0707.2594/ms.tex
1: %\documentclass[12pt,preprint]{aastex}
2: \documentclass{emulateapj}
3: \usepackage{natbib,amsmath}
4: 
5: \newcommand{\cmjj}{\mbox{${\rm cm^{-2}}$}}
6: \newcommand{\hI}{\mbox{${\rm H\,I}$}}
7: \newcommand{\nhi}{\mbox{$N({\rm H\,I})$}}
8: \newcommand{\lya}{\mbox{${\rm Ly}\alpha$}}
9: \newcommand{\lyb}{\mbox{${\rm Ly}\beta$}}
10: \newcommand{\apg}{\gtrsim}
11: \newcommand{\apll}{\lesssim}
12: \newcommand{\fesc}{\mbox{$f_{\rm esc}$}}
13: \newcommand{\tll}{\mbox{$\tau_{\rm LL}$}}
14: \newcommand{\etal}{\ensuremath{\mbox{et~al.}}}
15: 
16: \shorttitle{Escape Fraction in GRB Host Galaxies}
17: \shortauthors{Chen \etal}
18: 
19: \begin{document}
20: 
21: \slugcomment{Submitted to the Astrophysical Journal Letters}
22: 
23: \title{A New Constraint on the Escape Fraction in Distant Galaxies Using $\gamma$-ray Burst Afterglow Spectroscopy}
24: \author{Hsiao-Wen Chen\altaffilmark{1}, Jason X.\ Prochaska\altaffilmark{2,3}, and Nickolay Y.\ Gnedin\altaffilmark{4,1,5}}
25: %, and Enrico Ramirez-Ruiz\altaffilmark{3}} 
26: \altaffiltext{1}{Department of Astronomy \& Astrophysics, University of Chicago, Chicago, IL 60637, USA, {\tt hchen@oddjob.uchicago.edu}}
27: \altaffiltext{2}{University of California Observatories - Lick Observatory, University of California, Santa Cruz, CA 95064, USA; {\tt xavier@ucolick.edu}} 
28: \altaffiltext{3}{Department of Astronomy and Astrophyics, University of California, Santa Cruz, CA 95064, USA}
29: \altaffiltext{4}{Particle Astrophysics Center, Fermi National Accelerator Laboratory, Batavia, IL 60510, USA; {\tt gnedin@fnal.gov}}
30: \altaffiltext{5}{Kavli Institute for Cosmological Physics, University of Chicago, Chicago, IL 60637, USA}
31: \begin{abstract}
32: 
33: We describe a new method to measure the escape fraction \fesc\ of
34: ionizing radiation from distant star-forming galaxies using the
35: afterglow spectra of long-duration $\gamma$-ray bursts (GRBs).
36: Optical spectra of GRB afterglows allow us to evaluate the optical
37: depth of the host ISM, according to the neutral hydrogen column density 
38: \nhi\ observed along the sightlines toward the star-forming regions where 
39: the GRBs are found.  Different from previous effort in searching for 
40: faint, transmitted Lyman continuum photons, our method is not subject to 
41: background subtraction uncertainties and does not require prior knowledge 
42: of either the spectral shape of the host galaxy population or the IGM 
43: \lya\ forest absorption along these GRB sightlines.  Because most GRBs 
44: occur in sub-$L_*$ galaxies, our study also offers the first constraint 
45: on \fesc\ for distant low-mass galaxies that dominate the cosmic 
46: luminosity density.  We have compiled a sample of 28 GRBs at redshift 
47: $z\apg 2$ for which the underlying \nhi\ in the host ISM are known.  
48: These GRBs together offer a statistical sampling of the integrated optical 
49: depth to ionizing photons along random sightlines from star-forming 
50: regions in the host galaxies, and allow us to estimate the mean escape 
51: fraction $\langle\fesc\rangle$ averaged over different viewing angles.  
52: We find $\langle\fesc\rangle=0.02\pm 0.02$ and place a 95\% c.l.  upper 
53: limit $\langle\fesc\rangle \le 0.075$ for these hosts.  We discuss 
54: possible biases of our approach and implications of the result.  
55: Finally, we propose to extend this technique for measuring 
56: $\langle\fesc\rangle$ at $z\sim 0.2$ using spectra of core-collapse 
57: supernovae.
58: 
59: \end{abstract} 
60: 
61: \keywords{cosmology:observations---gamma-rays:bursts---galaxies:high-redshift---galaxies:ISM}
62: 
63: \section{Introduction}
64: 
65: Observations of distant QSOs indicate that the intergalactic medium
66: (IGM) became fully ionized by redshift $z\sim 6$ (e.g.\ Fan, Carilli,
67: \& Keating 2006).  While at $z<3$ QSOs are the dominant sources of the
68: ultraviolet background radiation (e.g.\ Haardt \& Madau 1996), at
69: higher redshifts where the number density of QSOs declines steeply
70: toward earlier epochs (e.g.\ Willott \etal\ 2005; %Mahabal \etal\
71: 2005; Richards \etal\ 2006) additional ionizing sources are necessary.
72: The spectral shape of the ultraviolet background radiation inferred
73: from intervening metal absorption line studies (e.g.\ Haehnelt \etal\
74: 2001) and intergalactic He\,II absorption spectra (e.g.\ Shull \etal\
75: 2004; Reimers \etal\ 2005) suggest that young stars may provide the
76: dominant ionizing sources during early epochs.
77: 
78: The escape fraction of ionizing radiation, \fesc, specifies the
79: fraction of stellar-origin ionizing photons ($h\nu > 1$ Ryd) that
80: escape the interstellar medium (ISM) of star-forming galaxies.
81: Accurate measurements of \fesc\ are important for quantifying the
82: relative contribution of ionizing photons to the ultraviolet
83: background radiation between galaxies and AGN.  For the Milky Way,
84: estimates based on diffuse H$\alpha$ emission of High Velocity Clouds
85: (HVC) yield an upper limit of $\fesc \apll 6$\% (e.g.\ Bland-Hawthorn
86: \& Maloney 1999 and see Weiner \etal\ 2001 for a review).  In the
87: nearby universe, early observations of starburst galaxies place
88: constraints at $\fesc\apll 6$\% (e.g.\ Heckman \etal\ 2001).  Bergvall
89: \etal\ (2006) have reported the first positive detection of Lyman
90: continuum photons in the spectra of the starburst galaxy Haro 11 using
91: the {\it Far-UV Space Explorer}.  Their analysis suggests an escape
92: fraction in this galaxy of $\fesc=1-10$\%.  This finding has, however,
93: been challenged by Grimes \etal\ (2007), who cannot confirm the
94: detection of Lyman continuum in the same data set.
95: 
96: At higher redshift, a range of values are reported from $\fesc<6$\% at
97: $z=1.1-1.4$ (Malkan \etal\ 2003; Siana \etal\ 2007), to between mean
98: values of no more than 8\% (Giallongo \etal\ 2002; Fern\'andez-Soto
99: \etal\ 2003; Shapley \etal\ 2006) and $\fesc\approx 13-38$\% (Inoue
100: \etal\ 2005) at $z\sim 3$, to $\fesc\approx 22$\% at $\langle
101: z\rangle=3.4$ (Steidel \etal\ 2001)\footnote{Note that many of the
102: previous publications reported measurements for $f_{\rm esc,rel}$,
103: which is defined as the ratio of escaped ionizing photons at 912 \AA\
104: to the observed flux density at 1500 \AA.  Here we have converted
105: these relative measurements to \fesc\ at 1 Ryd based on their
106: respective $\langle E(B-V)\rangle$ reported by the authors.}.  In
107: addition, direct detections of Lyman continuum photons are reported
108: for two $z\sim 3$ galaxies by Shapley \etal\ (2006), implying
109: $\fesc\approx 13-20$\% for these two sources.  The large scatter in
110: the reported $\fesc$ may imply a large variation in the optical depth
111: across different lines of sight toward the inner regions of the
112: galaxies (e.g.\ Gnedin, Kravtsov, \& Chen 2007), but it also
113: underscores the challenges in detecting low-luminosity features in a
114: background noise limited regime.
115: 
116: In this {\it Letter}, we describe a novel approach for constraining
117: \fesc\ from high-redshift star-forming galaxies.  We estimate \fesc\
118: based on the distribution of neutral hydrogen column density \nhi\
119: observed in the afterglow spectra of long-duration $\gamma$-ray bursts
120: (GRBs).  Long-duration GRBs are believed to originate in the death of
121: massive stars with $M > 20$ M$_\odot$ (see Woosley \& Bloom 2006 for a
122: recent review), and are signposts of active star-forming regions in
123: the ISM of their host galaxies (e.g.\ Bloom \etal\ 2002; Fruchter
124: \etal\ 2006).  Spectroscopic observations of the bright optical
125: afterglows following the initial bursts have allowed us to measure the
126: gas and dust content along the sightlines toward the GRBs, based on
127: absorption features imprinted in the afterglow spectra (e.g.\
128: Vreeswijk \etal\ 2004; Chen \etal\ 2005; Prochaska \etal\ 2007a).
129: These GRB sightlines together offer a statistical sampling of the
130: integrated optical depth to ionizing photons along random directions
131: in the host galaxies.
132: 
133: The escape fraction \fesc\ determined from the \nhi\ distribution
134: along random sightlines in the ISM of GRB host galaxies does not
135: require direct detection of Lyman continuum photons.  It is not
136: subject to systematic uncertainties due to background subtraction.
137: The \fesc\ value is derived based on the total gas column observed in
138: front of the star-forming region that hosts the GRB.  It does not
139: depend on the spectral shape of the ultraviolet radiation from the
140: host galaxy or the stochastic uncertainties in the IGM \lya\ forest
141: absorption along the lines of sight.  Finally, while some GRB host
142: galaxies are reported to have high star formation rate, SFR $\apg 100$
143: M$_\odot$, (c.f.\ Berger \etal\ 2003; Le Floc'h \etal\ 2006), growing
144: evidence indicates that the majority are sub-$L_*$ galaxies (e.g.\ Le
145: Floc'h \etal\ 2003; Sollerman \etal\ 2005; Fruchter \etal\ 2006).  Our
146: study therefore offers the first constraint on \fesc\ for low-mass
147: galaxies at $z>2$.
148: 
149: %In the following sections, we describe the basic methodology and apply
150: %the technique to the current sample of GRBs.  We place a direct
151: %constraint on \fesc\ for $z>2$ GRB host galaxies.  Finally, we discuss
152: %the implications and possible observational biases of our result.
153: 
154: \section{The Sample of GRBs at $z\apg 2$}
155: 
156: To obtain an accurate estimate of the mean escape fraction of ionizing
157: photons along GRB sightlines, we first compile a sample of GRBs that
158: are confirmed at $z\apg 2$.  We focus our analysis on bursts at
159: $z_{\rm GRB} \apg 2$ for two main reasons.  First, it minimizes the
160: ambiguity between an optically thin sightline and a low-redshift
161: interloper.  The $z>2$ IGM will imprint its signature on an optical
162: spectrum via the \lya\ forest and metal-line absorption features.
163: Therefore, an afterlow that exhibits a featurless spectrum is most
164: likely at $z_{\rm GRB} < 2$.  Second, measurements of \nhi\ from the
165: absorption profiles of \lya\ and Lyman series allow us to directly
166: evaluate the optical depth at the Lyman limit frequency \tll.
167: 
168: Table 1 lists 40 spectroscopically confirmed GRBs at $z_{\rm GRB}\apg
169: 2$.  In addition to the redshift of each source, we also list the
170: isotropic equivalent energy release in $\gamma$-ray photons ($E_{\rm
171: iso}$), the observed \nhi\ if available, a flag f$_{\alpha}$ to
172: indicate whether the afterglow spectrum covers the redshifted \lya\
173: transition ('0' means no coverage and '1' means \lya\ coverage), and a
174: flag f$_{i}$ to indicate whether metal-line features due to low ions
175: such Si\,II or C\,II are present.  Four of the GRBs do not have
176: spectral coverage of the \lya\ transition from the host, but the
177: presence of low ions (f$_{i}$) indicate that the gas is consistent
178: with being optically thick.  To exhibit strong low-ion absorption, an
179: optically thin gas would need to have very high density (to maintain a
180: non-negligible neutral fraction) and super-solar metallicity, both of
181: which are very unlikely.  Nine sources do not have published \nhi.  In
182: the subsequent analysis, we consider only those 28 sightlines with
183: published \nhi\ values as our main sample, and assume that the
184: remaining 13 sources with no available \nhi\ measurements share the
185: same distribution as the main sample.  This is justified based on the
186: similar $z_{\rm GRB}$ and $E_{\rm iso}$ distributions between GRB
187: sightlines with and without known \nhi\ measurements.
188: 
189: Figure 1 presents the cumulative \nhi\ distribution, ${\cal F}[<\nhi]$
190: from the main sample of 28 GRB host galaxies, together with the
191: 1-$\sigma$ uncertainties determined based on a bootstrap re-sampling
192: method.  Specifically, we establish a simulated sample of 28
193: sightlines from random sampling of the main sample, allowing
194: duplications of individual sightlines.  Then, we evaluate the
195: cumulative \nhi\ distribution of the simulated sample.  We repeat the
196: procedure 10,000 times to determine the 68\% scatter of ${\cal F}$
197: around the mean value in each \nhi\ bin.
198: 
199: \begin{figure}
200: \begin{center}
201: \includegraphics[scale=0.35, angle=270]{f1.ps}
202: \caption{Cumulative distribution of neutral hydrogen column density
203: ${\cal F}[<\nhi]$ observed in the host galaxies of long-duration GRBs
204: at $z\ge 2$ (solid histogram).  The shaded area shows the 1-$\sigma$
205: uncertainties evaluated using a bootstrap re-sampling method that
206: accounts for both \nhi\ measurement uncertainties and sampling errors.
207: The dash-dotted line represents the best-fit power-law model described
208: in \S\ 3.  The dotted curve respresents the predicted distribution
209: from Gnedin, Kravtsov, \& Chen (2007).}
210: \end{center}
211: \end{figure}
212: 
213: \section{The Method}
214: 
215: The optical depth of Lyman limit photons along individual lines of
216: sight is determined according to $\tll=\sigma_{\rm LL}\times \nhi$,
217: where $\sigma_{\rm LL}=6.28\times 10^{-18}$ cm$^2$ is the
218: photo-ionization cross section of hydrogen atoms.  In principle, {\it
219: considering a sample of random sightlines from the star-forming
220: regions in a galaxy together yields an estimate of the optical depth
221: averaged over all viewing angles}.  In practice, we consider an
222: ensemble of random sightlines toward GRBs in distant star-forming
223: galaxies.  The mean escape fraction of Lyman limit photons 
224: %for the GRB host galaxy population 
225: averaged over all directions is evaluated according to
226: \begin{equation}
227: \langle\fesc\rangle=\frac{1}{n}\sum_{i=1}^{i=n}\exp[-\sigma_{\rm LL}\,N_i({\rm H\,I})],
228: \end{equation}
229: where the sum extends over the total number of $n$ GRB sightlines in
230: the sample.  For our main sample presented in Figure 1, we find $n=28$
231: and Equation (1) yields $\langle\fesc\rangle=0.02\pm 0.02$.  The error
232: is estimated using the bootstrap re-sampling method described in \S\ 2
233: and represents the 68\% uncertainty in the mean value.  We also
234: determine a 95\% c.l. upper limit $\langle\fesc\rangle \le 0.075$.
235: 
236: Parameterizing the cumulative \nhi\ distribution by ${\cal F}[<\nhi] =
237: A\,[\nhi/N_0]^\alpha$, we find based on a $\chi^2$ analysis $\log\,A =
238: -0.58 \pm 0.05$ and $\alpha = 0.32\pm 0.03$ for $\log\,N_0 = 20.5$
239: over $\log\,\nhi=16.5-21.5$ (dash-dotted line in Figure 1).  Equation
240: (1) is expressed as
241: \begin{eqnarray*}
242: \langle\fesc\rangle&=&\int_0^\infty d\,N_{\rm H\,I}\frac{d\,{\cal
243: F}}{d\,N_{\rm H\,I}}\exp[-\sigma_{\rm LL}\,N({\rm H\,I})] \\
244: &=&\frac{A\,\alpha}{(\sigma_{\rm LL} N_0)^\alpha} \Gamma(\alpha).
245: \end{eqnarray*}
246: We determine $\langle\fesc\rangle = 0.020\pm 0.003$, where the errors
247: represent the 68\% uncertainties.  
248: 
249: \section{Discussion}
250: 
251: %Using the \nhi\ distribution established for the host ISM of
252: %long-duration GRBs at $z\apg 2$, we report a 95\% c.l.\ upper limit
253: %for the mean escape fraction of Lyman limit photons,
254: %$\langle\fesc\rangle \apll 0.075$ in distant star-forming galaxies.
255: We have applied the \nhi\ measured from the \lya\ (and in some cases
256: Lyman series as well) absorption strength in early-time afterglow
257: spectra to constrain the mean escape fraction in distant star-forming
258: galaxies.  Different from conventional methods to search for
259: transmitted Lyman continuum photons, our estimated
260: $\langle\fesc\rangle$ is not subject to background subtraction
261: uncertainties and does not depend on the intrinsic ultraviolet
262: spectral shape of the host galaxies or dust distribution in the host
263: ISM.  In addition, it does not depend on the stochastic IGM \lya\
264: absorption along the sightlines toward these GRBs.  
265: %Our result
266: %therefore offers an independent constraint on the escape fraction of
267: %Lyman limit photons from star-forming galaxies at $z=2-6$.
268: Here we discuss possible biases in our estimated $\langle\fesc\rangle$
269: due to the selection of GRB sightlines and implications of our result.
270: 
271: \subsection{Observational Biases}
272: 
273: Our analysis considers only galaxies that host a GRB event.  The
274: presence of a GRB indicates that the ISM immediately surrounding the
275: burst is photo-ionized by the bright afterglow.  The intensity of the
276: afterglow ionizing radiation is capable of photo-ionizing all \hI\ gas
277: to a distance of $r = 10$ to 30\,pc (Drain \& Hao 2002; Watson \etal\
278: 2007; Prochaska \etal\ 2007b) that can exceed the typical Stromgren
279: radius of an H\,II region.  It is conceivable that the presence of a
280: GRB may reduce \tll\ along the sightline and the constraint on
281: $\langle\fesc\rangle$ can be considered as an upper limit.
282: 
283: In addition, our constraint is derived explicitly at the 912-\AA\
284: Lyman limit transition.%, based on optical depth due to hydrogen
285: atoms.  Additional absorption due to dust would further reduce the
286: estimated $\langle\fesc\rangle$, but is almost negligible at
287: frequencies beyond 1 Ryd (see Gnedin \etal\ 2007).  However, some
288: fraction of the GRB afterglows are optically ``dark'' and missed in
289: our sample because afterglow spectroscopy was impossible.  While the
290: majority of these are presumably associated with high-$z$ events
291: (where the IGM absorbs most of the optical photons) or intrinsically
292: faint afterglows, the remainder are associated with highly
293: dust-extinguished sightlines (e.g.\ Pellizza \etal\ 2006; Rol \etal\
294: 2007).  These extremely dusty sightlines are optically thick to
295: ionizing radiation further strengthening the $\langle\fesc\rangle$.
296: 
297: Intrinsically faint UV afterglows could occur in a particularly
298: low-density environment (Kumar \& Panaitescu 2000) and, in turn,
299: preferentially probe optically thin sightlines.  We note, however,
300: that there is no notable correlation between the observed \nhi\ and
301: the afterglow UV luminosity.  For example, GRB\,021004 and GRB\,050820
302: have very different \nhi\ but comparable afterglow UV luminosity
303: (e.g.\ Lazzati \etal\ 2006; Vestrand \etal\ 2006).  Furthermore, the
304: isotropic equivalent energies of the GRB sample ($E_{\rm iso}$ in
305: Table 1) display a wide dispersion at all redshift, supporting that
306: these GRBs do not preferentially originate in low-density environment
307: (e.g.\ Piran 1999).  We argue that the pre-selection of bright UV
308: afterglows has not restricted the analysis to a special sub-sample of
309: GRBs.
310: 
311: \subsection{Implications and Future Work}
312: 
313: %These studies indicate an average luminosity
314: %of sub-L$_*$ at all redshifts surveyed. 
315: The association between GRBs and massive stars that typically have a
316: short lifetime indicates that the GRB events occur close to the
317: locations where their progenitor stars were formed and directly trace
318: the current star formation rate.  Late-time imaging surveys to search
319: for the host galaxies of long-duration GRBs have shown that these GRBs
320: occur in sub-$L_*$, galaxies (Le Floc'h \etal\ 2003; Sollerman \etal\
321: 2005; Fruchter \etal\ 2006) that exhibit on average large specific
322: star-formation rates, SFR/$L_B$, (Christensen \etal\ 2004).  This can
323: be understood as low-mass galaxies undergoing early generations of
324: star formation (Erb \etal\ 2006).  The derivation of \fesc\ from GRB
325: sightlines reveals the escape fraction during the lifetime of the most
326: intense ionizing sources, and applies to `normal', star-forming
327: galaxies that dominates the cosmic UV luminosity density.  The
328: low-mass nature also allows a direct comparison with predictions from
329: high-resolution cosmological simulations (the dotted curve in Figure
330: 1; Gnedin \etal\ 2007).
331: 
332: Our 95\% c.l. upper limit of $\langle\fesc\rangle$ is comparable to
333: the low values previously reported based on observations of Lyman
334: continuum photons from more luminous star-forming galaxies at $z\sim
335: 3$ (Giallongo \etal\ 2002; Fern\'andez-Soto \etal\ 2003; Shapley
336: \etal\ 2006).  Additional absorption due to dust would further reduce
337: the estimated $\langle\fesc\rangle$, but is almost negligible at
338: frequencies beyond 1 Ryd (see Gnedin \etal\ 2007).  We estimate the
339: total contribution of ionizing photons from sub-$L*$ galaxies,
340: adopting the UV luminosity function determined for luminous
341: star-forming galaxies at $z\sim 3$ from Adelberger \& Steidel (2000).
342: We derive a co-moving luminosity density at 1500 \AA\ of $2.2\times
343: 10^{26}\, h$ erg s$^{-1}$ Hz$^{-1}$ Mpc$^{-3}$ for sub-$L_*$
344: ($0.1-1\,L_*$) galaxies\footnote{We adopt a $\Lambda$ cosmology,
345: $\Omega_{\rm M}=0.3$ and $\Omega_\Lambda = 0.7$, with a dimensionless
346: Hubble constant $h = H_0/(100 \ {\rm km} \ {\rm s}^{-1}\ {\rm
347: Mpc}^{-1})$.}.  Applying an extinction correction to the observed
348: 1500-\AA\ flux (the authors estimated $f_{\rm esc}(1500 \AA)\approx
349: 0.2$) and assuming an intrinsic flux ratio between rest-frame 1500
350: \AA\ and 900 \AA\ of $f(1500)/f(900)=3$ adopted by Steidel \etal\ 2001
351: (but see Siana \etal\ 2007), we estimate a co-moving emissivity at 1
352: Ryd from sub-$L_*$ galaxies of $<2.8\times 10^{25}\, h$ erg s$^{-1}$
353: Hz$^{-1}$ Mpc$^{-3}$ for the 95\% c.l. upper limit
354: $\langle\fesc\rangle \apll 0.075$.  To estimate the QSO contribution
355: to the ultraviolet background radiation, we adopt 
356: %calculate the co-moving
357: %luminosity density at rest-frame 912\AA\ from QSOs using 
358: the $z=3$ QSO luminosity function estimated by Hopkins \etal\ (2007).
359: We find that the contribution to the ionizing background from QSOs of
360: bolometric luminosity $L_{\rm bol} > 10^8\,L_\odot$ is $\approx
361: 5\times 10^{24}\, h$ erg s$^{-1}$ Hz$^{-1}$ Mpc$^{-3}$.  While the
362: uncertainties in these various numbers are large, this exercise shows
363: that QSOs and sub-$L*$ galaxies with $\langle\fesc\rangle=1-2$ \% can
364: contribute a comparable amount of ionizing photons to the ultraviolet
365: background radiation at $z\sim 3$.
366: 
367: A larger sample of GRB sightlines with known \nhi\ measurements for
368: the host ISM is needed for improving the uncertainties in
369: $\langle\fesc\rangle$.  In addition, follow-up imaging surveys for
370: unvailing the emission properties of the GRB host galaxies are
371: valuable for testing model predictions of star formation at high
372: redshift.  Specifically, cosmological simulations show that low-mass
373: galaxies are inefficient in emitting ionizing radiation (Gnedin \etal\
374: 2007).  The luminosity distribution of these GRB hosts offers a direct
375: test of these models.
376: 
377: Finally, it is worth exploring whether a similar analysis could be
378: performed with core-collapse supernovae (SN).  These events also trace
379: the death of massive stars over a broader mass range and provide a
380: bright probe of the optical depth through the host galaxy.
381: %In addition, core-collapse SN trace a
382: %broader mass range of massive stars that produce substantial ionizing
383: %radiation.  
384: While the extreme line-blanketing in the far-ultraviolet of SN spectra
385: may preclude the study of Hydrogen absorption, it may be plausible to
386: pursue an analysis of ISM metal-absorption, e.g.\ via the Mg\,II
387: doublet.  Indeed, the {\it HST} spectrum of the Type\,II SN 1999em
388: shows narrow absorption lines of Fe\,II, Mg\,II, and Mg\,I (Baron
389: \etal\ 2000).  These features imply that NGC\,1637 is optically thick
390: to ionizing radiation along this particular sightline.  A survey of $z
391: \gtrsim 0.2$ SN could be carried out with a modest resolution,
392: blue-sensitive spectrometer on a 10~m-class telescope.
393: 
394: 
395: \acknowledgements
396: 
397: The authors acknowledge helpful discussion with J.\ O'Meara, E.\
398: Ramirez-Ruiz and M.\ Dessauges-Zavadsky.  We thank A.\ Kann and J.\
399: Bland-Hawthorn for valuable input.  H.-W.C. acknowledges support from
400: NASA grant NNG\,06GC36G and an NSF grant AST-0607510.
401: J. X. P. acknowledges support from NASA/Swift grant NNG\,05GF55G and a
402: CAREER grant (AST-0548180).
403: 
404: 
405: \begin{references}
406: 
407: \vskip 0.2in
408: 
409: \reference{} Adelberger, K. L. \& Steidel, C. C. 2000, ApJ, 544, 218.
410: \reference{} Baron, E. et al.\ 2000, ApJ, 545, 444.
411: \reference{} Berger, E. \etal\ 2003, ApJ, 588, 99
412: \reference{} Berger, E. 2006, GRB Coordinates Network, 5962, 1.
413: \reference{} Berger, E. \& Gladders, M. 2006, GRB Coordinates Network, 5170, 1. 
414: \reference{} Berger, E. et al.\ 2006a, GRB Coordinates Network, 4815, 1. 
415: \reference{} Berger, E. et al.\ 2006b, ApJ, 642, 979. 
416: \reference{} Berger, E., Fox, D. B., \& Cucchiara, A. 2007, GRB Coordinates Network, 6470, 1.
417: \reference{} Bergvall, N. et al.\ 2006, A\&A, 448, 513. 
418: \reference{} Bland-Hawthorn, J., \& Maloney, P. R. 1999, ApJ, 510, L33
419: \reference{} Bloom, J. S., Kulkarni, S. R., \& Djorgovski, S. G. 2002, AJ, 123, 1111. 
420: \reference{} Bloom, J. S., Frail, D. A., \& Kulkarni, S. R. 2003, ApJ, 594, 674.
421: %\reference{} Bulter, N. R. \& Kocevski, D. 2007, ApJ, 663, 407
422: \reference{} Castro, S. M. et al.\ 2000, GCN Report 605. 
423: \reference{} Castro, S. et al.\ 2003, ApJ, 586, 128. 
424: \reference{} Cenko, S. B. et al.\ 2006, GRB Coordinates Network, 5155, 1. 
425: \reference{} Chen, H.-W. et al.\ 2005, ApJ, 634, L25. 
426: \reference{} Christensen,L.,Hjorth,J., \& Gorosabel, J.2004, A\&A,425,913. 
427: %\reference{} Cucchiara, A., Fox, D. B., \& Berger, E. 2006, GRB Coordinates Network, 4729, 1. 
428: \reference{} Cucchiara, A. \etal\ 2006, High-Resolution Spectroscopy of the Afterglow of GRB 060210, in Swift and GRBs: Unveiling the Relativistic Universe, poster 
429: \reference{} D'Elia, V. et al.\ 2006, GRB Coordinates Network, 5637, 1. 
430: \reference{} Draine, B. T. \& Hao, L. 2002, ApJ, 569, 780. 
431: \reference{} Erb, D. K. \etal\ 2006, ApJ, 647, 128.
432: \reference{} Fan, X., Carilli, C. L., \& Keating, B. 2006, ARA\&A, 44, 415
433: \reference{} Fern\'andez-Soto, A., Lanzetta, K. M., \& Chen, H.-W. 2003, MNRAS, 342, 1215. 
434: \reference{} Fruchter, A. S. et al.\ 2006, Nature, 441, 463. 
435: \reference{} Fynbo, J. P. U. et al.\ 2006a, GRB Coordinates Network, 5809, 1. 
436: \reference{} Fynbo, J. P. U. et al.\ 2006b, A\&A, 451, L47. 
437: \reference{} Giallongo, E. \etal\ 2002, ApJ, 568, L9
438: \reference{} Gnedin, N. Y., Kravtsov, A. V., \& Chen, H.-W. 2007, ApJ submitted (arXiv:0707.0879).
439: \reference{} Grimes, J. P. \etal\ 2007, ApJ in press (arXiv:0707.0693).
440: \reference{} Haardt, F. \& Madau, P. 1996, ApJ. 461, 20.
441: \reference{} Haehnelt, M. G. \etal\ 2001, ApJ, 549, L151.
442: \reference{} Heckman, T. M. \etal\ 2001, ApJ, 558, 56.
443: %\reference{} Hjorth, J. et al.\ 2003a, Nature, 423, 847. 
444: \reference{} Hjorth, J. et al.\ 2003, ApJ, 597, 699. 
445: \reference{} Hopkins, P. F., Richards, G. T., \& Hernquist, L. 2007, ApJ, 654, 731
446: \reference{} Inoue, A. K. et al.\ 2005, A\&A, 435, 471. 
447: \reference{} Jakobsson, P. et al.\ 2004, A\&A, 427, 785. 
448: \reference{} Jakobsson, P. et al.\ 2006, A\&A, 460, L13. 
449: \reference{} Jakobsson, P. et al.\ 2007, GRB Coordinates Network, 6283, 1. 
450: \reference{} Jaunsen, A. O. et al.\ 2007, GRB Coordinates Network, 6010, 1. 
451: \reference{} Kawai, N. et al.\ 2006, Nature, 440, 184. 
452: \reference{} Kumar, P., \& Panaitescu, A. 2000, ApJ, 541, L51
453: \reference{} Lazzati, D. \etal\ 2006, MNRAS, 372, 1791
454: \reference{} Le Floc'h, E. et al.\ 2003, A\&A, 400, 499.
455: \reference{} Le Floc'h, E., Charmandaris, V., \& Forrest, W. J. 2006, ApJ, 642, 636
456: %\reference{} Mahabal, A. \etal\ 2005, ApJ, 634, L9.
457: \reference{} Malkan, M., Webb, W., \& Konopacky, Q. 2003, ApJ, 598, 878.
458: %\reference{} Mirabal, N. et al.\ 2006, ApJ, 643, L99.
459: %\reference{} Modjaz, M. \etal\ 2007, ApJ submitted (astro-ph/0701246)
460: \reference{} Pellizza, L. J. et al.\ 2006, A\&A, 459, L5.
461: \reference{} Piran, T. 1999, Phys. Rep., 314, 575
462: \reference{} Piranomonte, S. et al.\ 2006, GRB Coordinates Network, 4520, 1.
463: \reference{} Piranomonte, S. et al.\ 2007, A\&A in press (arXiv:0704.1729).
464: \reference{} Price, P. A. et al.\ 2007, ApJ, 663, L57
465: %\reference{} Prochaska, J. et al.\ 2007a, In prep.
466: \reference{} Prochaska, J. X. et al.\ 2007a, ApJS, 168, 231.
467: \reference{} Prochaska, J. X. et al.\ 2007b, ArXiv Astrophysics e-prints.
468: %\reference{} Prochaska, J. X., Herbert-Fort, S., \& Wolfe, A. M. 2005, ApJ, 635, 123. 
469: %\reference{} Prochter, G. et al. 2007, In prep. 
470: %\reference{} Prochter, G. E. et al.\ 2006, ApJ, 648, L93. 
471: \reference{} Quimby, R. et al.\ 2005, GRB Coordinates Network, 4221, 1. 
472: \reference{} Reimers, D., Fechner, C., Hagen, H.-J., et al.\ 2005, A\&A, 442, 63
473: \reference{} Richards, G. T. \etal\ 2006, AJ, 131, 49.
474: \reference{} Rol, E. et al.\ 2006, GRB Coordinates Network, 5555, 1. 
475: \reference{} Rol, E. et al.\ 2007, ApJ submitted (arXiv:0706.1518)
476: \reference{} Ruiz-Velasco, A. E. et al.\ 2007, ApJ in press (arXiv:0706.1257)
477: \reference{} Sakamoto, T. \etal\ 2005, ApJ, 629, 311.
478: \reference{} Savaglio, S. et al.\ 2007, GRB Coordinates Network, 6166, 1. 
479: \reference{} Shapley, A. E. et al.\ 2006, ApJ, 651, 688. 
480: \reference{} Shin, M.-S. \etal\ 2007, ApJ submitted (astro-ph/0608327).
481: \reference{} Shull, J. M. \etal\ 2004, ApJ, 600, 570.
482: \reference{} Siana, B. et al.\ 2007, arXiv:0706.4093. 
483: \reference{} Sollerman, J. et al.\ 2005, New Astronomy, 11, 103. 
484: \reference{} Steidel, C. C., Pettini, M., \& Adelberger, K. L. 2001, ApJ, 546, 665. 
485: \reference{} Thoene, C. C. et al.\ 2007a, GRB Coordinates Network, 6379, 1. 
486: \reference{} Thoene, C. C. et al.\ 2007b, GRB Coordinates Network, 6499, 1. 
487: \reference{} Vestrand, W.~T. \etal\ 2006, Nature, 442, 172
488: \reference{} Vreeswijk, P. M. et al.\ 2004, A\&A, 419, 927. 
489: \reference{} Vreeswijk, P. M. et al.\ 2006, A\&A, 447, 145. 
490: \reference{} Watson, D. et al.\ 2006, ApJ, 652, 1011. 
491: \reference{} Watson, D. \etal\ 2007, ApJ, 660, L101.
492: \reference{} Weiner, B.~J., Vogel, S.~N., \& Williams, T.~B. 2002, in ASP Conf. Ser. 254, Extragalactic Gas at Low Redshift, ed. J. S. Mulchaey \& J. Stocke (San Francisco: ASP), 256
493: \reference{} Willott, C.J. \etal\ 2005, ApJ, 626, 657.
494: \reference{} Woosley, S.~E. \& Bloom, J.~S. 2006, ARA\&A, 44, 507.
495: \end{references}
496: 
497: \clearpage
498: 
499: \begin{deluxetable}{lccrccr}
500: \tablecaption{The Sample of GRBs at $z\ge 2$}
501: \tablewidth{0pt}
502: %\tablecolumns{6}
503: \tablehead{\colhead{GRB} & \colhead{$z_{\rm GRB}$} & 
504: \colhead{$\log\,E_{\rm iso}$\tablenotemark{a}} & \colhead{$\log\,N(\hI)$} & 
505: \colhead{f$_\alpha$\tablenotemark{b}} & \colhead{f$_i$\tablenotemark{c}} & 
506: \colhead{Ref.}}
507: \startdata
508: 000301c & 2.03 & 52.64 & $21.2\pm 0.5$   & 1 & 7 & 1 \nl
509: 000926  & 2.04 & 53.76 & $21.30\pm 0.25$ & 1 & 7 & 2 \nl
510: 011211  & 2.14 & 53.17 & $20.4\pm 0.2$   & 1 & 7 & 3 \nl
511: 020124  & 3.20 & 52.81 & $21.7\pm 0.4$   & 1 & 7 & 33 \nl 
512: 021004  & 2.33 & 52.82 & $19.5\pm 0.5$   & 1 & 3 & 4 \nl
513: 030226  & 1.99 & 53.20 & $20.5\pm 0.3$   & 1 & 7 & 34 \nl
514: 030323  & 3.37 & 52.99 & $21.90\pm 0.07$ & 1 & 7 & 5 \nl
515: 030429  & 2.65 & 52.87 & $21.6\pm 0.2$   & 1 & 7 & 6 \nl
516: 050319  & 3.24 & 52.81 & $20.9\pm 0.2$   & 1 & 7 & 22 \nl
517: 050401  & 2.90 & 53.94 & $22.6\pm 0.3$   & 1 & 7 & 7 \nl
518: 050505  & 4.27 & 53.81 & $22.05\pm 0.10$ & 1 & 7 & 8 \nl
519: 050730  & 3.97 & 53.76 & $22.15\pm 0.10$ & 1 & 7 & 9 \nl
520: 050820  & 2.61 & 53.61 & $21.1\pm 0.1$   & 1 & 7 & 10 \nl
521: 050904  & 6.29 & 54.32 & $21.3\pm 0.2$   & 1 & 7 & 11 \nl
522: 050908  & 3.34 & 52.65 & $19.1\pm 0.1$   & 1 & 3 & 4 \nl
523: 050922c & 2.19 & 52.73 & $21.55\pm 0.10$ & 1 & 7 & 12 \nl
524: 051109  & 2.34 & 52.89 & $...$           & 0 & 1 & 13 \nl
525: 060115  & 3.53 & 53.28 & $...$           & 1 & 7 & 14 \nl
526: 060124  & 2.30 & 53.39 & $19.3\pm 0.2$   & 1 & 3 & 4 \nl
527: 060206  & 2.26 & 52.45 & $20.85\pm 0.1$  & 1 & 7 & 15 \nl
528: 060210  & 3.91 & 53.99 & $21.7\pm 0.2$   & 1 & 7 & 16 \nl
529: 060223  & 4.41 & 53.06 & $...$           & 1 & 7 & 17 \nl
530: 060510b & 4.94 & 53.96 & $21.1\pm 0.1$   & 1 & 7 & 18 \nl
531: 060522  & 5.11 & 53.42 & $20.5\pm 0.5$   & 1 & 7 & 19 \nl
532: 060526  & 3.21 & 52.67 & $20.0\pm 0.2$   & 1 & 3 & 20 \nl
533: 060605  & 3.78 & 52.73 & $...$           & 1 & 3 & 21 \nl
534: 060607  & 3.08 & 53.27 & $16.85\pm 0.10$   & 1 & 3 & 4 \nl
535: 060707  & 3.42 & 53.20 & $21.0\pm 0.2$   & 1 & 7 & 22 \nl
536: 060714  & 2.71 & 53.20 & $21.8\pm 0.1$   & 1 & 7 & 22 \nl
537: 060906  & 3.68 & 53.38 & $21.85\pm 0.1$  & 1 & 7 & 22 \nl
538: 060908  & 2.43 & 53.07 & $...$           & 0 & 1 & 23 \nl
539: 060926  & 3.20 & 52.24 & $22.7\pm 0.1$   & 1 & 7 & 24 \nl
540: 060927  & 5.47 & 53.49 & $22.5\pm 0.4$   & 1 & 7 & 25 \nl
541: 061110b & 3.44 & 53.08 & $...$           & 1 & 7 & 26 \nl
542: 061222b & 3.35 & 53.29 & $...$           & 0 & 1 & 27 \nl
543: 070110  & 2.35 & 52.78 & $...$           & 1 & 7 & 28 \nl
544: 070411  & 2.59 & 53.07 & $...$           & 1 & 1 & 29 \nl
545: 070506  & 2.31 & 51.88 & $...$           & 1 & 7 & 30 \nl
546: 070529  & 2.50 & 53.05 & $...$           & 0 & 1 & 31 \nl
547: 070611  & 2.04 & 52.01 & $...$           & 1 & 1 & 32 \nl
548: \enddata
549: \tablecomments{References:
550: 1.\ Castro et al.\ (2000); 2.\ Castro et al.\ (2003); 3.\ Vreeswijk et al.\ (2006); 
551: 4.\ Prochaska et al.\ in preparation; 5.\ Vreeswijk et al.\ (2004); 6.\ Jakobsson et al.\ (2004); 
552: 7.\ Watson et al.\ (2006);8.\ Berger et al.\ (2006b);9: Chen et al.\ (2005); 
553: 10.\ Prochaska et al.\ (2007a); 11.\ Kawai et al.\ (2006); 12.\ Piranomonte et al.\ (2007); 
554: 13.\ Quimby et al.\ (2005); 14.\ Piranomonte et al.\ (2006); 15.\ Fynbo et al.\ (2006b); 
555: 16.\ Cucchiara, Fox \& Berger (2006); 17: Berger et al.\ (2006a); 
556: 18.\ Price et al.\ (2007); 19.\ Cenko et al.\ (2006); 20.\ Berger \& Gladders (2006);
557: 21.\ Savaglio et al.\ (2007); 22.\ Jakobsson et al.\ (2006); 23.\ Rol et al.\ (2006);
558: 24.\ D'Elia et al.\ (2006); 25.\ Ruiz-Velasco et al.\ (2007); 26.\ Fynbo et al.\ (2006a);
559: 27.\ Berger (2006); 28.\ Jaunsen et al.\ (2007); 29.\ Jakobsson et al.\ (2007); 
560: 30.\ Thoene et al.\ (2007a); 31.\ Berger \etal\ (2007); 32.\ Thoene \etal\ (2007b); 
561: 33.\ Hjorth \etal\ (2003); 34.\ Shin \etal\ (2007).} 
562: \tablenotetext{a}{Isotropic equivalent energy release of $\gamma$-ray photons (erg s$^{-1}$). For pre-{\it Swift} bursts, the energy interval corresponds to $200-2000$ keV (Bloom \etal\ 2003; Sakamoto \etal\ 2005).  For {\it Swift} bursts (2005 and on), the energy interval corresponds to $10-150$ keV, and no k-corrections have been applied.} 
563: \tablenotetext{b}{f$_\alpha$: flag for available \lya\ coverage in the afterglow spectra: $0=$ No coverage; $1=$ \lya\ coverage.}
564: \tablenotetext{c}{f$_{i}$: cumulative flag describing the properties of the ISM surrouding the GRB: $1=$ presence of low ions (e.g.\ Si\,II, C\,II); $2=$ the gas is optically thick at the Lyman limit; $4=$ The \lya\ absorption strength indicates $N(\hI)> 2\times 10^{20}$ \cmjj.  For example, a flag$_{\rm ISM}=7$ indicates the system is a DLA with associated low ions.}
565: \end{deluxetable}
566: 
567: \end{document}
568: