: 25~\kms. However, since the GRB redshifts are not known to better
817: than a precision of several tens of \kms, these offsets are not
818: significant, and we treat the strong high-ion absorption as arising
819: exactly at \zgrb.},
820: in the form of saturated \os, \cf, and \sif\ components,
821: and (where the data are unblended) also in \sfo\ and \ssix.
822: \nf\ is present in 6/7 cases, with the advantage of being less
823: saturated than the other high-ion lines. Because of this, \nf\ is the
824: best line to investigate the line width and optical
825: depth in the strong absorption components. In the six cases where \nf\ is
826: detected, we fit eight \nf\ components, with $b$-values of
827: 7$\pm$3, 9$\pm$3, 11$\pm$2, 14$\pm$8, 16$\pm$3, 18$\pm$3,
828: 19$\pm$3 and 28$\pm$3~\kms. The median of these eight values is 16~\kms. %*
829: A \nf\ component formed in gas at 200\,000~K, the temperature at which
830: the production of \nf\ by collisions is maximized \citep{GS07},
831: has a $b$-value of 15.4~\kms. Therefore, we are unable to rule out
832: collisional ionization for the gas traced by \nf.
833:
834: \subsubsection{Photoionized circumburst gas?}
835: Noticing that the \nf\ absorption at \zgrb\ tends to be stronger and narrower
836: (in total velocity width) than the \nf\ absorption observed in the
837: Milky Way ISM and in DLAs, \citet{Pr08b} have recently argued that the
838: \nf\ absorption components at \zgrb\ do not trace ISM material in the
839: halo of the host galaxy, but rather arise in circumburst gas in the
840: immediate vicinity of the GRB. Our high-resolution UVES dataset,
841: covering a wider range of ionization states than has been observed
842: before, allows us to investigate this intriguing idea.
843:
844: Our data are qualitatively consistent with the circumburst
845: hypothesis in the cases of GRBs~021004, 050922C, and 071031, where we
846: see strong, single components of \nf\ at \zgrb, each with
847: log\,$N$(\nf)$>$14.0, aligned with other high ions including \os\ with
848: log\,$N$(\os)$\ga$14.5. For these cases,
849: our new measurements of the \os, \sfo, and \ssix\ column densities at
850: \zgrb\ (listed in Tables 2, 4, and 7) could be used to constrain
851: photoionization models of the circumburst region.
852: However, in the remaining four spectra in our sample
853: (GRBs~050730, 050820, 060607, and 080310) the situation is
854: more complex. In each of these four cases, log\,$N$(\nf)$<$14.1, and
855: in two cases log\,$N$(\nf)$<$13.2, i.e the observed \nf\ is \emph{not} in
856: excess of what is expected from the host galaxy's ISM.
857: In two of these cases (GRBs 050730 and 080310) the \nf\
858: absorption profiles show two components separated by a few tens of
859: \kms\ -- this is not the signature of a single burst.
860: In another case (GRB~050820) there is an offset of 110~\kms\ between a
861: single \nf\ component and the strongest absorption in the other high ions.
862: And finally, in the case of GRB~060607, no \nf\ is detected at all.
863: In addition to these case-by-case details, we repeat that
864: the $b$-values of the detected \nf\ absorption components at \zgrb\
865: are \emph{not} particularly narrow, so photoionization is not required
866: by the line widths (though it is not ruled out either).
867:
868: \subsubsection{Total ionized column density}
869: We can further investigate the nature of the strong components seen at
870: \zgrb\ by calculating the total column density of ionized hydrogen
871: $N$(hot~\hw)\footnote{Here we use the word ``hot'' simply to label the
872: \nf-bearing gas; the upper limit on the temperature is
873: 215,000\,K for the median \nf\ $b$-value of 16~\kms.}
874: contained in these components. To do this we have to correct for ionization
875: and metallicity, so the calculation is only possible for cases where
876: the nitrogen abundance has been measured in the DLA at \zgrb. This is
877: true for three GRBs in our sample: [N/H]=$-$3.16$\pm$0.10 for GRB~050730,
878: $>\!-$1.35 for GRB~050820, and $<\!-$4.09 for GRB~050922C \citep{Pr07b}.
879: The relationship between the \nf\ and hot \hw\ column densities can be
880: written in the following way:
881: \begin{equation}
882: N({\rm hot~\hw})=\frac{N{\rm (\nf)}}{{\rm (\nf/N)(N/H)_n}}\frac{{\rm
883: (N/H)_{n}}}{({\rm N/H)_{\rm i}}},
884: \end{equation}
885: where (N/H)$_{\rm i}$ denotes the nitrogen abundance in the ionized gas
886: and (N/H)$_{\rm n}$ denotes the nitrogen abundance in the neutral gas
887: (which is measured).
888: We take the solar nitrogen abundance of (N/H)$_\odot=10^{-4.22}$
889: from \citet{Gr07}, and then (N/H)$_{\rm n}=10^{\rm [N/H]}$(N/H)$_\odot$.
890: If the gas is collisionally ionized, as would be
891: appropriate for \nf\ in the host galaxy ISM, \nf/N$<$0.25 at all
892: temperatures \citep{GS07}\footnote{This is true for either equilibrium
893: or non-equilibrium models. However, in the \citet{Pr08b}
894: GRB photoionization models, a \nf\ ionization
895: fraction as high as 0.6 is predicted.}.
896: Therefore we adopt 0.25 as the maximum allowed ionization
897: fraction, corresponding to a minimum ionization correction of a factor of 4.
898: If the ionized gas has the same metallicity as the
899: neutral gas, so that the term (N/H)$_{\rm n}$/(N/H)$_{\rm i}$ in
900: Equation 1 is equal to one, then $N$(hot~\hw) is:\\
901: $>$21.8 for GRB~050730 [which has log\,$N$(\hi)=22.15],\\
902: $>$19.3 for GRB~050820 [which has log\,$N$(\hi)=21.00], and \\
903: $>$22.7 for GRB~050922C [which has log\,$N$(\hi)=21.55],\\
904: where the results are lower limits since (a) we have used the smallest
905: allowed ionization correction, and (b) they only apply to the
906: \nf-bearing gas, and \hw\ may exist at other temperatures as well.
907: Comparing the values of $N$(hot~\hw) and $N$(\hi), we find
908: the hot-ionized-to-neutral ratio $N$(hot~\hw)/$N$(\hi) takes values of
909: $>$0.4 and $>$0.02 for GRBs~050730 and 050820, and $>$13 for
910: GRB~050922C. Whereas the first two values are modest, the
911: GRB~050922C value is extremely (and implausibly) high -- it
912: requires over ten times as much mass in hot ionized gas in a single
913: component as there is in neutral gas in the entire host galaxy.
914: However, if the nitrogen abundance in the ionized gas was higher than in
915: the neutral gas, so that (N/H)$_{\rm n}$/(N/H)$_{\rm i}<1$, then solutions
916: with lower $N$(hot~\hw) would be possible. Thus our calculation of the
917: total ionized column density at \zgrb\ suggests that the
918: metallicity in the highly ionized gas is higher than the metallicity in the
919: neutral gas (particularly for GRB~050922C); however, this calculation
920: is inconclusive in determining whether the gas is circumburst or interstellar.
921:
922: \subsubsection{Photo-excitation modeling of $N$(\sfo) and $N$(\sfo$^*$)}
923: \begin{figure}
924: \includegraphics[width=9cm]{f7.eps}
925: \caption{Analysis of the \sfo\ level populations at \zgrb\ toward
926: GRB~050730. The solid and dashed lines show the predictions from
927: UV photo-excitation models for the time-variation of the \sfo\
928: and \sfo$^*$ column densities (see text for details).
929: The data points show the observed $N$(\sfo)
930: and upper limit to $N$(\sfo$^*$) in the UVES spectrum, plotted at
931: the $x$-position corresponding to the mid-observation epoch
932: (five hours after the trigger).
933: The solid lines show the model with the best-fit distance of 418$\pm$9~pc.
934: The dashed lines show a model with the distance $d$ fixed at 200~pc,
935: for illustration. In this case $N$(\sfo$^*$) would be 0.5 dex stronger than
936: observed, so this model is clearly ruled out. The result that the
937: \sfo-absorbing gas must lie at $d>400$~pc implies an interstellar
938: rather than circumburst origin.}
939: \end{figure}
940:
941: Another method for distinguishing between circumburst gas and more
942: distant interstellar gas is to look for absorption from
943: fine-structure levels. In circumburst gas, UV pumping by GRB photons will
944: populate the excited electronic levels, which decay and cascade to
945: populate the fine-structure levels of the ground state
946: \citep{Pr06, Ch07, Vr07}. Because of the rapid fading of the
947: afterglow, the level populations evolve in a non-equilibrium,
948: time-dependent manner. Our data allow us to investigate this
949: process in the \sfo\ ion, since in four GRB spectra
950: the combination of a detection of \sfo\ $\lambda$1062.662 and a
951: non-detection of \sfo$^*$~$\lambda$1072.973 allows us to
952: place an upper limit on the \sfo$^*$/\sfo\ ratio at \zgrb.
953: The Einstein $A$-value of the forbidden transition between
954: the ground state of \sfo\ and its fine-structure
955: level is $7.7\times10^{-3}$~s$^{-1}$
956: \citep[taken from NIST;][]{Ra08}\footnote{See
957: http://physics.nist.gov/PhysRefData/ASD/index.html}.
958: In other words, in the absence of stimulated photo- and collisional
959: de-excitation, the fine-structure level probed by the \sfo$^*$
960: 1072.973 transition will decay in 1/$A$=130\,s.
961: In fully ionized gas (as expected near GRBs) there is no \lya\ opacity
962: and so the \sfo\ lines (which lie shortward of \lya) will
963: not be shielded from UV radiation.
964: For GRBs~021004, 050730, and 080310 we measure 3$\sigma$
965: limits to the $N$(\sfo$^*$)/$N$(\sfo) ratio of $<$0.08, $<$0.12,
966: and $<$0.41 respectively (for the other four afterglows we are unable to
967: constrain this ratio due to blending). %*
968:
969: Following the technique described in \citet{Vr07}, we can model
970: the \sfo\ excitation level in gas subject to UV radiation from the GRB,
971: assuming that photo-excitation is the dominant excitation process.
972: We chose to model the \sfo\ level population measured toward
973: GRB~050730, because this case has the strongest constraint
974: on $N$(\sfo$^*$). The model has five free parameters: the distance
975: $d$ between the GRB and the absorbing gas (assuming all the
976: \sfo\ arises at the same distance), the total \sfo\ column
977: density $N_{\rm tot}$ (equivalent to the pre-burst ground-state \sfo\
978: column density), the afterglow spectral slope $\beta$ (where
979: $F_\nu\propto\nu^{\beta}$), the Doppler $b$-parameter of the
980: absorbing gas, and the rest-frame time when the calculation begins
981: $t_0$ (i.e. the time when the afterglow photons begin to be tracked in
982: the code). For a given set of these parameters, the model predicts the
983: \sfo\ and \sfo$^*$ column densities as a function of time
984: (both in the observer frame and in the rest frame).
985: Since the time of mid-observation of GRB~050730 is known, five hours
986: (observer frame) after the trigger, predicted values for $N$(\sfo) and
987: $N$(\sfo$^*$) can be extracted from each model.
988:
989: The parameters in our GRB~050730 model were determined as follows.
990: $N_{\rm tot}$ was chosen to equal $N$(\sfo)+$N$(\sfo$^*$), assuming the
991: actual $N$(\sfo$^*$) is equal to the measured upper limit
992: (if the actual value is lower, our conclusions are strengthened;
993: see below). The value for $\beta$ ($-$0.56) was determined using the
994: observed light curve for this afterglow.
995: The parameter $t_0$ is poorly constrained, but our runs showed
996: that the \sfo\ level populations are very insensitive to $t_0$ for
997: values between 0 and 300~s; we used a value $t_0$=10\,s \citep[see][]{Vr07}.
998: Finally a $b$-value of 10~\kms\ was chosen as a typical value for
999: interstellar gas, and indeed matches (within the 1-$\sigma$ error) the
1000: measured $b$-value for the \sfo\ component at 0~\kms\ toward
1001: GRB~050730. This leaves the distance $d$ as the one genuinely floating
1002: parameter in the model; a grid of models at different distance $d$ was
1003: generated, and the predicted and observed $N$(\sfo) and
1004: $N$(\sfo$^*$) were compared using a chi-squared minimization routine.
1005:
1006: The model results are shown in Figure 7.
1007: The best-fit distance $d$ is 418$\pm$9~pc.
1008: Because we assumed that the actual \sfo$^*$ column density
1009: was equal to the measured upper limit, this represents a strong
1010: \emph{lower limit} of 400~pc (2$\sigma$) on the distance of the
1011: absorbing gas from the GRB. In other words, if the gas was at
1012: $d\!<\!400$~pc, we would have seen \sfo$^*$ absorption, whereas we
1013: clearly do not. This represents an important result, since it implies
1014: that the \sfo\ absorption has an interstellar rather than circumburst origin.
1015: The distance limit on the \sfo\ absorption does not necessarily have
1016: to apply to the \nf\ absorption (or the other high ions).
1017: However, the \nf\ and \sfo\ profiles
1018: at \zgrb\ toward GRB~050730 are similar, both showing the
1019: component near 0~\kms\ and a weaker component at 25~\kms, suggesting
1020: that the two ions are co-spatial and hence that the \nf\ absorption
1021: also arises at $>$400~pc from the GRB.
1022:
1023: \subsubsection{Origin of strong absorption}
1024: In summary, we find that while the circumburst hypothesis
1025: explains the observation that in three afterglow spectra, strong
1026: single-component \nf\ is aligned with saturated absorption in the
1027: other high ions, it is challenged by the \nf\ profiles in the other four
1028: datasets, it is not strictly required by the \nf\ line widths, and the
1029: expected high-ion fine-structure lines and time-variable column
1030: densities are yet to be observed. Furthermore, in one case (GRB~050730)
1031: we are able to place a lower limit on the distance to the high-ion
1032: absorbing gas of 400~pc, using the simultaneous detection of \sfo\
1033: and non-detection of \sfo$^*$.
1034: Therefore, based on our current dataset, we cannot rule out an
1035: interstellar explanation for the strong high-ion components at \zgrb,
1036: noting that since the higher $N$(\hi) values in GRB-DLAs than in
1037: QSO-DLAs suggest that GRB sightlines pass through the inner star-forming
1038: regions of the host galaxies \citep{Pr08a, Fy08}, one would
1039: \emph{expect} high column densities of interstellar plasma at
1040: \zgrb, so the strength of the \nf\ absorption does not, in our view,
1041: argue against an interstellar origin.
1042: We also note that a strong, apparently narrow (though unresolved) \nf\
1043: component is observed in the spectrum of LBG cB58 at the galaxy redshift
1044: \citep[$z$=2.73;][]{Pe02}, showing that such components can be formed
1045: even in the absence of a GRB.
1046: Time-series observations of \nf\ in GRB afterglow spectra, which
1047: unfortunately are not available in the current dataset, would help to
1048: resolve this issue. A significant detection of time-variation in the
1049: \nf\ column density at \zgrb\ \citep[see models by][]{Pr08b}
1050: would represent the discovery of gas that is unambiguously close to
1051: the GRB. %*
1052:
1053: \subsection{Multi-component absorption in \cf\ and \sif}
1054: The multi-component structure seen in \cf\ and \sif\ spreading over
1055: several hundred~\kms\ around \zgrb\ is not seen in the other high
1056: ions. However, the \cf\ and \sif\ profiles lie out of the \lya\ forest, so
1057: are free from blending and show higher S/N than \os\ or \nf,
1058: and it is unclear how the \os\ and \nf\ profiles would appear if
1059: observed at the same S/N. These complex, multi-component \sif\ and \cf\
1060: profiles are reminiscent of the profiles
1061: of these ions in damped \lya\ (DLA) absorbers
1062: \citep{Lu96, Le98, WP00, Fo07b, Le08}, which
1063: represent galaxy halos seen in absorption toward a background source.
1064: The median (mean) value of the total \cf\ velocity width
1065: in 74 DLAs and sub-DLAs at $z$=2--3 reported by \citet{Fo07b} is
1066: 255~\kms\ (342~\kms), where the width is the total observed range of
1067: absorption, regardless of optical depth. In the seven GRB spectra in
1068: our current sample, the median (mean) total \cf\ velocity width
1069: measured in an identical manner is 280~\kms\ (320~\kms), %*
1070: similar to the DLA value
1071: (note we have excluded the HV components, which we treat separately).
1072: A precise comparison of the median \cf\ column density between
1073: the GRB sample and the DLA sample is not possible, because these
1074: measurements are affected by saturation, but we can state that
1075: log\,$N$(\cf) at \zgrb\ is $\ga$15 in all seven GRB spectra, which
1076: is significantly stronger than the median DLA value
1077: log\,$N$(DLA \cf)=14.27$\pm$0.57 measured by \citet{Fo07b}.
1078: Nonetheless, the similarities in total \cf\ line width between GRB and
1079: DLAs support a galactic origin for the multiple components seen in
1080: \cf\ and \sif\ in a velocity range of several hundred \kms\ around \zgrb.
1081:
1082: \subsection{Negative-velocity (blueshifted) absorption-line wings}
1083: One notable feature present in four of the seven GRB afterglow spectra
1084: in our sample is
1085: a negative-velocity absorption-line wing seen in \cf, \sif, and
1086: (at lower S/N) \os. These wings are characterized by relatively smooth,
1087: asymmetric absorption extending for 65--140~\kms\ %*
1088: to negative velocities relative to the strong absorption at \zgrb.
1089: To explore the high-ion velocity structure in the wings,
1090: we present in Figure 8 apparent column density profiles of the high
1091: ions over the velocity of the wings, calculated using
1092: $N_a(v)=3.768\times10^{14}\tau_a(v)/f\lambda$, where
1093: the apparent optical depth
1094: $\tau_a(v)={\rm ln}\,[F_c(v)/F(v)]$, $F(v)$ and $F_c(v)$ are the
1095: observed flux level and estimated continuum level
1096: as a function of velocity, and where $\lambda$ is in \AA\
1097: \citep{SS91}. We follow the error treatment of \citet{SS92}.
1098: This form of display allows the velocity structure seen in each ion to
1099: be closely compared. In two cases we add the profile of \cw\
1100: $\lambda$1334.532, which also shows wing absorption. The high-ion column
1101: densities in the wings are summarized in Table 10.
1102:
1103: \input{tab10.tex}
1104: \begin{figure*}
1105: \includegraphics[width=18cm]{f8.eps}
1106: \caption{High-ion apparent column density profiles
1107: in the four negative-velocity wings in our sample.
1108: Errors reflect both statistical noise and continuum placement uncertainties.
1109: The \cf\ and \sif\ profiles have each been scaled by the factors
1110: annotated on the plot to allow for comparison of their shape with \os.
1111: Only lower limits (arrows) can be presented
1112: for saturated pixels. The \os\ data have been rebinned by four
1113: pixels in the case of GRB~071031, eight pixels in the case of
1114: GRB~050922C, and five pixels in the other cases.
1115: \nf, \sfo, and \ssix\ are not detected in the wings.}
1116: \end{figure*}
1117:
1118: We find that the \cf\ and \sif\ profiles track one another closely in
1119: three of the four wings, indicating that the two ions form in the same
1120: regions of gas. The exception is the very weak wing seen toward
1121: GRB~060607, in which \sif\ is not detected, but \sit\ is seen.
1122: For the wings seen toward GRBs~050730 and 080310, the \os\ and \cf\
1123: profiles share a common gradient d$N_a$/d$v$. However, for GRB~071031,
1124: the \os\ absorption shows a shallower slope.
1125: We measure a $N$(\cf)/$N$(\sif) ratio of $\approx$3 in two of the four
1126: wings (GRBs~050730 and 071031), similar to the Galactic halo average of
1127: 3.46$\pm$1.09 measured by \citet{Zs03}.
1128: This supports an interstellar origin for the absorption-line wings.
1129: In the wings toward GRBs~060607 and 080310 the $N$(\cf)/$N$(\sif) ratio is much
1130: higher, taking values of $>$13 (3$\sigma$) and 25$\pm$4 respectively. %*
1131:
1132: \os\ absorption-line wings with similar velocity extent have been
1133: observed at zero redshift in the ultraviolet spectra of quasars taken
1134: with the \emph{Far-Ultraviolet Spectroscopic Explorer (FUSE)} satellite
1135: \citep{Se01, Se03, Fo05, Fo06, Sa05a, Ke06}\footnote{Analogous \cf\
1136: and \sif\ wings are yet to be observed at zero redshift, but these
1137: two ions fall in the near-UV, requiring \emph{HST} observations
1138: rather than \emph{FUSE}, and the number of high-resolution QSO
1139: spectra taken with \emph{HST} is small compared to the number of
1140: extragalactic sight-lines observed with \emph{FUSE}.}.
1141: \citet{Fo06} measured eleven Galactic \os\ wings,
1142: and report a mean and standard deviation \os\ column density
1143: of log\,$N$(\os)=13.85$\pm$0.45, with each wing extending over
1144: $\sim$100~\kms; these authors successfully modeled the \os\
1145: absorption-line wings as tracing Galactic outflows moving under
1146: ballistic trajectories.
1147: Furthermore, a strong \os\ wing covering a similar velocity range
1148: is seen in a sub-DLA (an absorber with
1149: 19.0$<$log\,$N$(\hi)$<$20.3) at $z$=2.67 \citep{Fo07c}.
1150: The \cf\ profile observed in the LBG cB58 at $z$=2.7 shows a clear
1151: absorption wing extending from $-$250 to $-$750~\kms\
1152: ($\approx$ nine resolution elements)
1153: in the galaxy rest-frame \citep{Pe02}, and an asymmetric blueshifted
1154: absorption feature in \ion{Mg}{ii} has been reported in the composite
1155: spectrum of 1400 $z\!\approx\!1.4$ galaxies by \citet{We08}, who also
1156: interpret it as a galactic outflow signature.
1157:
1158: \begin{figure*}
1159: \includegraphics[width=18.5cm]{f9.eps}
1160: \caption{Montage of apparent column density profiles of \os\
1161: absorption-line wings measured in various environments: near \zgrb\
1162: in three GRB afterglow spectra (this paper, three left-most lines),
1163: at zero redshift in the Milky Way halo in the UV spectra of three
1164: AGN \citep[][ three right-most lines]{Fo06}, and in a
1165: sub-DLA at $z$=2.6562 \citep{Fo07c}. The Milky Way and sub-DLA cases
1166: have been inverted and velocity offsets have
1167: been added for ease of comparison, but the gradient of each profile
1168: is preserved. Linear fits are shown with color-coded dashed lines,
1169: and the gradient of each fit in units of $10^9$~ions~\sqcm~(\kms)$^{-2}$
1170: is annotated with its 1$\sigma$ error on the panel.
1171: In these diverse galactic environments sampling over twelve billion
1172: years, the \os\ absorption-line profiles can be used to trace and
1173: quantify galactic outflows.}
1174: \end{figure*}
1175:
1176: Whereas the $z$=0 \os\ wings are always detected at \emph{positive}
1177: velocities (redshifted) relative to the Local Standard of Rest (LSR),
1178: the GRB wings are seen at \emph{negative} velocities (blueshifted)
1179: relative to \zgrb.
1180: This behavior is strongly suggestive of galactic outflows:
1181: in the Milky Way case (at least for high-latitude and anti-center sight
1182: lines, where Galactic rotation effects are insignificant),
1183: outflowing gas appears redshifted relative to the LSR, whereas in the
1184: GRB case, outflowing gas moves toward us along the line-of-sight and
1185: hence appears blueshifted relative to the host galaxy.
1186: The possibility remains that some wing absorbers represent chance
1187: alignments of unrelated components separated by small velocities,
1188: which are unresolved by the spectrograph.
1189: However, we find this explanation unlikely for the strongest wings
1190: in our sample, seen toward GRBs~071031 and 080310, since in these
1191: cases the column density in the wing increases
1192: monotonically with velocity\footnote{In the case of GRB~071031, the
1193: \cf\ profile shows a smooth gradient with velocity once the
1194: $-$115~\kms\ component has been subtracted off.}
1195: in an interval of over 100~\kms, and there is no reason why a series
1196: of random components should be aligned in this way.
1197:
1198: To illustrate the similarity in the wing absorption features observed at
1199: both low and high redshift, we show in Figure 9 a comparison of the
1200: \os\ apparent column density profiles in the zero redshift and GRB
1201: sight line wings. We have inverted the positive-velocity wings to
1202: compare their gradient with the wings observed in the GRB spectra.
1203: While the similarity in the overall
1204: shapes of the wing profiles is striking, the high-$z$ wings in the GRB
1205: host galaxies show higher \os\ column densities (by a factor of $\approx$5) %*
1206: and steeper gradients d$N_{\rm a}$/d$v$ (by a factor of 5--10) %*
1207: than the Milky Way wings, implying higher mass flow rates.
1208:
1209: \subsection{High-velocity components at 500--5\,000~\kms}
1210: In six of our seven GRB spectra (all GRBs except 050922C), high-ion
1211: components at 500--5\,000~\kms\ relative to \zgrb\ are observed (see
1212: Figure 5). Such HV components have been observed before
1213: in GRB~020813 \citep{Ba03},
1214: GRB~021004 \citep{Mo02, Mi03, Sc03, Fi05, La06},
1215: GRB~030226 \citep{Kl04},
1216: GRB~050505 \citep{Be06}, and
1217: GRB~050730 \citep{DE07, Ch07}.
1218: \citet{Ch07} report that \cf\ components with $W_{\rm r}\!>\!200$~m\AA\
1219: occur at 1\,000--5\,000~\kms\ in 20 per cent of GRB afterglow spectra.
1220: In our data, the incidence of HV \cf\ absorbers (6/7, or 86\%)
1221: is much higher because (a) we have a higher sensitivity: the weakest
1222: HV \cf\ absorber in our sample shows $W_{\rm r}$=48~m\AA,
1223: and (b) we classify absorbers in the range 500--1\,000~\kms\ as HV.
1224:
1225: The HV absorbers seen in GRB afterglow spectra
1226: do not represent a homogeneous population, as do the blueshifted
1227: absorption-line wings and (potentially) the strong, narrow absorbers at \zgrb.
1228: On the contrary, we observe a broad range of properties among the HV
1229: absorbers. In three sight lines (toward GRBs~021004, 050730, and 060607), %*
1230: the HV components exhibit low-ionization absorption
1231: (clearest in \cw\ $\lambda$1334.532 and \siw\ $\lambda$1260.422).
1232: However, the HV components toward GRBs~050820, 071031 and 080310 show no
1233: (or very weak) neutral-phase absorption, in either \cw\ or \siw. So whereas the
1234: presence of neutral gas in the first three cases implies the absorbing
1235: material is not formed in the circumburst medium
1236: \citep[because this medium is expected to be fully ionized by the
1237: burst; e.g.][]{Vr07, Ch07, Pr08b}, we find that the properties of the
1238: HV components in the latter three cases \emph{are}
1239: consistent with an origin in the close circumburst medium.
1240: This is particularly true for the HV gas toward GRB~071031, which
1241: is detected in the form of strong, \os\ and \cf\
1242: components that are not only absent in \cw\ and \siw, but also
1243: show non-detections of \sif. We have placed limits of
1244: $N$(\cf)/$N$(\sif) $>$110 at $-$560~\kms, and $>$140 at $-$510~\kms\
1245: in the GRB~071031 spectrum. These extreme ratios are far higher than
1246: typical interstellar values \citep[Galactic ISM values for this ratio
1247: are $\approx$3;][]{Zs03} implying these HV components have a
1248: different, non-interstellar origin.
1249:
1250: Given the connection between GRBs and massive stars \citep{MW99, WB06},
1251: a leading explanation for the HV components in GRB afterglow spectra
1252: is that they trace outburst episodes from the massive-star
1253: progenitors of the GRBs \citep[e.g.][]{Sc03}, and thus arise in the
1254: circumburst region. As pointed out by \citet{Be06}, a Wolf-Rayet
1255: outflow model would give rise to absorbers that are enriched in carbon
1256: and deficient in silicon, naturally explaining the
1257: high \cf/\sif\ ratios measured in the HV absorbers.
1258:
1259: If the wind-blown bubbles around GRB progenitors
1260: are similar in size to those observed around Wolf-Rayet stars in the
1261: Milky Way, their typical radius will be $\approx$2--10~pc \citep{Gr00}.
1262: Several groups have recently modeled Wolf-Rayet winds and their
1263: relationship to the HV components in GRB afterglow spectra
1264: \citep{Ra01, Ra05, Cv04, El06}, and in particular, predictions for the
1265: high-ion signature of Wolf-Rayet winds are given by \citet{vM05,vM07,vM08}.
1266: Depending on observing epoch and viewing angle,
1267: components at 150-700~\kms\ and 1\,000-2\,000~\kms\ are
1268: predicted at various stages of Wolf-Rayet evolution \citep{vM05};
1269: these velocities match the observed range of HV components in our sample.
1270: Recent models of the absorption-line signatures of galactic winds
1271: \citep[as opposed to single-star winds;][]{Fa07, KR07, Sa08} generally
1272: predict high-ion components at velocities of only 100--200~\kms\
1273: relative to the galaxy, though observations support the idea that
1274: galactic winds can reach speeds of $\approx$1\,000~\kms\ \citep{Fr02,
1275: Tr07, We08}. Nonetheless, the fairly narrow $b$-values measured in
1276: the HV components toward GRBs~071031 and 080310
1277: (not to mention the lack of \sif) are at odds with the
1278: broader, collisionally ionized components expected in galactic winds
1279: \citep{OD06}. Thus we conclude that among the available models,
1280: Wolf-Rayet winds are the best explanation for the highly ionized HV
1281: components toward GRBs~071031 and 080310.
1282:
1283: \section{Summary}
1284: We have performed a systematic study of the high-ion absorption near
1285: \zgrb\ in seven GRB afterglow spectra observed with UVES.
1286: Three of the seven were observed within 15 minutes of the trigger
1287: by the {\it Swift} satellite. This search has provided high-resolution
1288: (6.0~\kms\ FWHM) profiles of absorption in the ions \os, \nf, \cf,
1289: \sif, \sfo, and \ssix. The spectra contain a wealth of useful
1290: information on the properties of circumburst and interstellar gas in
1291: the GRB host galaxies. We have presented the results of Voigt
1292: profile fits to the high-ion absorption. Analysis of the line
1293: profiles shows that several types of high-ion absorption exist at
1294: $z\!\approx\!z_{\rm GRB}$, which we summarize in the following points.
1295:
1296: \begin{enumerate}
1297: \item
1298: We detect strong high-ion absorption components exactly at \zgrb,
1299: always in \os, \cf, and \sif, usually in \nf, and occasionally
1300: (where the data are unblended) in \sfo\ and \ssix. The high-ion
1301: column densities are significantly higher than those seen in the
1302: Milky Way ISM and in DLAs. We focus on \nf\ since it is the least
1303: saturated and hence best-measured high ion. In three of seven
1304: afterglow spectra, the strong \nf\ absorption takes the form of a
1305: single component coincident in velocity with saturated absorption in
1306: the other high ions, appearing to support the recent conclusion of
1307: \citet{Pr08b} that GRBs can photoionize \nf\ in circumburst gas.
1308: However, in the remaining four cases, the \nf\ absorption is either
1309: multi-component, offset from the other high-ions, or absent, and it
1310: is unclear whether the circumburst hypothesis can explain these
1311: cases. In addition, the \nf\ $b$-values (median of 16~\kms) are not
1312: narrow enough to rule out collisional ionization. Finally, for
1313: GRB~050730 we have placed a lower limit on the distance to the
1314: \sfo-absorbing gas of 400~pc, based on comparing photo-excitation
1315: models to the observed column density of \sfo\ and upper limit to
1316: \sfo$^*$. This supports an interstellar (rather than circumburst)
1317: origin for the strong high-ion absorption components at \zgrb.
1318:
1319: \item In addition to the strong component at \zgrb, the \cf\ and \sif\
1320: absorption exhibits complex, multi-component profiles extending over
1321: several hundred \kms\ (excluding the HV components, which we treat
1322: separately). The median total \cf\ line width in our seven spectra
1323: is 280~\kms, similar to the value it takes in high-redshift DLA %*
1324: galaxies. This supports a galactic ISM origin for these multiple components.
1325:
1326: \item We detect asymmetric, blueshifted absorption-line wings in the
1327: \cf, \sif, and (at lower S/N) \os\ profiles in four of the seven GRB
1328: afterglow spectra. These wings are similar in shape and velocity
1329: extent ($\approx$100--150~\kms) to positive-velocity wings seen in
1330: \os\ absorption at zero redshift in the Milky Way halo, and two of
1331: the four wings show \cf/\sif\ ratios of $\approx$3, %*
1332: equivalent to the value measured in the Milky Way ISM.
1333: We thus interpret the wings as tracing outflowing interstellar
1334: gas. The \os\ column densities in the GRB wings are, in the mean, five
1335: times higher than the \os\ column densities in the Milky Way wings
1336: (implying similarly higher mass flow rates), %*
1337: and the gradient d$N_{\rm a}$/d$v$ is typically 5--10 times higher. %*
1338:
1339: \item HV (500--5000~\kms) components are seen in \cf\ in six of the seven
1340: spectra (all GRBs except 050922C).
1341: These components are not necessarily a single homogeneous
1342: population, and many may arise in unrelated foreground galaxies.
1343: However, in the (previously unpublished) cases of GRBs~071031 and
1344: 080310, the ionization properties of the HV components (very high
1345: \cf/\sif\ ratios and an absence of neutral-phase absorption in \siw\
1346: or \cw) are consistent with a circumburst origin. Models of
1347: Wolf-Rayet winds from the GRB progenitor stars are capable of
1348: explaining the kinematics and ionization level of these HV components.
1349: \end{enumerate}
1350:
1351: \emph{Acknowledgments}
1352: We thank the ESO staff for their implementation of the rapid-response mode.
1353: PMV acknowledges the support of the EU under a Marie Curie
1354: Intra-European Fellowship, contract MEIF-CT-2006-041363.
1355: We thank Nicolas Lehner, Chris Howk, and Hsiao-Wen Chen
1356: for comments on the manuscript.
1357:
1358: \begin{thebibliography}{}
1359: \bibitem[Adelberger et al.(2005)]{Ad05} % IGM C IV outflow
1360: Adelberger, K. L., Shapley, A. E., Steidel, C. C., Pettini, M., Erb,
1361: D. K., \& Reddy, N. A., 2005, \apj, 629, 636
1362:
1363: \bibitem[Arav et al.(1999)]{Ar99} % S IV in BALs
1364: Arav, N., Korista, K. T., de Kool, M., Junkkarinen, V. T., \&
1365: Begelman, M. C. 1999, \apj, 516, 27
1366:
1367: \bibitem[Ballester et al.(2000)]{Ba00} % UVES reduction pipeline
1368: Ballester, P., Modigliani, A., Boitquin, O., et al. 2000, The
1369: Messenger, 101, 31
1370:
1371: \bibitem[Barth et al.(2003)]{Ba03} %GRB 020813 afterglow spectroscopy z=1.255
1372: Barth, A. J., Sari, R., Cohen, M. H., et al. 2003, \apj, 584, L47
1373:
1374: \bibitem[Berger et al.(2006)]{Be06} % GRB-DLA in 050505 LRIS AG spec z=4.27
1375: Berger, E., Penprase, B. E., Cenko, S. B., et al. 2006, \apj, 642, 979
1376:
1377: \bibitem[Bloom et al.(2002)]{Bl02}%GRB-SF connection
1378: Bloom, J. S., Kulkarni, S. R., \& Djorgovski, S. G. 2002, \aj, 123, 1111
1379:
1380: \bibitem[Boksenberg et al.(2003)Boksenberg, Sargent, \& Rauch]{Bo03} %IGM CIV
1381: Boksenberg, A., Sargent, W. L. W., \& Rauch, M. 2003, preprint
1382: (astro-ph/0307557)
1383:
1384: \bibitem[Bowen et al.(2008)]{Bo08} %OVI in Disk
1385: Bowen, D. V., Jenkins, E. B., Tripp, T. M., et al. 2008, \apjs,
1386: 176, 59
1387:
1388: \bibitem[Bregman \& Harrington(1986)]{BH86} %54eV break
1389: Bregman, J. N., \& Harrington, J. P. 1986, \apj, 309, 833
1390:
1391: \bibitem[Castro et al.(2003)]{Ca03} % GRB000926 Keck/HST imaging
1392: Castro, S., Galama, T. J., Harrison, F. A., et al. 2003, \apj, 586, 128
1393:
1394: \bibitem[Chen et al.(2005)]{Ch05} % GRB DLA 050730 MIKE/Magellan
1395: Chen, H.-W., Prochaska, J. X., Bloom, J. S., \& Thompson, I. B. 2005,
1396: \apj, 634, L25
1397:
1398: \bibitem[Chen et al.(2007)]{Ch07} % HV C IV in 1/5 GRB spectra
1399: Chen, H.-W., Prochaska, J. X., Ramirez-Ruiz, E., et al. 2007, \apj, 663, 420
1400:
1401: \bibitem[Chevalier et al.(2004)]{Cv04}%WR wing models
1402: Chevalier, R. A., Li, Z.-Y., \& Fransson, C. 2004, \apj, 606, 369
1403:
1404: \bibitem[Christensen et al.(2004)]{Ch04}%GRB host galaxies
1405: Christensen, L., Hjorth, J., \& Gorosabel, J. 2004, \aap, 425, 913
1406:
1407: \bibitem[Churchill \& Charlton(1999)]{CC99} %S VI detection
1408: Churchill, C. W., \& Charlton, J. C. 1999, \aj, 118, 59
1409:
1410: \bibitem[Crowther(2007)]{Cr07} %ARAA WR STARS
1411: Crowther, P. A. 2007, \araa, 45, 177
1412:
1413: \bibitem[Cummings et al.(2008)]{Cu08} %GCN 080310
1414: Cummings, J. R., Baumgartner, W. H., Beardmore, A. P., et al. 2008, GCN, 7382
1415:
1416: \bibitem[Dekker et al.(2000)]{De00} %UVES design
1417: Dekker, H., D'Odorico, S., Kaufer, A., Delabre, B., \& Kotzlowski,
1418: H. 2000, SPIE, 4008, 534
1419:
1420: \bibitem[Dessauges-Zavadsky et al.(2006)]{DZ06} %TV in 020813 SiII*, FeII*
1421: Dessauges-Zavadsky, M., Chen, H.-W., Prochaska, J. X., Bloom, J. S.,
1422: \& Barth, A. J. 2006, \apj, 648, L89
1423:
1424: \bibitem[D'Elia et al.(2005a)]{DE05a} % GCN 050730 UVES
1425: D'Elia, V., Melandri, A., Fiore, F., et al. 2005a, GCN, 3746
1426:
1427: \bibitem[D'Elia et al.(2005b)]{DE05b} % GCN 050922C UVES
1428: D'Elia, V., Piranomonte, S., Fiore, F., et al. 2005b, GCN, 4044
1429:
1430: \bibitem[D'Elia et al.(2007)]{DE07} % OVI/CIV 050730 Full Study
1431: D'Elia, V., Fiore, F., Meurs, E. J. A., et al. 2007, \aap, 467, 629
1432:
1433: \bibitem[Eldridge et al.(2006)]{El06} % WR wind in GRB model
1434: Eldridge, J. J., Genet, F., Daigne, F., \& Mochkovitch, R. 2006,
1435: \mnras, 367, 18
1436:
1437: \bibitem[Fangano et al.(2007)]{Fa07} % Gal Wind Absorption Sign.
1438: Fangano, A. P. M., Ferrara, A., \& Richter, P. 2007, \mnras, 381, 469
1439:
1440: \bibitem[Field \& Steigman(1971)]{FS71} % OI/HI
1441: Field, G. B., \& Steigman, G. 1971, \apj, 166, 59
1442:
1443: \bibitem[Fiore et al.(2005)]{Fi05} % UVES obs of GRB 021004 (and 020813)
1444: Fiore, F., D'Elia, V., Lazzati, D., et al. 2005, \apj, 624, 853
1445:
1446: \bibitem[Fox et al.(2005)]{Fo05} % PG 0953 wing
1447: Fox, A. J., Wakker, B. P., Savage, B. D., et al. 2005, \apj, 630, 332
1448:
1449: \bibitem[Fox et al.(2006)Fox, Savage, \& Wakker]{Fo06} % OVI wings
1450: Fox, A. J., Savage, B. D., \& Wakker, B. P. 2006, \apjs, 165, 229
1451:
1452: \bibitem[Fox et al.(2007a)]{Fo07a} % OVI in DLAs
1453: Fox, A. J., Petitjean, P., Ledoux, C., \& Srianand, R. 2007a, A\&A,
1454: 465, 171
1455:
1456: \bibitem[Fox et al.(2007b)]{Fo07b} % CIV in DLAs
1457: Fox, A. J., Ledoux, C., Petitjean, P., \& Srianand, R. 2007b, A\&A,
1458: 473, 791
1459:
1460: \bibitem[Fox et al.(2007c)]{Fo07c} % OVI wing in sub-DLA Q0331
1461: Fox, A. J., Petitjean, P., Ledoux, C., \& Srianand, R. 2007c, ApJ,
1462: 668, L15
1463:
1464: \bibitem[Frye et al.(2002)]{Fr02} % 500-800 km/s winds in lensed hi-z gals
1465: Frye, B. L., Broadhurst, T., \& Ben\'itez, N. 2002, \apj, 568, 558
1466:
1467: \bibitem[Fynbo et al.(2006)]{Fy06} %GRB 060206 z=4.048 N.Opt.Tel (Canaries)
1468: Fynbo, J. P. U., Starling, R. L. C., Ledoux, C., et al. 2006,
1469: \aap, 451, L47
1470:
1471: \bibitem[Fynbo et al.(2008)]{Fy08} %GRB host galaxies vs DLA vs LBG
1472: Fynbo, J. P. U., Prochaska, J. X., Larsen, J. M.,
1473: Dessauges-Zavadsky, M., \& M\o ller, P. 2008, \apj, 683, 321
1474:
1475: \bibitem[Ganguly et al.(2006)]{Ga06} % S IV in intrinsic systems
1476: Ganguly, R., Sembach, K. R., Tripp, T. M., Savage, B. D., \& Wakker,
1477: B. P. 2006, \apj, 645, 868
1478:
1479: \bibitem[Gehrels et al.(2004)]{Ge04} % SWIFT REFERENCE
1480: Gehrels, N., Chincarini, G., Giommi, P., et al. 2004, \apj, 611, 1005
1481:
1482: \bibitem[Gnat \& Sternberg(2007)]{GS07} %new CIE,NECI
1483: Gnat, O., \& Sternberg, A. 2007, \apjs, 168, 213
1484:
1485: \bibitem[Grevesse et al.(2007)]{Gr07} %Latest Abundances Solar
1486: Grevesse, N., Asplund, M., \& Sauval, A. J. 2007, Space Sci. Rev., 130, 105
1487:
1488: \bibitem[Gruendl et al.(2000)]{Gr00} % WR wind-blown bubbles
1489: Gruendl, R. A., Chu, Y.-H., Dunne, B. C., \& Points, S. D. 2000, \aj,
1490: 120, 2670
1491:
1492: \bibitem[Hammer et al.(2006)]{Ha06} %WR stars in GRB host galaxies
1493: Hammer, F., Flores, H., Schaerer, D., Dessauges-Zavadsky, M.,
1494: Le Floc'h, E., \& Puech, M. 2006, \aap, 454, 103
1495:
1496: \bibitem[Hjorth et al.(2003)]{Hj03} % GRB-DLA 020124 z=3.198, N.O.T.
1497: Hjorth, J., M\o ller, P., Gorosabel, J., et al. 2003, \apj, 597, 699
1498:
1499: \bibitem[Holland et al.(2005)]{Ho05} % GCN 050730
1500: Holland, S. T., Barthelmy, S., Burrows, D. N., et al. 2005, GCN, 3704
1501:
1502: \bibitem[Indebetouw \& Shull(2004)]{IS04} % OVI,NV in halo
1503: Indebetouw, R., \& Shull, J. M. 2004, \apj, 607, 309
1504:
1505: \bibitem[Jakobsson et al.(2004)]{Ja04} % GRB-DLA 030429 z=2.658 FORS
1506: Jakobsson, P., Fynbo, J. P. U., Ledoux, C., et al. 2004, \aap, 427, 785
1507:
1508: \bibitem[Jakobsson et al.(2006)]{Ja06} % GRB-DLA sample N(HI) distribution
1509: Jakobsson, P., Fynbo, J. P. U., Ledoux, C., et al. 2006, \aap, 460, L13
1510:
1511: \bibitem[Jensen et al.(2001)]{Je01} % 1st GRB-DLA 000301C NOT z=2.04 N(HI)=21.2
1512: Jensen, B. L., Fynbo, J. P. U., Gorosabel, J., et al. 2001, \aap, 370, 909
1513:
1514: \bibitem[Kawata \& Rauch(2007)]{KR07} %Galactic Wind Sign Hi-z Galaxies
1515: Kawata, D., \& Rauch, M. 2007, \apj, 663, 38
1516:
1517: \bibitem[Keeney et al.(2006)]{Ke06}%HV OVI absorbers
1518: Keeney, B. A., Danforth, C. W., Stocke, J. T., et al. 2006, \apj,
1519: 646, 951
1520:
1521: \bibitem[Klose et al.(2004)]{Kl04} % 030226 z>1.96, low-resn FORS spec+imaging
1522: Klose, S., Greiner, J., Rau, A., et al. 2004, \aj, 128, 1942
1523:
1524: \bibitem[Lazzati et al.(2006)]{La06} %021004 z=2.32 photo models foll-up Fi05
1525: Lazzati, D., Perna, R., Flasher, J., Dwarkadas, V. V., \& Fiore, F.
1526: 2006, \mnras, 372, 1791
1527:
1528: \bibitem[Ledoux et al.(1998)]{Le98} % Early anal C IV vs neutral line width
1529: Ledoux, C., Petitjean, P., Bergeron, J., Wampler, E. J., \& Srianand,
1530: R. 1998, \aap, 337, 51
1531:
1532: \bibitem[Ledoux et al.(2005)]{Le05} %GCN 050820 UVES
1533: Ledoux, C., Vreeswijk, P. M., Ellison, S., et al. 2005, GCN, 3860
1534:
1535: \bibitem[Ledoux et al.(2006)]{Le06} % RRM on UVES 060607
1536: Ledoux, C., Vreeswijk, P. M., Smette, A., Jaunsen, A., \& Kaufer,
1537: A. 2006, GCN, 5237
1538:
1539: \bibitem[Ledoux et al.(2007)]{Le07} %GCN z_GRB 071031
1540: Ledoux, C., Jakobsson, P., Jaunsen, A. O., et al. 2007, GCN, 7023
1541:
1542: \bibitem[Lehner \& Howk(2007)]{LH07} % LMC High Ions
1543: Lehner, N., \& Howk, J. C. 2007, \mnras, 377, 687
1544:
1545: \bibitem[Lehner et al.(2008)]{Le08} % High-ion DLA
1546: Lehner, N., Howk, J. C., Prochaska, J. X., \& Wolfe, A. M. 2008,
1547: \mnras, 390, 2
1548:
1549: \bibitem[Lehner et al.(2003)]{Le03} % SIV/SIV* in OB star winds
1550: Lehner, N., Fullerton, A. W., Massa, D., Sembach, K. R., \& Zsarg\'o,
1551: J. 2003, \apj, 589, 526
1552:
1553: \bibitem[Levshakov et al.(2003)]{Lv03} % Narrow OVI HE 0515-4414 & S VI
1554: Levshakov, S. A., Agafonova, I. I., Reimers, D., \& Baade, R. 2003,
1555: \aap, 404, 449
1556:
1557: \bibitem[Lodders(2003)]{Lo03}%Solar abundances
1558: Lodders, K. 2003, \apj, 591, 1220
1559:
1560: \bibitem[Lu \& Savage(1993)]{LS93} % NV/OVI limit
1561: Lu, L., \& Savage, B. D. 1993, \apj, 403, 127
1562:
1563: \bibitem[Lu et al.(1996)]{Lu96} % DLA abundances, DLA C IV
1564: Lu, L., Sargent, W. L. W., Barlow, T. A., Churchill, C. W., \& Vogt,
1565: S. S. 1996, \apjs, 107, 475
1566:
1567: \bibitem[MacFadyen \& Woosley(1999)]{MW99} %WR Collapsar
1568: MacFadyen, A. I., \& Woosley, S. E. 1999, \apj, 524, 262
1569:
1570: \bibitem[Massa et al.(2003)]{Ma03} %SIV/SIV* in Hot Star Winds
1571: Massa, D., Fullerton, A. W., Sonneborn, G., \& Hutchings, J. B. 2003,
1572: \apj, 586, 996
1573:
1574: \bibitem[Metzger et al.(1997)]{Me97} % GRB extragalactic
1575: Metzger, M. R., Djorgovski, S. G., Kulkarni, S. R., et al. 1997,
1576: Nature, 387, 878
1577:
1578: \bibitem[Mirabal et al.(2003)]{Mi03} % 021004 High-ion shells Keck/LRISs <1pc
1579: Mirabal, N., Halpern, J. P., Chornock, R, et al. 2003, \apj, 595, 935
1580:
1581: \bibitem[M\o ller et al.(2002)]{Mo02} % 021004 NOT Line-locking mentioned
1582: M\o ller, P., Fynbo, J. P. U., Hjorth, J., et al. 2002, \aap, 396, L21
1583:
1584: \bibitem[Morton(2003)]{Mo03} % Atomic Data
1585: Morton, D. C. 2003, \apjs, 149, 205
1586:
1587: \bibitem[Norris et al.(2005)]{No05} %GCN 050922C
1588: Norris, J., Barbier, L., Burrows, D., et al. 2005, GCN, 4013
1589:
1590: \bibitem[Oppenheimer \& Dav\'e(2006)]{OD06} % Wind sim IGM enrichment
1591: Oppenheimer, B., \& Dav\'e, R. 2006, \mnras, 373, 1265
1592:
1593: \bibitem[Page et al.(2005)]{Pa05} % GCN 050820
1594: Page, M., Burrows, D., Beardmore, A., et al. 2005, GCN, 3830
1595:
1596: \bibitem[Penprase et al.(2006)]{Pe06} %high-res, high S/N HIRES GRB 051111
1597: Penprase, B. E., Berger, E., Fox, D. B., et al. 2006, \apj, 646, 358
1598:
1599: \bibitem[Pettini et al.(2000)]{Pe00} % Wind from MS1512 LBG Paper I
1600: Pettini, M., Steidel, C. C., Adelberger, K. L., Dickinson, M., \&
1601: Giavalisco, M. 2000, \apj, 528, 96
1602:
1603: \bibitem[Pettini et al.(2002)]{Pe02} % Wind cB58 LBG II, ESI dv=58 km/s(FWHM)
1604: Pettini, M., Rix, S. A., Steidel, C. C., et al. 2002, \apj, 569, 742
1605:
1606: \bibitem[Prochaska et al.(2006)]{Pr06} % ISM 050730/051111; atomic,FS lines
1607: Prochaska, J. X., Chen, H.-W., \& Bloom, J. S. 2006, \apj, 648, 95
1608:
1609: \bibitem[Prochaska et al.(2007a)]{Pr07a} % ISM host 050730,050820,051111,060418
1610: Prochaska, J. X., Chen, H.-W., Bloom, J. S., et al. 2007a, \apjs, 168, 231
1611:
1612: \bibitem[Prochaska et al.(2007b)]{Pr07b} % ISM in GRB host II, gas,stars,dust
1613: Prochaska, J. X., Chen, H.-W., Dessauges-Zavadsky, M., \& Bloom,
1614: J. S. 2007b, \apj, 666, 267
1615:
1616: \bibitem[Prochaska et al.(2008a)]{Pr08a} % Vel Fields High-z Gal
1617: Prochaska, J. X., Chen, H.-W., Wolfe, A. M., Dessauges-Zavadsky, M.,
1618: \& Bloom, J. S. 2008a, ApJ, 672, 59
1619:
1620: \bibitem[Prochaska et al.(2008b)]{Pr08b} % N V in GRB hosts
1621: Prochaska, J. X., Dessauges-Zavadsky, M., Ramirez-Ruiz, E., \& Chen,
1622: H.-W. 2008b, ApJ, in press (astro-ph/0806.0399)
1623:
1624: \bibitem[Ralchenko et al.(2008)]{Ra08} %NIST
1625: Ralchenko, Y., Kramida, A. E., Reader, J., and NIST ASD Team 2008,
1626: NIST Atomic Spectra Database (version 3.1.5), National Institute
1627: of Standards and Technology, Gaithersburg, MD
1628:
1629: \bibitem[Ramirez-Ruiz et al.(2001)]{Ra01} % WR wind models
1630: Ramirez-Ruiz, E., Dray, L., Madau, P., \& Tout, C. A. 2001, \mnras,
1631: 327, 829
1632:
1633: \bibitem[Ramirez-Ruiz et al.(2005)]{Ra05} % WR wind models
1634: Ramirez-Ruiz, E. Garc\'ia-Segura, G., Salmonson, J. D., \&
1635: P\'erez-Rend\'on, B. 2005, \apj, 631, 435
1636:
1637: \bibitem[Ricker et al.(2003)]{Ri03} % HETE-2
1638: Ricker, G. R., et al. 2003, in AIP Conf. Proc. 662,
1639: Gamma-Ray Burst and Afterglow Astronomy 2001,
1640: ed. G. R. Ricker \& R. K. Vanderspek (New York: AIP), 3
1641:
1642: \bibitem[Samui et al.(2008)]{Sa08} % Sim Gal wind
1643: Samui, S., Subramanian, K., \& Srianand, R. 2008, \mnras, 385, 783
1644:
1645: \bibitem[Savage \& Sembach(1991)]{SS91} % AOD I
1646: Savage, B. D., \& Sembach, K. R. 1991, \apj, 379, 245
1647:
1648: \bibitem[Savage et al.(2003)]{Sa03} % OVI survey MW halo
1649: Savage, B. D., Sembach, K. R., Wakker, B. P., et al. 2003, \apjs, 146, 125
1650:
1651: \bibitem[Savage et al.(2005a)]{Sa05a} %OVI Wing Mrk 421
1652: Savage, B. D., Wakker, B. P., Fox, A. J., \& Sembach, K. R. 2005a,
1653: \apj, 619, 863
1654:
1655: \bibitem[Savage et al.(2005b)]{Sa05b} %SVI IGM (HE 0226 Ne VIII)
1656: Savage, B. D., Lehner, N., Wakker, B. P., Sembach, K. R., \& Tripp, T. M.
1657: 2005b, \apj, 626, 776
1658:
1659: \bibitem[Savaglio \& Fall(2004)]{Sv04} %Dust depletion CBM 020813 z=1.255 LRIS
1660: Savaglio, S., \& Fall, S. M. 2004, \apj, 614, 293
1661:
1662: \bibitem[Savaglio et al.(2002)]{Sv02} %021004 UVES
1663: Savaglio, S., Fiore, F., Israel, G., et al. 2002, GCN, 1633
1664:
1665: \bibitem[Savaglio et al.(2003)]{Sv03} %Abun+Dust hosts 990123,010222,000926
1666: Savaglio, S., Fall, S. M., \& Fiore, F. 2003, \apj, 585, 638
1667:
1668: \bibitem[Savaglio et al.(2008)]{Sv08} %GRB host gal properties
1669: Savaglio, S., Glazebrook, K., \& Le Borgne, D. 2008, \apj, submitted
1670: (astro-ph/0803.2718)
1671:
1672: \bibitem[Schaefer et al.(2003)]{Sc03} %021004 HET spectra ; HV comps-> shells
1673: Schaefer, B. E., Gerardy, C. L., H\"oflich, P., et al. 2003, \apj,
1674: 588, 387
1675:
1676: \bibitem[Sembach \& Savage(1992)]{SS92} % AOD
1677: Sembach, K. R., \& Savage, B. D. 1992, \apjs, 83, 147
1678:
1679: \bibitem[Sembach et al.(2001)]{Se01} % 3C 273 wing
1680: Sembach, K. R., Howk, J. C., Savage, B. D., Shull, J. M., \& Oegerle,
1681: W. R. 2001, \apj, 561, 573
1682:
1683: \bibitem[Sembach et al.(2003)]{Se03} % First survey of OVI wings
1684: Sembach, K. R., Wakker, B. P., Savage, B. D., et al. 2003, \apjs, 146, 165
1685:
1686: \bibitem[Shirasaki et al.(2002)]{Sh02} % GCN 021004
1687: Shirasaki, Y., Graziani, C., Matsuoka, M., et al. 2002, GCN 1565
1688:
1689: \bibitem[Shapley et al.(2003)]{Sh03} %Comp UV spec z~3 LBGs
1690: Shapley, A. E., Steidel, C. C., Pettini, M., \& Adelberger,
1691: K. L. 2003, \apj, 588, 65
1692:
1693: \bibitem[Simcoe et al.(2006)]{Si06} % Keck O VI/CIV/imag 1 SL (IGM hi-ion rat)
1694: Simcoe, R. A., Sargent, W. L. W., Rauch, M., \& Becker, G.
1695: 2006, \apj, 637, 648
1696:
1697: \bibitem[Starling et al.(2005)]{St05} % 050730 WHT/ISIS NHI=22.1 DLA
1698: Starling, R. L. C., Vreeswijk, P. M., Ellison, S. L., et al. 2005,
1699: \aap, 442, L21
1700:
1701: \bibitem[Stroh et al.(2007)]{St07} % Announcement of GRB 071031
1702: Stroh, M. C., Gronwall, C., Grupe, D., et al. 2007, GCN, 7020
1703:
1704: \bibitem[Th\"one et al.(2007)]{Th07} % UVES 030329 z=0.169 host gal prop
1705: Th\"one, C., Greiner, J., Savaglio, S., \& Jehin, E. 2007, \apj, 671, 628
1706:
1707: \bibitem[Th\"one et al.(2008a)]{Th08a} % Spat Resol Prop GRB 060505z=0.0889
1708: Th\"one, C., Fynbo, J. P. U., \"Ostlin, G., et al. 2008a, \apj, 676, 1151
1709:
1710: \bibitem[Th\"one et al.(2008b)]{Th08b} % WHT/ISIS 060206 (SB wind model)z=4.048
1711: Th\"one, C., Wiersema, K., Ledoux, C., et al. 2008b, \aap,
1712: in press (astro-ph/0708.3448)
1713:
1714: \bibitem[Tremonti et al.(2007)]{Tr07} % 1000 km/s winds z=0.6 postSB gal
1715: Tremonti, C. A., Moustakas, J., \& Diamond-Stanic, A. M. 2007, \apj, 663, L77
1716:
1717: \bibitem[van Marle et al.(2005)]{vM05} % WR wind models for GRBs
1718: van Marle, A. J., Langer, N., \& Garc\'ia-Segura, G. 2005, \aap, 444, 837
1719:
1720: \bibitem[van Marle et al.(2007)]{vM07} % WR wind models for GRBs
1721: van Marle, A. J., Langer, N., \& Garc\'ia-Segura, G. 2007, \aap, 469, 941
1722:
1723: \bibitem[van Marle et al.(2008)]{vM08} % WR wind models for GRBs
1724: van Marle, A. J., Langer, N., Yoon, S.-C., \& Garc\'ia-Segura,
1725: G. 2008, \aap, 478, 769
1726:
1727: \bibitem[van Paradijs et al.(1997)]{vP97} % GRBs are extragalactic
1728: van Paradijs, J., Groot, P. J., Galama, T., et al. 1997, Nature, 386, 686
1729:
1730: \bibitem[Vilkoviskij et al.(1999)]{Vi99} % Line-locking
1731: Vilkoviskij, E. Y., Efimov, S. N., Karpova, O. G., \& Pavlova,
1732: L. A. 1999, \mnras, 309, 80
1733:
1734: \bibitem[Vreeswijk et al.(2004)]{Vr04} % FORS2 RRM GRB 030323 host-galaxy DLA
1735: Vreeswijk, P. M., Ellison, S. L., Ledoux, C., et al. 2004, \aap, 419, 927
1736:
1737: \bibitem[Vreeswijk et al.(2007)]{Vr07} % UVES RRM GRB 060410 UV FeII* NiII*
1738: Vreeswijk, P. M., Ledoux, C., Smette, A., et al. 2007, \aap, 468, 83
1739:
1740: \bibitem[Vreeswijk et al.(2008)]{Vr08} % GCN 080310 UVES
1741: Vreeswijk, P. M., Jakobsson, P., Jaunsen, A. O., \& Ledoux C., 2008,
1742: GCN, 7391
1743:
1744: \bibitem[Wakker et al.(2003)]{Wa03} % OVI Survey
1745: Wakker, B. P., Savage, B. D., Sembach., K. R., et al. 2003, \apjs, 146, 1
1746:
1747: \bibitem[Watson et al.(2006)]{Wa06} % DLA-GRB 050401 z=2.899 N(HI)=22.6 FORS-2
1748: Watson, D., Fynbo, J. P. U., Ledoux, C., et al. 2006, \apj, 652, 1011
1749:
1750: \bibitem[Weiner et al.(2008)]{We08} % 1000 km/s outflows z=1.4 galaxies
1751: Weiner, B. J., Coil, A. L., Prochaska, J. X., et al. 2008, \apj, submitted
1752: (astro-ph/0804.4686)
1753:
1754: \bibitem[Wiersema et al.(2007)]{Wi07} % GRB host galaxies
1755: Wiersema, K., Savaglio, S., Vreeswijk, P. M., et al. 2007, \aap, 464, 529
1756:
1757: \bibitem[Wolfe \& Prochaska(2000)]{WP00} % Obsn of CIV in DLAs
1758: Wolfe. A. M., \& Prochaska, J. X. 2000, \apj, 545, 591
1759:
1760: \bibitem[Wolfe et al.(2005)Wolfe , Gawiser, \& Prochaska]{Wo05} % ARAA on DLAs
1761: Wolfe. A. M., Gawiser, E., \& Prochaska, J. X. 2005, \araa, 43, 861
1762:
1763: \bibitem[Woolsey \& Bloom(2006)]{WB06}
1764: Woolsey, S. E., \& Bloom, J. S. 2006, ARA\&A, 44, 507
1765:
1766: \bibitem[Ziaeepour et al.(2006)]{Zi06} % GCN 060607
1767: Ziaeepour, H. Z., Barthelmy, S. D., Gehrels, N., et al. 2006, GCN, 5233
1768:
1769: \bibitem[Zsarg\'o et al.(2003)]{Zs03} % Halo Star O VI Survey
1770: Zsarg\'o, J., Sembach, K. R., Howk, J. C., \& Savage, B. D. 2003,
1771: \apj, 586, 1019
1772:
1773: \end{thebibliography}
1774:
1775:
1776: \end{document}
1777:
1778: