The Pennsylvania State University
The Graduate School
Department of Astronomy and Astrophysics
GAMMA-RAY BURST AFTERGLOWS
AS PROBES
OF THEIR HOST GALAXIES
AND THE COSMOS
A Dissertation in
Astronomy and Astrophysics
by
Antonino Cucchiara
c 2010 Antonino Cucchiara
Submitted in Partial Fulfillment of the Requirements for the Degree of
Doctor of Philosophy
August 2010 The dissertation of Antonino Cucchiara was reviewed and approved1 by the following:
Derek B. Fox Assistant Professor of Astronomy and Astrophysics Dissertation Adviser Chair of Committee
Michael Eracleous Associate Professor of Astronomy and Astrophysics
Jane Charlton Professor of Astronomy and Astrophysics
Peter Meszaros Eberly Family Chair in Astronomy and Astrophysics Professor of Astronomy and Astrophysics Professor of Physics
Irina Mocioiu Assistant Professor of Physics
Lawrence Ramsey Professor of Astronomy and Astrophysics Head of the Department of Astronomy and Astrophysics
1Signatures on file in the Graduate School. iii Abstract
Gamma-ray Bursts (GRBs) represent the sole class of catastrophic phenomena seen over almost the entire history of the Universe. Their extreme luminosities in high energy γ-ray radiation make them readily detectable, even with relatively small satellite- based detectors, out to the earliest cosmic epochs. Moreover, the brilliance of their fading afterglow light, routinely observed in X-ray, optical, near-infrared, and radio wavelengths, allows them to be exploited – for hours, days, or weeks – as cosmic lighthouses, probing the conditions of gas and dust along the line of sight, through their host galaxies and the cosmos at large. Since the November 2004 launch of Swift, this GRB-focused NASA mission has discovered more than 500 GRBs, in almost all cases reporting the burst coordinates to ground-based observers within seconds of the event. The availability of prompt burst positions from Swift, combined with promptly-reported flux measurements from instru- ments on Swift and an array of ground-based robotic telescopes, have enabled targeted spectroscopic campaigns that have gathered detailed observations of the young, bright afterglows of hundreds of these events. This thesis reports the results of my own ef- forts over the past 5 years, analyzing imaging and spectroscopic observations of Swift- detected GRBs as triggered according to my own requests, or as gathered from public data archives. In Chapter 2, I discuss our follow-up campaign for GRB 090429B, one of our best “extreme redshift” (z > 8) candidates. This burst followed closely on the spectroscopically- confirmed z = 8.2 GRB 090423, and our multiwavelength observations and SED model- ing demonstrate the value and limitation of such studies, in cases where a spectroscopic redshift cannot be gathered in a timely fashion. I also address the importance of such extreme-redshift events from a cosmological perspective. In Chapter 3, I use high-resolution GRB afterglow spectra to study the properties of intervening absorbers along GRB lines of sight, in particular, the “very strong” Mg ii absorbers, which are often associated with Damped Lyman-α systems. By comparing the properties of the absorbers detected along GRB and quasar lines-of-sight, we attempt to shed light on the mysterious excess of such absorbers in GRB (as compared to quasar) spectra. Carrying out a battery of kinematic and qualitative tests, we fail to identify any respect in which the GRB systems systematically differ from the quasar systems, thus disfavoring the hypothesis that some or many of the GRB absorbers are associated with fast-moving gas in the GRB environs or host galaxy. iv
In Chapter 4, I present a summary of five years of GRB afterglow spectroscopic follow-up efforts, the Gemini catalog of mid-resolution afterglow spectra. This dataset consists of more than 50 GRBs observed with the GMOS instruments on the Gemini North and South telescopes; among these objects, we select 22 high signal-to-noise spec- tra for comprehensive analysis, reporting equivalent widths and upper or lower limits for a large number of atomic species often observed in absorption in GRB afterglow spectra. Applying these results, we explore the metallicity and dust content of the host galaxies of these GRBs, since these properties likely have important implications for the nature of the progenitors and the characterization of the GRB environment. Finally, in my conclusions, I review what we have learned over the past five years, and address some of the open questions in the use of GRB afterglows as cosmic probes that I expect will attract the greatest attention of researchers over the next few years. v TABLE OF CONTENTS
List of Tables ...... vii
List of Figures ...... viii
Acknowledgments ...... xi
Chapter 1. Introduction ...... 1 1.1 Gamma-RayBursts:AHistory ...... 1 1.2 GRBTheoryinBrief...... 4 1.3 Thesismotivation...... 7 1.3.1 GRBs in the early Universe ...... 7 1.3.2 Intervening systems along GRB lines of sight ...... 8 1.3.3 GRBhostchemistry ...... 9
Chapter 2. GRB 090429B: another extreme-redshift phenomena? ...... 11 2.1 Introduction...... 11 2.2 ObservationsandAnalysis...... 13 2.2.1 Optical and Near-IR data ...... 13 2.3 ResultsandDiscusson ...... 14 2.3.1 Spectral Energy Distribution ...... 16 2.3.2 Photometric Redshift Analysis ...... 16 2.3.3 Theoretical Considerations ...... 18 2.4 Conclusions ...... 19
Chapter 3. Testing the Possible Intrinsic Origin of the Excess of Very Strong MgII Absorbers Along GRB Lines-of-Sight ...... 25 3.1 Introduction...... 25 3.2 MotivationforThisStudy ...... 26 3.2.1 BiasDuetoDust...... 27 3.2.2 Partial Beam Coverage Explanations ...... 28 3.2.3 Lensing Amplification of GRB Beam-size ...... 29 3.2.4 An Origin Intrinsic to GRBs or Their Environments . . . . . 29 3.2.5 Motivation for This Study ...... 30 3.3 Datasetandanalysis ...... 31 3.4 Results...... 33 3.4.1 Mg ii Kinematics...... 41 vi
3.4.2 Equivalent Width Comparisons ...... 41 3.4.3 Fe ii to Mg ii Ratio ...... 43 3.4.4 Dust...... 46 3.4.5 C iv KinematicsandPhaseStructure ...... 46 3.4.6 Fine Structure and Metastable Transitions ...... 48 3.5 Discussion of Intrinsic Origin of Excess GRB Very Strong Mg ii Ab- sorbers...... 49 3.6 SummaryofResults ...... 52
Chapter 4. Gemini Spectroscopic Catalog of Swift GRB Afterglows ...... 55 4.1 Introduction...... 55 4.2 Datasampleandanalysis ...... 56 4.2.1 Datareductionprocedure ...... 60 4.3 Equivalent widths measure and uncertainties ...... 61 4.3.1 Low-resolution spectra ...... 62 4.3.2 Metallicity and dust tracers ...... 66 4.4 Conclusion...... 71
Chapter 5. Conclusions and Future work ...... 74 5.1 GRB afterglow spectroscopy: where do we stand? ...... 74 5.2 AfterglowsasProbes...... 74 5.3 Intothefuture...... 76
Appendix A. GRB high-resolution absorption features ...... 78
Appendix B. QSO high-resolution absorption features ...... 88
Appendix C. Catalog of GRB Gemini spectra ...... 116
Bibliography ...... 126 vii List of Tables
2.1 GroundBasedobservationssummary...... 15
3.1 Observationlog...... 32 3.2 Systemsidentified ...... 37 3.3 Equivalentwidths ...... 38 3.4 Equivalentwidths ...... 39 3.5 Equivalentwidths ...... 40
4.1 GRBAfterglowSpectralSample ...... 57 4.2 Excluded GRB Afterglow Spectral Sample ...... 58 4.3 Observed EW comparison based on GRB 060418 data ...... 64 4.4 Equivalent widths of the transitions ...... 67 4.5 Equivalent widths of the transitions (continue) ...... 68 4.6 Equivalent widths of the transitions (continue) ...... 69 4.7 Equivalent widths of the transitions (continue) ...... 70 viii List of Figures
1.1 BATSEGRBsspatialdistribution ...... 2 1.2 BATSEGRBdurationdistribution ...... 3 1.3 PromptGRBspectrum ...... 4 1.4 GRBschematic...... 5 1.5 GRBphysicalphenomena ...... 6
2.1 XRTlightcurve ...... 21 2.2 GeminiNorthImaging...... 22 2.3 hyper-z ...... 23 2.4 hyper-z ...... 23 2.5 SpectralEnergyDistribution ...... 24
3.1 kinematicplots ...... 35 3.2 kinematicplots ...... 36 3.3 kinematicplots ...... 42 3.4 kinematicplots ...... 44 3.5 kinematicplots ...... 45 3.6 kinematicplots ...... 47 3.7 kinematicplots ...... 50
4.1 Gratings...... 59 4.2 GRB060418comparison ...... 63 4.3 Zinc-Cromiumratioplot...... 72
A.1 GRB 060418 absorber absorption lines ...... 79 A.2 GRB 060418 absorber absorption lines ...... 80 A.3 GRB 050820 absorber absorption lines ...... 81 A.4 GRB 060418 absorber absorption lines ...... 82 A.5 GRB 021004 absorber absorption lines ...... 83 A.6 GRB 050820 absorber absorption lines ...... 84 A.7 GRB 021004 absorber absorption lines ...... 85 A.8 GRB 060607 absorber absorption lines ...... 86 A.9 GRB 050730 absorber absorption lines ...... 87
B.1 QSO0100+0130 absorber absorption lines ...... 89 ix
B.2 Q1127-145 absorber absorption lines ...... 90 B.3 Q1229-021 absorber absorption lines ...... 91 B.4 3C336absorberabsorptionlines...... 92 B.5 HE1122-1648 absorber absorption lines ...... 93 B.6 Q0453-423 absorber absorption lines ...... 94 B.7 Q0002-422 absorber absorption lines ...... 95 B.8 3C336absorberabsorptionlines...... 96 B.9 Q1629+120 absorber absorption lines ...... 97 B.10 Q0130-4021 absorber absorption lines ...... 98 B.11 CTQ0298 absorber absorption lines ...... 99 B.12 Q0328-272 absorber absorption lines ...... 100 B.13 Q1621-0042 absorber absorption lines ...... 101 B.14 Q0453-423 absorber absorption lines ...... 102 B.15 Q0112+0300 absorber absorption lines ...... 103 B.16 HE1341-1020 absorber absorption lines ...... 104 B.17 Q0109-3518 absorber absorption lines ...... 105 B.18 PKS0237-23 absorber absorption lines ...... 106 B.19 Q1418-064 absorber absorption lines ...... 107 B.20 PKS0237-23 absorber absorption lines ...... 108 B.21 HE2217-2818 absorber absorption lines ...... 109 B.22 QSO0100+0130 absorber absorption lines ...... 110 B.23 Q1331+170 absorber absorption lines ...... 111 B.24 Q0100+130 absorber absorption lines ...... 112 B.25 Q1151+068 absorber absorption lines ...... 113 B.26 Q0551-3637 absorber absorption lines ...... 114 B.27 Q0002-422 absorber absorption lines ...... 115
C.1 GRB060510B...... 116 C.2 GRB060210 ...... 117 C.3 GRB081029 ...... 117 C.4 GRB050908 ...... 118 C.5 GRB070529 ...... 118 C.6 GRB080413A ...... 119 C.7 GRB080804 ...... 119 C.8 GRB081008 ...... 120 C.9 GRB080319C ...... 120 C.10GRB080928 ...... 121 C.11GRB060502 ...... 121 C.12GRB060418 ...... 121 x
C.13GRB080604 ...... 122 C.14GRB071122 ...... 122 C.15GRB091024 ...... 123 C.16GRB091208B ...... 123 C.17GRB071010B ...... 123 C.18GRB080710 ...... 124 C.19GRB060729 ...... 125 C.20GRB081007 ...... 125 xi Acknowledgments
I started this journey in May 2005, when John Nousek, Dave Burrows and Pete Roming decided to offer me the position of Science Planner for the Swift satellite. I was incredibly happy and I will never forget their support, their patience and the astonishing opportunities they gave me in the 8 months I work for Swift. I will always be thankful to them, and all the people at Swift. Of course you Claudio and Debora, who hot me for more than a month while I was looking for an apartment to live in. Now, after 5 years, from when on December, 22 2005 I became officially a PSU grad student lot of things happened and lot of people I have met. Many thanks go for sure to my advisor, Derek B. Fox, who believe in me as his first student. I can say it was not always easy for my and his background, for the language difference and many other differences. But I have to admit, we came along way together and I will proud to walk with you and earn my PhD. Leaving your own country is never easy and I have been taught by my parents what that means 26 years ago when they left sicily and their families for getting jobs in north Italy. Mamma, Pap´aand Enrica, my sister, always have been part of my life even if there was an ocean in between. I love you so much and I will reward you forever. I also want to thank my grandparents, Concetta and Nino, who prayed for me and watch me from Heaven with my other grandparents. I know I made you proud of this and I will make you proud every day of my life. I cannot forget how many people helped me out and bring me to this thesis. I need to thank Dr. Jane Charlton for all the wonderful words, help and encouragement she gave me before my Comps, during my first paper work (which, thanks to her, was actually published with minor comments from the referee). Also, I want to acknowledge Dr. Abe Falcone and Dr. Mike Eracleous for the terrific advises in those moments in which I faced the dilemma of “living or staying”, and helped me in my relation with Derek. Finally, I need to thank my girlfriend, Giulia, who believed in me in the last part of this PhD, when our relationship, still young, needed lot of things to work on. Grazie, piccola mia. For all of these people and many more I did not list here, I thank God, Who brought them close to me to show me the right (and the wrong) way. Thanks to all, Nino. xii
The important thing is not to stop question- ing. Curiosity has its own reason for existing. One cannot help but be in awe when he con- templates the mysteries of eternity, of life, of the marvelous structure of reality. It is enough if one tries merely to comprehend a little of this mystery every day. Never lose a holy curiosity – Albert Einstein (1879 - 1955) 1
Chapter 1
Introduction
1.1 Gamma-Ray Bursts: A History
In July 1967 the Vela satellites, launched with the intent to verify the Nuclear Test Ban Treaty between the United States and the Soviet Union, observed a γ-ray pulse originating from outer space. The cosmic origin of this emission – not from any man-made source or Solar System object – was clear from early in the team’s analysis, but given the classified nature of the mission, several years had to pass before the data could be announced publicly, opening a new window on the exploration of the cosmos (Klebesadel et al. 1973). The first gamma-ray bursts (GRBs) had been discovered. In the subsequent decade thousands of GRBs were detected in the γ-ray en- ergy regime thanks primarily to the Burst and Transient Source Experiment (BATSE) on board the Compton Gamma-Ray Observatory (CGRO). As an observational phe- nomenon, these rapid flashes last from a fraction of a second up to a few minutes; during this time, each GRB typically outshines the summed γ-ray emissions of every other source in the γ-ray sky. BATSE detected almost 3000 GRBs during its nearly 10-year lifetime, and deduced some important results about this fascinating phenomena. First of all, the distribution of detected GRBs over the sky is consistent with isotropy to a high degree (Fig. 1.1), strongly suggesting an extragalactic nature for these events. Second, with the large and uniform BATSE sample it proved possible to perform, for exam- ple, classification studies. Based on their duration, usually defined by the time interval over which the middle 50% (T50) or 90% (T90) of the counts above the background are measured, two classes of GRBs were identified: GRBs that last less than 2 seconds are classified as short GRBS, while GRBs with duration longer than 2 seconds are classified as long (Fig. 1.2). The relatively poor (degree-scale) localizations and transient nature of GRBs made it difficult to efficiently observe their location with X-ray satellites or, at longer wavelengths, from ground-based facilities. In 1997, the Italian-Dutch satellite BeppoSAX was able to detect for the first time the X-ray counterpart of GRB 970111 (Costa et al. 1997) and, with the discovery of the X-ray (Piro et al. 1999) and optical (van Paradijs 2
2704 BATSE Gamma-Ray Bursts +90
+180 -180
-90
10-7 10-6 10-5 10-4 Fluence, 50-300 keV (ergs cm-2)
Fig. 1.1 Distribution of the ∼ 2700 GRBs observed by BATSE. The isotropic distribu- tion in galactic coordinates is an indication of the extragalactic origin of GRBs. From: http://www.batse.msfc.nasa.gov/batse/grb/skymap/ et al. 1997) counterparts of GRB 970228 , opened an entirely new field of study: the science of GRB afterglows. After these ground-breaking discoveries others quickly fol- lowed, including GRB 970508, with a bright and long-lived afterglow that was observed over several orders of magnitude in energy range. At radio wavelengths the afterglow was studied for over a year, offering important clues about the size, relativistic expansion, and eventual deceleration of the GRB blastwave (Frail et al. 2000). The optical spectrum of the GRB 970508 afterglow, the first to be gathered for any GRB (Metzger et al. 1997), shows absorption features associated with the burst’s host galaxy, and thus provided both a redshift (z = 0.835) and a definitive demonstration of a cosmological origin for at least this GRB. Absorption features at lower redshift were also identified; these are interpreted as being produced by intervening gas along the line of sight towards the GRB, in similar fashion to the lower-redshift absorption systems often observed in quasar spectra. Thanks to continuing efforts of the BeppoSAX team, other satellite teams, and ground-based observers, the number of GRBs with X-ray, optical, or infrared afterglows detected increased year after year, providing new insights into the GRB progenitors and the emission mechanisms, as well as their host galaxies. Another important discovery in these early stages of GRB history was that of the GRB-supernova connection: follow-up observations of GRB 980425 discovered a su- pernova (SN 1998bw) rather than an afterglow (Galama et al. 1998). The properties of 3
Fig. 1.2 Distribution of the T90 values for GRBs detected by BATSE. The two populations of long and short bursts are clearly evident. From: http://www.batse.msfc.nasa.gov/batse/grb/skymap/ this supernova – including its extreme optical luminosity, high-velocity features, and the presence of relativistic ejecta (Kulkarni et al. 1998) – that many inferred an association between the SN and the GRB, even though the resulting GRB had an energy more than 104 times lower than the energies of the more frequent z ≈ 1 events. Over years of subsequent afterglow studies, evidence accumulated that GRB af- terglows, when observed closely and at relatively low redshift (z < 1), could be seen to exhibit a “supernova bump” with the right peak magnitude, peak timescale, and colors to be explained as a coincident type Ibc supernova, much like SN 1998bw. Spec- troscopic characterization of these “bumps” was missing, however, until the explosion of GRB 030329, detected by HETE-2. The time sequence of spectra gathered for this z = 0.17 event (Stanek et al. 2003; Hjorth et al. 2003) show an early power-law contin- −0.9 uum (Fν ∝ ν ), typical of GRB afterglows, that slowly evolves into the spectrum of a type Ic supernova. Thus did GRB 030329 serve to confirm a massive star progenitor origin for at least some of the long-duration GRBs. Successful operations of Swift over the past five years have indeed served to shed light on many of the previous era’s greatest GRB mysteries (Gehrels et al. 2009).33 of the short-duration bursts have been localized, with eight having redshift determinations from their host galaxies; the short bursts have thus also been revealed as a cosmological population, albeit with a substantially larger fraction of low-redshift (z < 1) events than in the long-burst sample. The quantity and quality of the X-ray lightcurves of GRB afterglows has increased enormously, revealing unexpected features that likely signal ex- tended (minutes to hours-long) activity of the GRB central engine. The speed and accu- racy of the burst positions from Swift have enabled high signal-to-noise, high-resolution 4 spectroscopy of the young, bright afterglows, revealing detailed abundance patterns, sig- natures of fine-structure (metastable) atomic states, and even, time-variability of these features. And finally, the frontier of GRB redshifts has been pushed out to the earliest epochs, at z > 8, within the first billion years of cosmic time. Thanks to dedicated efforts by multiple groups around the world, an array of small-aperture telescopes are operating and primed to robotically point towards burst positions, as indicated by Swift, on timescales of seconds to minutes. The number of GRBs detected by these facilities in the optical and infrared has increased considerably, providing a better understanding of the early phases of the afterglow (see section 1.2). Optical coverage of the early stages of the afterglow has provided remarkable insights into the nature of the emitting regions (see the “naked-eye” GRB 080319B, Racusin et al. 2008).
Fig. 1.3 Typical GRB high-energy spectrum. The photon energy ranges covered by the Gamma-ray Burst Monitor and the Large Area Telescope on board the Fermi Gamma- Ray Observatory are indicated. From: http://fermi.gsfc.nasa.gov
Looking ahead, high-energy detectors on the new satellites AGILE and Fermi are now detecting GRB-associated photons at >GeV energies, thus providing access to a new territory for GRB studies (Fig. 1.3)
1.2 GRB Theory in Brief
In this section I will attempt to summarize current knowledge of GRBs, their pro- genitors, and the emission mechanisms at work from the prompt phase until the afterglow synchrotron emission dominates over the prompt emission leading to the afterglow phase. 5
The fireball model describes the physical mechanisms that produce a GRB, while explicitly ignoring any physics (or other complications) associated with the engine itself; −5 all that is required is that the engine impute a small mass M ∼ 10 M⊙ of baryonic 51 material with sufficient energy E ∼> 10 erg, such that its free expansion and acceleration, out to large distance, yields a highly-relativistic outflow, Γ ∼> 100. Starting from this simple premise, the fireball model explains the basic obser- vational components of GRBs as resulting from the collisions of relativistic shells of material emitted from the central engine, both with each other (“internal shocks”) and with the circumburst material surrounding the GRB progenitor (the “external shock”). The internal shocks between the blobs produce the synchrotron radiation which is ob- served as the prompt γ-ray emission of the GRB. The “external” shock, on the other hand, is responsible for the optical, X-ray, and radio afterglow. A schematic diagram of a GRB is shown in Fig. 1.4.
Fig. 1.4 Schematic representation of a GRB, from the possible progenitors to the in- teraction of the blastwave with the interstellar medium, which produces the afterglow emission. From: http://www.astro.psu.edu/ niel/astro130/images
The prompt γ-ray spectra of GRBs is generally adequately fit by a Band function (Band et al. 1993), a smooth joining of two power-law spectra, with peak energy (in a νFν sense) in the hard X-ray to soft γ-ray regime, ∼>100 keV. Although the Fermi Large Area Telescope (LAT) has now detected GRB photons with energies up to 30 GeV, even at these energies deviations from the Band spectral form are extremely rare. 6
This has prompted intense speculation about the possible origins of the highest- energy emissions, with proposals including mechanisms such as synchrotron self-compton (SSC) emission, SSC from the reverse shock, and inverse Compton scattering off thermal X-rays.
Fig. 1.5 Key GRB physical phenomena and their relevant length scales. On the smallest, 106 cm scale, we find the GRB engine (black hole or neutron star) and its magnetosphere. At 108 cm there is the accretion flow, and an associated relativistic outflow from the engine that is collimated into two jets. At scales of 1010 cm, collimation is produced in the collapsar model as the outflow is channelled preferentially along the rotation axis. 1012 cm is the typical size of an evolved massive star progenitor; a thermal break-out signal is expected at some level to precede the GRB itself. At 1014 cm, velocity differences between multiple expanding shells lead to collisions, shocks, and γ-ray emission; this is the most favorable region for producing the highly variable γ-ray light curves. At 1016 cm the external shock is produced by the interaction of the expanding ejecta with the swept-up circumburst material. At 1018 cm the relativistic phase ends; and on a length scale of 1020 cm we see the GRB environment on a subgalactic scale. Figure from Gehrels et al. (2009).
GRB activity is thought to be manifested, in observationally significant ways, over a range of 13 orders of magnitude in radius (Fig. 1.5). The nature of the central engine remains a mystery, although given the γ-ray energies and kinetic luminosities involved, a relativistic and compact progenitor, in par- ticular a neutron star or black hole, is strongly implicated. Accretion onto this central object, or magnetic fields threading the disk and tapping the central object’s rotational energy, are the most promising energy reservoirs, powering twin jets that expand to large distance and accelerate to highly-relativistic (Γ ∼> 100) velocities. As we have seen al- ready in our discussion of the fireball model above, internal shocks generated by colliding shells in the relativistic jet are thought to explain the prompt γ-ray emission, while on a substantially larger length scale, an external shock develops as the blastwave sweeps up the external medium, powering the subsequent afterglow. Eventually, at length scales of r ∼ 1 pc, the energy of the blastwave is dissipated and it transitions to subrelativistic flow. 7
1.3 Thesis motivation
Much of what we have learned about GRBs has been derived through studies of the burst afterglows, as they illuminate the circumburst environment, their host galaxies, and material at cosmological distances along the line of sight. In the paragraphs below I will briefly present the motivation of this thesis, which emphasizes the unique insights offered by spectroscopic and (to a lesser degree) photo- metric observations of GRB afterglows, both with respect to the nature of the GRBs themselves, and the wider Universe.
1.3.1 GRBs in the early Universe
The Swift-era discovery of GRB 050904 (Kawai et al. 2006) and GRB 080913 (Greiner et al. 2009) at z > 6 first opened the door to the exploration of the re-ionization era using GRB afterglow spectroscopy. The association of the long-duration GRBs with the death of massive stars circumvents the growth limitations that cause the number density of quasars to decline exponentially at high redshift. As such, this makes GRBs a highly valuable tool not only for studying the star formation history of the Universe, but also to act as “cosmic lighthouses” illuminating some of the most distant galaxies known. The first generations of massive stars are considered to be primary actors in the ionization of the intergalactic medium (IGM), its present state. From studies of cosmic microwave background anisotropies, authors have argued that this process was over at around z = 10 ± 1, while spectroscopy of high-redshift quasars suggests that the the IGM was still partly neutral at z ≈ 6 (Becker et al. 2001; Mesinger 2009). A detailed description of the cosmic reionization is thus still required, as well as a characterization of the early star and galaxy formation processes which yielded the ultra-violet radiation that powered it. The intrinsic faintness of these galaxies is a true challenge for the observations, although the newly-commissioned WFC3 on HST has proved capable of identifying the brightest of these photometrically (Wilkins et al. 2010), and the forthcoming James Webb Space Telescope will provide an enormous further leap forward in capabilities once it is in full operation. In the meantime, the fact that GRBs can be seen up to z ∼ 20 (Lamb & Reichart 2000) gives us an opportunity to study the re-ionization era with current facilities. In order to address the likely returns from such studies, authors have been modeling the cosmic star-formation rate and predicting the corresponding GRB redshift distribution (Bromm & Loeb 2006), considering alternative scenarios where GRBs either are, or are 8 not, associated as well with the death of Population III stars at the highest redshifts. Within their model, Bromm & Loeb 2006 were able to predict that ∼10% of Swift bursts should originate at z ≥ 5. Observations largely agree with these predictions, tightening constraints also on the escape fraction of ionizing photons including GRBs up to high redshift (Wyithe et al. 2010a). The desire to identify and study GRBs from the reionization era is further stim- ulated by the prospect of using their bright afterglows, via high signal-to-noise spec- troscopy, to address crucial questions about that era. Such observations could provide a measure of the hydrogen neutral fraction in the IGM at the GRB redshift and line of sight. At the same time, understanding the chemistry of GRB progenitors will shed some light on the nature of the first generation of stars, and on how metals have been produced and recycled throughout the history of the Universe. Our discovery and observations of GRB 090429B, as described in Chapter 2, ad- dresses an important piece of this puzzle. After describing our observations, we present our SED modeling which leads to a photometric redshift of z ≈ 9.5, but also illustrates the challenges of this technique: the degeneracy between extreme-redshift and, lower redshift, highly-absorbed interpretations of candidate high-redshift afterglows.
1.3.2 Intervening systems along GRB lines of sight
The importance of GRB afterglow spectroscopy is now well known. This tech- nique is fundamental for determining the redshift, and so the distance, of this extremely energetic phenomena. Apart from the absorption signatures imprinted by the GRB environs and host galaxy, these spectra often exhibit additional features produced by intervening systems (galaxies and other collections of absorbing gas) along the line of sight. This phenomenon is familiar from studies of quasar spectra, where the intervening material is studied and classified as damped Lyman-α systems (DLAs), galactic halos, galactic super-winds, and so forth; and each such class of absorber has been characterized in numerous ways. One particular class of these objects has proved to be of particular interest for comparative studies of GRB and quasar lines of sight: the very strong Mg ii absorbers, which exhibit equivalent widths of the blue component of the doublet larger than 1 A,˚ and are frequently used as tracers of DLAs (Rao et al. 2006). The reason the very strong Mg ii absorbers have attracted this attention is that the redshift path density (dN/dz) is almost 4 times greater along GRB lines of sight, compared to the density along the lines of sight to quasars (Prochter et al. 2006b). This is 9 an entirely unexpected result that, in an extreme interpretation, would cast doubt on the interpretation of these absorbers as cosmologically distant from the GRB, at least in some cases. Several authors have thus explored whether this discrepancy can be explained by invoking dusty foreground objects which bias optically-selected quasar samples (M´enard et al. 2008); geometric effects which facilitate detection of strong absorbers along GRB lines of sight (Frank et al. 2007); gravitational lensing by absorbers which would increase the probability of observing an absorber toward a bright GRB (Wyithe et al. 2010b; Vergani et al. 2009); or finally, the presence of high-velocity gas in the GRB environment or host galaxy. In this last scenario some of the absorbers would have been mistakenly interpreted as unrelated systems at lower redshift, because of their large velocity (blue shift) relative to the host galaxy itself. Intrigued by the latter explanation, in Chapter 3 we investigated this possibility by exploring the properties of very strong Mg ii absorbers along quasar and GRB lines of sight, in a sample of high-resolution spectra obtained with the VLT-UVES spectro- graph. Although the small number of GRBs in our sample prevent us from obtaining a definitive solution for the dN/dz discrepancy, it seems that an intrinsic origin for the excess absorbers is disfavored. More data are needed and probably a systematic high- resolution imaging campaign of the GRB host galaxies that exhibit very strong Mg ii absorbers will help in understanding this phenomenon.
1.3.3 GRB host chemistry
In recent years there has been a large discussion about whether GRB host galaxies and, hence, GRB progenitors are metal-poor by comparison to a larger population of star-forming galaxies. The metallicity of GRB hosts and also detailed abundance profiles can be measured directly from high-resolution spectroscopy of bright afterglows. To date several spectra of this quality have been collected for single bursts (e.g. Vreeswijk et al. 2004; Chen et al. 2005; Fynbo et al. 2006). Separately, the host light-averaged metallicities of multiple GRB host galaxies have been measured via standard emission- line diagnostics (Stanek et al. 2006a; Savaglio et al. 2009). Although these studies agree, in a general sense, that the metallicities of host galaxies at z . 1 are significantly (Z ∼ 0.1Z⊙) sub-solar (Savaglio et al. 2009), consistent with the sub-solar metallicity measured for GRB hosts via absorption spectroscopy at z & 2, it is unclear whether this distinguishes them strongly from the general field galaxy population at high redshift, and thus, whether a low-metallicity progenitor is uniquely indicated. 10
Separately, the contribution of host galaxy dust to the general problem of optically- dark GRBs continues to be investigated. Perley et al. (2009) performed an observational campaign of 14 GRBs classified as dark in order to investigate the dichotomy between high extinction vs. high-z nature. The number of high-redshift (z > 7) events has been constrained to be between 0.2% and 7% of the Swift sample, with several GRBs in the work argued to have AV = 2 to 5 mag. Even though these events were simply highly obscured, we are far from a complete understanding on the diagnostic of the GRB envi- ronment and its effect on optically selected afterglow samples (Levesque et al. 2010). To further explore these questions, it is worth exploiting, as far as possible, the large number of mid- and low-resolution spectra of GRB afterglows that have been col- lected in the Swift era, in order to study the chemistry of the GRB hosts and intervening absorbers. Recently Fynbo et al. (2009) presented a large catalog of GRB afterglow spec- tra obtained by several facilities around the world. In a similar way, in Chapter 4 we present our large dataset obtained with the Gemini telescopes and the GMOS spectro- graphs. We observed 22 long-duration GRBs with sufficient signal to noise for this study, spanning a large redshift range (0.49 Chapter 2 GRB 090429B: another extreme-redshift phenomena? Gamma-ray bursts (GRBs) and their afterglows have the potential to serve as powerful probes of the early Universe, revealing the locations and physical properties of active star-forming regions out to the highest redshifts, and providing direct con- straints on the reionization of the intergalactic medium. Thanks to the burst detection and rapid-response capabilities of the Swift satellite observatory, this potential is now being realized, with three spectroscopically-confirmed GRBs identified to-date at red- shifts z > 6. In this paper we present Swift and ground-based photometric observations of GRB 090429B and its near-infrared afterglow, which exhibits the drop-out signature of an “extreme redshift” (z > 8) burst. Just such a signature successfully identified GRB 090423 (z = 8.2) as an extreme-redshift candidate prior to its first spectroscopic observations. In the absence of confirming spectroscopy for GRB 090429B, we make a deep host galaxy search and perform an extensive photometric redshift modeling effort, seeking to characterize the regimes of redshift and extinction that are consistent with the data, under a range of physical scenarios. While our analysis yields a photometric red- +0.25 shift zphot = 9.5−0.60 (90%-confidence) with negligible host extinction, as observed from GRB 090423, we find the data do not rule out lower-redshift, higher-extinction scenar- ios; for example, GRB 090429B may have occurred along an extinguished (AV ≈ 1 mag) line of sight in a galaxy at z ≈ 5, or along a very-extinguished (AV ≈ 2 mag) line of sight in a galaxy at z ≈ 2. Our non-detection (RAB > 27 mag) of any underlying host galaxy allows us to set a firm upper limit on its luminosity of LR < 0.03L∗ for z = 1. GRBs 090429B and 090423 represent two examples of multiband NIR follow-up yield- ing candidate extreme-redshift (z > 8) events in a prompt fashion. Current and future multiband NIR follow-up instruments like GROND and RATIR are likely to increase the rate of discovery of such afterglows and enable rapid-response NIR-spectroscopic obser- vations with optical/NIR spectrometers like X-shooter on the Very Large Telescope and GNIRS on Gemini. 2.1 Introduction Gamma-ray bursts (GRB) have now been confirmed as among the most distant objects in the Universe (Tanvir et al. 2009b), illuminating conditions of star formation 12 in the earliest epochs. As burst detections push towards progressively higher redshifts, the mere existence of GRBs at these times will provide an increasingly stringent test of models of gravitational collapse, galaxy formation, and the first generations of stars. At the same time, high-quality spectroscopy of the burst afterglows can be expected to reveal element abundances, host galaxy kinematics, and potentially, the HI fraction of the intergalactic medium (IGM), as the process of cosmic reionization unfolds (Totani et al. 2006, e.g.,). In order to realize these ambitions, and efficiently exploit high-redshift GRBs as probes of the extreme-redshift (z > 8) Universe, the GRB follow-up community has been devoting increasing effort to the rapid identification of GRB near-infrared (NIR) afterglows. For several years, our efforts have centered on the rapid-response capabil- ity of the Gemini Telescopes on Mauna Kea and Cerro Pachon, and on construction, commissioning, and operation of the seven-band grizJHK Gamma-Ray Burst Optical and Near-Infrared Detector (GROND; Greiner et al. 2008) at the European Southern Observatory. The initial redshift constraint for GRB 090423, the first confirmed extreme-redshift GRB, was due to our Gemini multiband NIR imaging sequence, which established the afterglow of this burst as a Y -band drop-out having redshift 7 . z . 9 (Cucchiara et al. 2009b). This photometric redshift was confirmed in spectroscopic observations performed at the Very Large Telescope (Tanvir et al. 2009a) and the Telescopio Nazionale Galileo (Fernandez-Soto et al. 2009; Thoene et al. 2009). The full multi-wavelength dataset for this event has been discussed in several papers (Tanvir et al. 2009b; Salvaterra et al. 2009). In this paper we will discuss our high-energy discovery and optical/NIR obser- vations of GRB 090429B, discovered by the Swift satellite. As with GRB 090423, early- time optical limits and NIR afterglow detections established this event – with its z-band dropout afterglow, and blue H − K color – as a candidate extreme-redshift GRB. Due to observational constraints, we were not able to measure a spectroscopic redshift from any facility. If GRB 090429B is indeed an extreme-redshift event, then it is likely at an even higher redshift than GRB 090423, z ≈ 9.5. The paper is structured as follow: in Sec 2.2 we present our full dataset on GRB 090429B; in Sec. 2.3 we will discuss the possibility that GRB 090429B is a second extreme-redshift GRB using photometric redshift modeling and theoretical arguments. Finally, in Sec. 2.4 we summarize our conclusions, highlighting the importance of future infrared robotic facilities and NIR spectroscopy for the study of the early Universe using these rare and highly important class of GRBs. 13 2.2 Observations and Analysis The Burst Alert Telescope (BAT; Barthelmy et al. 2005) aboard the Swiftsatellite triggered on GRB 090429B at T0 = 05:30:03 UT. The 15–350 keV light curve is composed of three distinct peaks with a total duration T90 = 5.5 s, and the time-integrated spectrum can be fitted by a single power-law with an exponential cut-off. The derived total fluence −7 −2 in the 15–150 keV band is 3.1 × 10 ergs cm , with Epeak = 49 keV. This peak energy is among the few detected by Swift within the BAT bandpass. After 106 s, the narrow-field instruments began their standard burst-response observation sequence. The X-ray Telescope (XRT; Burrows et al. 2005) identified an uncataloged fading source at RA(J2000)=14h02m40s.10, Dec(J2000)=+32◦10′14′′.6; no optical/UV counterpart was seen in the UV-Optical Telescope (UVOT; Roming et al. 2005) data. The X-ray data has been characterized using standard routines in Heasoft, Xspec, and QDP, with the light curve fitting process as described in Evans et al. (2009). For some analyses, we have used the automatic data products produced by the UK Swift Science Data Centre (Evans et al. 2007, 2009). The time-averaged 0.3–10 keV X-ray spectrum from 97–29893 seconds after the burst is best fit by a power-law with photon +0.16 +4.6 21 spectral index αX = 2.01−0.24 and with an absorption column of NH = 1.01−5.3 × 10 cm−2, in excess of the Galactic absorption of 1.2 × 1020 cm−2 (at >2σ confidence level). The X-ray light curve, shown in Figure 2.1, is adequately fit by a combination of bright- ening and fading temporal power laws: initially, the X-ray flux rises with temporal index +0.43 +146 β1 = −0.96−0.52, referenced to the burst time; following the peak time TX = 589−80 s, +0.08 the light curve then breaks to a power-law decay with β2 = 1.20−0.07. In the lower panel of Figure 2.1, we show evidence that the X-ray spectrum softens after the peak. 2.2.1 Optical and Near-IR data The Gamma-Ray Burst Optical and Near-Infrared Detector (GROND, Greiner et al. 2008) observed the field of GRB 090429B using simultaneously in its (dichroic- ′ ′ ′ ′ defined) g r i z JHKs filter set beginning 14 minutes after the Swift discovery (Olivares et al. 2009). Reduction and photometric analysis were performed using IRAF software 1. Images were calibrated using field stars from the SDSS catalog in the optical the 2MASS 1IRAF is distributed by the National Optical Astronomy Observatories, which are operated by AURA, Inc., under cooperative agreement with NSF 14 catalog in the NIR. No optical/NIR source was detected at the X-ray afterglow position; upper limits on the brightness of any associated source are presented in Table 2.1. Beginning roughly 3 hours after the burst trigger, we attempted TOO observations from Gemini-North. We gathered optical i′z′imaging with the Gemini Multi-Object Spectrograph (GMOS; Hook et al. 2004) and NIR JHK imaging with the Near-Infared Imager (NIRI; Hodapp et al. 2003). GMOS observations consist of 5 exposures of 3 minutes each, per filter; NIRI observations consist of 5 dithered positions of 60 s each. The Gemini GMOS and NIRI tasks under the IRAF environment were used to sky-subtract, align, and combine the images; photometry was performed relative to SDSS stars for the GMOS data, and relative to secondary calibrators from GROND for the NIRI data (Table 2.1). While no optical counterpart was present in our i′ or z′ images, comparison with catalog images allowed us to identify an infrared afterglow candidate in J, H, and K bands, at RA=14h02m40′′5, Dec =+32◦10 ′14′′.32. Although the candidate was identi- fied during the night, due to deteriorating atmospheric conditions no spectroscopic data could be gathered. A subsequent Gemini-N + NIRI K-band observation on April 30 UT revealed a clear fading of ∼ 1.2 mag of the identified source, confirming its transient nature (corresponding to a power-law index of αK = 0.48 ± 0.10, steeper than the X- raydata). Figure 2.2 presents the imaging data from our IR observations while Figure 2.1 presents the resulting photometric dataset, including also data from the literature. No evidence of a host galaxy is present in our images. A deep R-band image of the field, taken under good conditions (0.4”seeing) at 14 days after the GRB (Fig. 2.2), allows us to place a 3σ upper limit on the host galaxy apparent magnitude of RAB > 27.07 mag. The nearest resolved galaxy to the GRB position is at a distance of 45 arcseconds. 2.3 Results and Discusson The lack of an optical or infrared spectrum for GRB 090429B makes it impossible to determine its redshift in definitive fashion. Instead, we find it necessary to pursue alternative approaches, making use of our multiband photometry and our understand- ing of typical GRB afterglow properties in the Swift era. In the following sections we describe how our multiwavelength observations allow us to model the afterglow spectral energy distribution and interpret the X-ray and optical/NIR properties self-consistently as being those of an extreme-redshift (z > 8) GRB. However, based on our photomet- ric uncertainties and the possibilities for substantial intrinsic (host galaxy) extinction towards the burst, we ultimately cannot exclude lower-redshift solutions. 15 Table 2.1. Ground Based observations summary T-T0 (sec) Magnitude Filter Telescope Note 960 > 23.0 g’ GROND 5-sigma 960 > 23.0 r’ GROND 5-sigma 960 > 22.3 i’ GROND 5-sigma 960 > 21.8 z’ GROND 5-sigma 960 > 21.3 J GROND 5-sigma 960 > 20.7 H GROND 5-sigma 960 > 19.4 K GROND 5-sigma 3600 > 25.3 R VLT/FORS2 3-sigma 10437 > 25.6 i’ Gemini-N 3-sigma 10437 > 24.1 z’ Gemini-N 3-sigma 10437 22.78 ± 0.22 J Gemini-N 10437 21.53 ± 0.05 H Gemini-N 10437 21.24 ± 0.03 K Gemini-N 97200 22.40 ± 0.16 K Gemini-N 1.2 × 106 > 27.07 R Gemini-N 3-sigma Note. — Optical/NIR observations of GRB 090429B. Mag- nitudes are quoted in the AB system, and corrected for the ex- pected Galactic extinction along the line of sight (E(B −V )= 0.01). 16 2.3.1 Spectral Energy Distribution First of all, with our optical upper limits and infrared detections we can place some constraints on the redshift of GRB 090429B. If the lack of optical detection is due to a Lyman-break we can infer a lower limit of z & 6.3 (z′-band). Similarly, the spectral break observed between the J and H bands suggests a redshift of 8.0 βJH = 3.9 ± 1.2 (2σ) and very difficult to interpret in a standard synchrotron scenario, whether by comparison to the H − K color or by comparison to the simultaneous X-ray afterglow data. Conservatively speaking, we observe βJH & 2.7, with a steepening (or break) in the spectral slope of ∆βJHK & 1.71 (2σ). This spectral break suggests that the high-redshift Lyman-break or extinction are suppressing the J-band flux below the power-law extrapolation from H − K, and at the same time, providing even greater suppression of the afterglow flux in the optical (i′z′; see Fig 2.1). Overall, the similarity with GRB 090423 is striking; although the observed break is not as sharp this is easily accommodated in the extreme-redshift scenario with a Lyman-break midway though the J bandpass – by contrast, the Lyman-break for GRB 090423 at z = 8.2 was placed perfectly between the Y and J filters. The absence of a host galaxy to a 3σ limit of RAB > 27 mag is also relevant in this context, since it corresponds to an absolute magnitude of MR = −17.22 mag at z = 1. Such a value is among the lowest of GRB host galaxies (Berger 2009). Comparing our results with the dark bursts host search by Perley et al. (2009) we note that their 14 detected host galaxies have R magnitude between 23.6 and 25.5. A recent survey of GRB host galaxies (Jakobsson et al. 2009; Fynbo et al. 2009) shows that only 4% of 186 SwiftGRBs in the considered sample have R> 27 mag and redshift 0 2.3.2 Photometric Redshift Analysis Motivated by the extreme-redshift hypothesis for GRB 090429B, we construct an observed SED based on our Gemini-N observations at t ∼ 104 s, and fit it with the 17 latest version of hyper-z code (for a detailed explanation of the fitting procedure see Bolzonella et al. 2000). In performing this analysis we exclude the X-ray data since it is possible the high-energy and infrared emission originate from different components (§2.3.3). In order to analyze this dataset we provided customized hyper-z inputs as follows: First, we introduced the Gemini-NIRI filter transmission curves into the code database; second, for the template spectral energy distribution we assume a simple power-law, as appropriate for afterglow synchrotron emission. We tested a range of different values for the intrinsic spectral index, ranging from β = 0.5 up to β = 1.3. This range is consistent with the observed range of extinction- corrected afterglow spectral slopes (Schady et al. 2007, 2009; Kann et al. 2007), and consistent as well with the spectral slope observed in the X-ray, βX = 1.12 ± 0.12, while accounting (at the low end) for a possible cooling break (δβ = 0.5) between the X-ray and optical/NIR regimes. We work with an SMC extinction law for the high-redshift host galaxy as preferred from other Swift-era GRB studies (Schady et al. 2007, 2009; Kann et al. 2007). The redshift is a free parameter in the range 0 ≤ z ≤ 10, and the host galaxy extinction is varied from zero up to AV = 5 mag. We collapse the three-dimensional grid along the β (spectral slope) axis, choosing 2 the best χ value for any (z, AV ). In addition to a standard (i.e., frequentist) flat prior distribution in z and AV , we have also applied a “Swift” prior distribution to our analysis. For the “Swift” priors, the z distribution is provided by the observed distribution of Swift absorption redshifts, and the AV distribution by the predicted intrinsic (unbiased) distribution from Schady et al. (2009). Both datasets are approximated with truncated Gaussian distributions, in log(1 + z) and log(AV + offset), respectively, for the purposes of continuity and extensibility in our analysis. The best overall fit to our dataset, using flat priors, corresponds to a photometric +0.25 2 redshift zphot = 9.5−0.60 (90%-confidence) when examined as a χ function of a single variable. This solution implies negligible extinction at the host (Fig. 2.3), as observed for GRB 090423 and, indeed, for a majority of Swift GRBs across all redshifts (Kann et al. 2007). A secondary local minimum at zphot < 0.5, allowed at just greater than 90% confi- dence, can be excluded on the basis of our deep host galaxy limit; however, examination of χ2 contours in the two-dimensional space of redshift and extinction reveals the pres- ence of other solutions within the 90%-confidence contour: specifically, solutions with 18 zphot ≈ 5 and AV ≈ 0.8 mag, and with zphot ≈ 2 and AV = 2.0±0.4 mag. More generally, within the two-dimensional space any redshift over 1.5 < zphot < 10 can be accommo- dated within the 99%-confidence contour, given an appropriate choice of extinction. As expected, lower-redshift solutions require greater levels of host galaxy extinction to yield the observed suppression and reddening of the assumed power-law continuum. These constraints are noticeably weakened by application of the Swift priors, which suggest, on the basis of our experience with Swift bursts, that we should expect low extinction and redshifts in the range 1 ∼< zphot ∼< 5. With the Swift priors, a one- dimensional analysis yields three 90%-confidence intervals: zphot = 2.25 ± 0.30, zphot = +1.8 +0.3 4.8−0.8, and zphot = 9.5−0.5. The 90%-confidence contour in z and AV encompasses the full range of redshifts, 1.5 2.3.3 Theoretical Considerations The X-ray afterglow after the peak, at Tp, can be interpreted as the synchrotron emission from the forward shock that propagates into the external medium with a uni- form density profile (e.g. Sari et al. 1998; Meszaros & Rees 1997). The spectral and temporal indices are consistent with the case of νm,νc <νX or νm <νX <νc, where νm and νc are the characteristic synchrotron frequency and the cooling frequency, respec- tively. The flux peak can be explained to be the deceleration of the ejecta (Sari & Piran 1999; Xue et al. 2009). The early time observed flux rising is shallower than predicted 2 3 (t for νm,νc <νX or t for νm <νX <νc), but this discrepancy can be reconciled with a possible contamination of the X-ray tail of the prompt emission. In this model, we find it difficult to explain the NIR light curve simultaneously. 4 5 The shallow decay of the NIR flux at 10 s . t . 10 s might correspond to νc < 5 14 νIR < νm, but νm(10 s) > 3 × 10 Hz is quite unlikely for the initial kinetic energy of the ejecta E . 1052 erg and a source redshift z . 9. Then we would expect that the X-ray and NIR fluxes decay in a similar behavior, i.e., αX − αIR = 0 or 0.25, but +0.13 αX − αIR = 0.72−0.12 > 0.25 at the 4σ confidence level. This suggests that the X-ray and IR emission originate from two different components. Also, if the X-ray spectrum corresponds to the afterglow unabsorbed spectrum and there is no temporal variation in the spectral index (i.e. νc,νm < νIR < νX ), then we would expect a single power-law over the observed frequency range with β ∼ p/2, (p is the usual electron energy distribution power index). We notice that the infrared 19 spectral slope (βH−K = 0.85 ± 0.20) and the X-ray spectral index (βX = 1.12 ± 0.23) are marginally consistent, although as Fig. 2.5 shows, they are not likely to represent two aspects of a single underlying power-law continuum. A fast cooling scenario (νc < νIR < νm < νX ) would imply a shallower infrared spectral slope and an intrinsic extinction of AV . 1. This would be unusual for the time of our observation, requiring a low value of the afterglow kinetic energy. We cannot neglect the fact that some extinction, due to the GRB host, is probably present and absorbs the flux at bluer wavelengths (Perley et al. 2009), but the infrared intrinsic spectral energy distribution would be shifted at higher values, implying a flat- tening in the H-K spectral slope. In this case the cooling frequency νc or the typical frequency νm may fall in between νIR and νX , again unusual for the time of our obser- vation compared to other GRBs. Nevertheless, if this is the case, we expect β = 1/2 for νc <ν<νm or β = (p − 1)/2 for νm <ν<νc. 2.4 Conclusions We have presented our discovery and multi-wavelength observations of GRB 090429B and its NIR afterglow, incorporating data from the Swift satellite and ground-based fa- cilities including GROND and the Gemini-North telescope. Initial observations, beginning only 14 min after trigger, did not show a detection in the GROND optical or infrared bands. At t+3 hr we observed with the Gemini-North telescope using the GMOS and NIRI instruments, identifying an optically-extinguished, bright near-infrared afterglow with J − H and H − K colors suggesting an extreme redshift, 8.0 Fig. 2.1 top panel: Light curves from different instruments; our full photometric dataset and additional data from the literature are shown. An inset figure shows the BAT light curve for this event. bottom panel: X-ray hardness ratio as a function of time. A change in hardness ratio is observed after the peak at t+600 s, consistent with the hypothesis that the X-ray peak indicates the deceleration time for the burst ejecta (see §2.3 for details). 22 J H XRT April 29.33 April 29.33 K K April 29.33 April 30.33 z' i' R − Host April 29.33 April 29.33 May 13.25 Fig. 2.2 Gemini-N imaging. Top panels: J and H-band images taken on April 29. Bottom panels: K-band images taken on April 29 and 30. A clear fading of the afterglow in infrared is visible. The 2” X-ray (XRT) localization for the afterglow is shown. 23 Flat Fig. 2.3 Confidence (δχ2) regions for our hyper-z photometric redshift fits to the Gemini- N optical and NIR photometric dataset for GRB 090429B. In each panel, the upper plot shows two-dimensional joint confidence contours in redshift z and host extinction AV , while the lower plot shows the one-dimensional projection, δχ2(z). At each grid point, the analysis uses the minimum chi2 value across the allowed β range, 0.5 ≤ β ≤ 1.3. This plot shows the confidence regions when using “Flat” (frequentist) priors. Swift Fig. 2.4 Confidence (δχ2) regions for our hyper-z photometric redshift fits to the Gemini- N optical and NIR photometric dataset for GRB 090429B. The upper plot shows two- dimensional joint confidence contours in redshift z and host extinction AV , while the lower plot shows the one-dimensional projection, δχ2(z). At each grid point, the analysis uses the minimum chi2 value across the allowed β range, 0.5 ≤ β ≤ 1.3. This plot shows the confidence regions when using “Swift” priors; see text for details. 24 4 Fig. 2.5 IR to X-ray spectral energy distribution at T0 + 10 sec. The green solid line represent the IR spectral slope derived from the K and H detections. The blue dot-dashed is the spectral fit derived from the XRT averaged spectrum. Finally, in red (dashed) we present our SED modeling fitting curve. Parameters values for the latter two spectral energy distribution are reported. i′ and z′ upper limits are shown as black triangles. 25 Chapter 3 Testing the Possible Intrinsic Origin of the Excess of Very Strong MgII Absorbers Along GRB Lines-of-Sight This chapter discusses the extraordinary opportunity offered by high-resolution spectroscopy of GRB afterglows in order to test the GRB environment and, in particular, the presence of intervening systems along their lines of sight. An interesting class of absorbers is the one that presents very strong Mg ii absorption features. We performed, for the first time, kinematic tests in order to investigate the nature of these absorbers along GRBs lines of sight in comparison with similar absorbers along quasars. This work has been published in Cucchiara et al. (2009c, ApJ, 697, 345). 3.1 Introduction The startling discovery of Prochter et al. (2006a) that the frequency of very strong (Wr(2796) > 1 A)˚ Mg ii absorbers along gamma-ray burst (GRB) lines of sight ([dN/dz]GRB = 0.90) is more than three times the frequency along quasar lines of sight ([dN/dz]QSO = 0.24), over similar redshift ranges, has yet to be understood. In par- ticular, explanations appealing to dust anti-bias in quasar samples, partial covering of the quasar sources, and gravitational-lensing amplification of the GRBs have all been carefully examined and found wanting. We therefore reconsider the possibility that the excess of very strong Mg ii absorbers toward GRBs is intrinsic either to the GRBs themselves or to their immediate environment, and associated with bulk outflows with velocities as large as vmax ∼ 0.3c. In order to examine this hypothesis, we accumulate a sample of 27 Wr(2796) > 1 A˚ absorption systems found toward 81 quasars, and com- pare their properties to those of 9 Wr(2796) > 1 A˚ absorption systems found toward 6 GRBs; all systems have been observed at high spectral resolution (R = 45, 000) using the Ultraviolet and Visual Echelle Spectrograph on the Very Large Telescope. We make multiple comparisons of the absorber properties across the two populations, testing for differences in metallicity, ionization state, abundance patterns, dust abundance, kine- matics, and phase structure. We find no significant differences between the two absorber 26 populations using any of these metrics, implying that, if the excess absorbers toward GRB lines of sight are indeed intrinsic, they must be produced in a process which has strong similarities to the processes yielding strong Mg ii systems in association with intervening galaxies. Although this may seem a priori unlikely, given the high outflow velocities required for any intrinsic model, we note that the same conclusion was reached, recently, with respect to the narrow absorption line systems seen in some quasars. 3.2 Motivation for This Study Intervening metal absorption line systems have been routinely observed along the lines of sight to cosmological gamma-ray bursts (GRB) since the first optical/UV spec- trum of a GRB afterglow was obtained in 1997 (Metzger et al. 1997). In particular, Mg ii λλ2796, 2803 resonance doublet absorption, observable from the ground over red- shifts 0.4 . z . 2.2, has been detected in many GRB afterglow spectra, at redshifts well separated from that of the highest-redshift absorber in the system, which is typically associated with the GRB host galaxy. Strong intervening Mg ii absorbers are a familiar subject of quasar absorption- line studies, and have been used for decades in studies of intermediate-redshift galaxies and their environments. Indeed, it is now known that the presence of absorption with Mg ii λ2796 rest frame equivalent width Wr(2796) > 0.3 A˚ is commonly (∼ 75% of all cases) associated with the presence of a nearby (within 60 kpc projected distance) ∼ 0.1–5 L⋆ galaxy along the line-of-sight (Kacprzak et al. 2008). The detailed physical picture of the strong Mg ii absorbers continues to be elab- orated. Among the strongest (Wr(2796) > 1 A)˚ Mg ii absorbers at z < 1.65, approxi- mately 80% are also damped Lyman-alpha systems (DLAs; Rao et al. 2006). Imaging of the quasar fields for a subset of even-stronger (Wr(2796) > 2.7 A)˚ Mg ii absorbers at low redshift (0.42 Wr(2796) > 1 A˚ absorbers, that these structures are related to superwinds, rather than to large-scale gas infalls in galaxy halos. Indeed, Nestor et al. (2005) find a rapid decline in the incidence of Wr(2796) > 2 A˚ systems with decreasing redshift, consistent with the expectations from superwinds since these are thought to increase in concert with the global star formation rate. GRBs and their afterglows are promising probes of the high-redshift Universe due to their brightness and observed redshift distribution, now extending beyond z = 6.7. High-resolution spectroscopy of the brightest afterglows has been used to study GRB host galaxies, and the subgalactic environs of GRBs, up to z = 6. With respect to intervening absorption systems, one would naively expect GRB lines of sight to be equivalent to quasar lines of sight, including strong Mg ii absorbers. However, for very strong (Wr(2796) > 1 A)˚ Mg ii absorbers, a puzzling difference in the redshift path density, dN/dz, between GRB and QSO sightlines has been discovered. The frequency of these absorbers along GRB lines of sight is more than three times larger ([dN/dz]GRB = 0.90) than the frequency along quasar sightlines ([dN/dz]QSO = 0.24) covering a similar range of redshifts (Prochter et al. 2006b). Since the discovery of this factor of ≈ 3 discrepancy, several hypotheses have been advanced to explain it: (1) The number of Mg ii systems along quasar lines of sight has been suppressed, due to a bias within quasar samples towards brighter objects lacking, e.g., dusty foreground systems along the line of sight; (2) The relative beam sizes of quasar and GRB afterglow emitting regions lead to partial covering in quasar spectra, and increased numbers of strong absorption systems in GRB spectra; (3) Gravitational lensing by the mass concentrations associated with strong absorption systems magnifies the GRB and its afterglow, increasing the probability of detection and spectroscopic observation; (4) A dominant number of the strong absorption systems in GRB spectra are physically associated with the GRB environment, the GRB itself, or both. We will now briefly review the status of each of these hypotheses as a way of motivating the present study, which focuses on the fourth hypothesis, that the excess of very strong Mg ii absorption systems in GRB afterglow spectra is intrinsic to the GRB or its environment. For any of the other hypotheses to provide an explanation of the full effect, the resulting biases would have to be quite large. The intrinsic hypothesis, by contrast, is required to produce on average roughly one intrinsic, high-velocity absorption system per GRB. 3.2.1 Bias Due to Dust If the very strongest intervening Mg ii absorbers tended to arise in the dustiest environments they would diminish the observed magnitude of a background quasar. In 28 such a way, optical magnitude-limited quasar absorption-line surveys might be biased toward the brighter quasars that do not have the strongest Mg ii systems in the fore- ground. Although gamma ray bursts would suffer the same dust bias effect, this effect is less important in determining whether a high signal-to-noise, high-resolution spectrum can be obtained than is the speed at which follow-up observations are possible. Thus we would expect that gamma ray bursts that have very strong, dusty Mg ii absorbers in the foreground would still be present in a sample of high-resolution GRB optical spectra. GRB spectra should therefore have more very strong Mg ii absorbers on average than do quasar spectra. The main problem with this explanation for the excess of Wr(2796) > 1 A˚ ab- sorbers in GRB spectra is that the magnitude of the effect is not nearly large enough to explain a factor of more than three excess. M´enard et al. (2008) found that . 2 % of quasars are absent from optical surveys due to reddening and extinction from strong Mg ii systems (1 More directly, Ellison et al. (2006) estimated the extinction due to a Wr(2796) = 1.87 A˚ intervening system along the line of sight of GRB060418 at redshift z = 1.106. The amount of extinction derived is E(B − V) = 0.08 mag, which is not sufficient to obscure sufficient quasars to explain the observed discrepancy in dN/dz. 3.2.2 Partial Beam Coverage Explanations An interesting explanation has been proposed by Frank et al. (2007), assuming that GRB emitting regions are generally smaller than those of quasars. In order to avoid obvious signatures of partial covering in the case of quasar strong Mg ii absorbers, Frank et al. (2007) derived a cloud density profile with a constant density core surrounded by a power-law density profile. The density profile sampled by the beam depends on the size of the beam relative to the core size of the absorber and on the impact parameter. Small beams that pass thorough only the center of the cloud sample much higher column densities than beams larger than the core for which high optical depth absorption is diluted. One prediction of this scenario, because the GRB beam size changes in time, is variability in the strength and the structure of their absorption lines. Such variability has not yet been seen (Th¨one et al. 2008), although it was once suggested (Hao et al. 2007) in the case of GRB 060206. More importantly, Pontzen et al. (2007) point out that signatures of partial covering should still be seen in high-resolution spectra unless the cloud density profiles are fine-tuned to an unreasonable degree. They also show that it is unlikely that an excess of the needed magnitude could arise from this effect. Finally, there is no way to set up the clouds producing the numerous distinct absorption components seen in strong Mg ii absorbers such that the line of sight passes through the inner high-density core of all of the clouds. 29 3.2.3 Lensing Amplification of GRB Beam-size Following an analogy with BL Lac objects as studied by Stocke & Rector (1997), the idea that a GRB line of sight intersects more absorbers because the emitting region is microlensed by a gravitational potential has been considered (e.g. Loeb & Perna (1998); Garnavich et al. (2000); Baltz & Hui (2005); Hirose et al. (2006)). In particular, considering binary lensing and double magnification bias, Baltz & Hui (2005) estimate that there is a 60% chance that a given GRB is microlensed. On the other hand, using a sample from Sloan Digital Sky Survey (SDSS) quasars, the magnification factor has been calculated to be µ . 1.10 (M´enard et al. 2008). This value is not sufficient to lead to the observed excess of very strong Mg ii absorbers towards GRBs. Furthermore, if the lensing effect is significant for GRBs, we would expect to see multiple images, “repeating” GRBs, and “bumps” in GRB optical light curves (Loeb & Perna 1998), neither of which are commonly seen. At this time, it is important to consider whether a more significant magnification bias could result from a true source number density function that is steep at its faint end. Nardini et al. (2006, 2008) have argued that this is the case, with α< −3 for GRBs with more than one MgII absorber (where α is the slope of the power-law assumed to fit the GRB luminosity function); however, a statistical sample is needed in order to test this assumption which is based only on energetic arguments. Zitouni et al. (2008) found that using an internal shock model the faint end slope of the GRB optical luminosity function can be fitted by a power law index with α ≈−0.6, as already noticed by Porciani et al. (2007). 3.2.4 An Origin Intrinsic to GRBs or Their Environments In view of the dramatic excess of Wr(2796) > 1 A˚ absorbers toward GRBs and the problems with the explanations discussed above, we should consider further the idea that the GRBs themselves or their immediate environments are responsible for material observed as Mg ii absorption. Since the observed excess is apparent over a substantial redshift range, the material would have to be accelerated to at least 0.2–0.3c relative to the GRB host galaxy. At first glance it might seem implausible that such high velocity ejection could be consistent with such narrow lines as are observed as components in the strong Mg ii absorbers (Prochter et al. 2006b). However, it is important to note that similar high velocity narrow absorption lines (NALs) are known to be ejected in the accretion disk winds of quasars. (Misawa et al. 2007a; Ganguly & Brotherton 2008; 30 Rodriguez Hidalgo et al. 2007). This phenomena also has yet to be explained by accretion disk wind models. Clearly it is possible for dense, coherent clouds of gas to hold together, even in mildly-relativistic outflows, at least in this case. So it seems possible that the same could be the case for GRBs. First, consider mechanisms involving the GRB itself. A typical GRB forward shock strongly ionizes the interstellar medium within 1016 to 1017 cm of the central engine. Thus it would be hard for Mg ii to survive within this distance. The shock front could accelerate a cloud of material to 0.2–0.3c out to a distance of ∼ 1 pc, however this would lead to an increase in temperature and Mg ii could not be produced at T ≫ 104 K. It would be necessary we would have to rely on the existence of a dense, metal-rich cloud which is not penetrated by the relativistic shock and which cools faster than the surrounding material. As proposed by van Marle et al. (2007), a Wolf-Rayet wind can actually bring dense material to the required distance. Then, this bubble can be accelerated by the GRB itself, producing the observed absorption features. The intrinsic “lower redshift” systems (at least some of them) would be the signature of the wind accelerated at mildly (. 0.2c) relativistic speed by the forward shock. Since GRBs are generally located in star forming regions, there are some less intrinsic local origins that can be considered. For example, it has been proposed that some of the strong Mg ii absorbers observed along the GRB lines-of-sight are associated with supernova remnants present at the time the GRB occurred. In fact, Wang et al. (2003) suggested that GRB 021004 was actually located in a region where the ISM was metal-rich due to high velocity ejecta from a hydrogen-rich supernova that exploded a few months before the GRB. Such ideas seem reasonable, but we must also understand how the host galaxy absorption fits into this picture. It is apparently coming from > 1 kpc, the host galaxy at large rather than the immediate GRB environment, yet it is clearly affected by the GRB radiation field at least in cases where metastable lines are observed. These systems are not counted in the excess of very strong Mg ii absorbers found by Prochter et al. (2006b), but it is still important to consider how the absorption signature from high velocity material more local to the GRB might compare to that of the host galaxy. 3.2.5 Motivation for This Study We aim to consider the similarities and differences between the absorption profiles of: 1) very strong Mg ii absorbers found in quasar spectra; 2) very strong Mg ii absorbers found in GRB spectra, which are a combination of the same objects found in quasar 31 spectra and a separate population yet to be understood; 3) GRB host galaxy absorption. Naively we might expect that the absorbers responsible for the factor of three excess of very strong Mg ii absorbers along GRB sightlines should differ in some way from the usual quasar absorption line systems. We thus compare quasar absorption line systems taken from the VLT/UVES archive with systems seen in absorption in GRB spectra obtained with the same instrument and configuration. We consider the kinematics of the Mg ii profiles, the ratios of Mg ii equivalent widths to those of other transitions, including dust tracers, the relationship between high and low ionization transitions, and the possible presence of metastable lines. In the next section, we describe the VLT/UVES datasets and our analysis meth- ods. In §3.4 we present the results of our comparisons between the very strong Mg ii absorbers seen toward quasars and GRBs. In §3.5 we discuss the implications of our result that there are no apparent differences between these populations, in particular, considering the implications for the hypothesis of an intrinsic origin for a majority of the GRB absorbers. §3.6 summarizes our conclusions. 3.3 Dataset and analysis Our GRB dataset consists of 6 GRBs observed with the Ultraviolet and Visible Echelle Spectrograph (UVES) mounted on VLT. These include all high resolution UVES spectra that were accessible to the public before August 2008. The details of the observa- tions, including the time that the observations were obtained, the wavelength coverage, and the exposure time, are listed in Table 3.1. The quasar dataset included 81 QSOs, also obtained with UVES/VLT. These quasars, listed in Table 3.2 of Narayanan et al. (2007), are those for which data were available before June 2006. They were reduced using the MIDAS pipeline as described in Narayanan et al. (2007). All observations of a given quasar were combined with S/N weighting, after scaling by the median ratio of the number of counts in the best exposure to the counts in the given exposure. The GRB spectra were reduced in a similar manner, also using the standard UVES/VLT tools under the MIDAS environment. Because of possible variability, the different exposures of the same GRB were not combined. A conversion to a heliocentric vacuum scale was applied to the final spectra. Continuum fitting and normalization was performed using the IRAF SFIT procedure 1, by dividing the spectrum into 3000 1IRAF is distributed by the National Optical Astronomy Observatories, which are operated by AURA, Inc., under cooperative agreement with NSF 32 Table 3.1. Observation log GRB UT start Exposure (min) coverage(nm) 021004 2002, April 5.22 30 376-498;670-852; 866-1043 021004 2002, April 5.23 30 302-392;473-580;576-680 021004 2002, April 5.23 10 565-660 021004 2002, April 5.24 60 452-560;568-665 021004 2002, April 5.25 60 302-392;473-580;576-680 021004 2002, April 5.28 60 376-498;670-852;866-1043 050730 2005, July 31.01 3 302-392;473-580;576-680 050730 2005, July 31.04 5 373-505;665-854;864-1008 050820 2005, August 20.29 30 328-452;452-560;568-665 050820 2005, August 20.32 30 328-452;452-560;568-665 050820 2005, August 20.35 40 373-505;665-854;864-1008 050922C 2005, September 22.98 50 302-392;473-580;576-680 050922C 2005, September 23.02 50 373-505;665-854;864-1008 060418 2006, April 14.13 3 328-452;452-560;568-665 060418 2006, April 14.14 5 328-452;452-560;568-665 060418 2006, April 14.15 10 328-452;452-560;568-665 060418 2006, April 14.15 20 328-452;452-560;568-665 060418 2006, April 14.17 40 328-452;452-560;568-665 060418 2006, April 14.20 80 376-498;670-852;866-1043 060607 2006, June 7.22 3 328-452;452-560;568-665 060607 2002, June 7.23 5 328-452;452-560;568-665 060607 2002, June 7.23 10 328-452;452-560;568-665 060607 2002, June 7.24 20 328-452;452-560;568-665 060607 2002, June 7.25 40 328-452;452-560;568-6650 060607 2002, June 7.28 80 376-498;670-852;866-1043 060607 2002, June 7.28 42 376-498;670-852;866-1043 Note. — 33 pixels segments and fitting each segment separately. The spectra were normalized by the resulting continuum fit. The signal-to-noise ratio ratio(S/N) of the QSO spectra ranges from ∼ 20 − 100 per pixel over most of the wavelength coverage. The GRB spectra are somewhat noisier due to the limitations caused by the need for rapid followup before the GRB fades. We searched both the QSO and GRB spectra for Mg ii doublets in regions redward of the Lyα forest. Formally, we applied a 5σ detection limit to absorption components in the Mg ii λ2796 line. Since our focus is on very strong Mg ii absorbers (with equivalent width W (λ2796) > 1 A),˚ we are easily detecting all systems of interest. For detected systems we then searched the expected locations of other ions, including Lyα , Lyβ, Lyγ, Mg i λ2853, Oi λ1302, Fe ii λ2374, 2383, 2587, and 2600, Si ii λ1260, C ii λ1335, Al iii λλ1855, 1863, Si iv λλ1394, 1403, C iv λλ1548, 1550, Nv λλ1238, 1242. Several of these transitions were only covered for the very highest redshifts in our sample, and thus could not be used for a statistical comparison. We also considered dust tracers such as Zn ii λ2026,2063, Cr ii λ2056,2062,2066, Ni ii λ1710,1752, and Mnii λ2577,2594,2606. For whichever of these transitions were covered we measured the equivalent widths or 3σ equivalent width limits. When blends with transitions from systems at other redshifts were identified we published the measured equivalent width as an upper limit. 3.4 Results In our search of the 6 GRB spectra listed in Table 3.2, we found 9 Wr(2796) > 1 A˚ Mg ii doublets. The redshift pathlength for our GRB search was 9.9, giving a dN/dz = 0.90 ± 0.3, consistent with the value from Prochter et al. (2006b), obtained from 14 GRBs spectra. Similarly, we identified 27 Wr(2796) > 1 A˚ Mg ii doublets over a redshift pathlength of 77.3 towards 81 QSOs. We derive dN/dz = 0.35 ± 0.07 for very strong Mg ii absorbers observed toward quasars, which is also consistent with the much larger Sloan sample of Nestor et al. (2005). System plots for the GRB absorbers, including various transitions that provide useful constraints, are shown in velocity space in Appenix A. Similar plots for the quasar absorbers are given in Appendix B. Basic information about the absorbers, for both GRBs and quasars, is given in Table 3.2. Rest frame equivalent widths of selected transitions that were detected in one or more systems are given in Tables 3.3, 3.4 and 3.5, along with equivalent width limits in cases where the transition was covered but not detected. The excess of strong Mg ii absorbers along GRB lines of sight, as compared to quasars, applies only for Wr(2796) > 1 A,˚ though there appears to be no greater an effect 34 for even stronger systems (Prochter et al. 2006b). For our VLT/UVES samples we plot Wr(2796) as a function of redshift in Fig. 3.1d. The equivalent width distributions of the GRB and quasar Mg ii absorbers were compared (including only those with Wr(2796) > 1 A)˚ using a Kolmogornov-Smirnov test (K-S test) and it was found that they are consistent of being drawn from the same distribution (p=12%). It is also important to consider whether the redshift distributions of the two samples are the same, since Wr(2796) > 2 A˚ Mg ii absorbers are known to evolve in the sense that they are less common at low redshift. A redshift difference could therefore lead to a difference in the equivalent width distributions. We find, however, that a K-S test shows that the redshifts of the GRB and quasar Mg ii absorber samples are consistent of being drawn from the same distribution (p=30%). A difference in the redshift distributions between the two samples could also be indicative of a concentration of GRB Mg ii absorbers at relatively small ejection velocities. We therefore also show in Fig. 3.1c the cumulative distributions of relative velocities for the two samples, normalized to 1. We are converting the quasar Mg ii absorber redshifts to relative velocities only to facilitate the comparison. There is no significant difference between these distributions, though we note that only one of the GRB absorbers has an implied ejection velocity greater than 0.4c. In this section we describe various comparisons of the kinematics and the absorp- tion in numerous chemical transitions for the 9 very strong Mg ii absorbers seen towards GRBs and the 27 seen towards quasars. These detailed comparisons are facilitated by the high resolution of the UVES/VLT spectra. Fig. 3.2 shows the absorption profiles of the Mg ii λ2796 line for the 9 GRB Mg ii systems in the left panel and a random sampling of 9 quasar systems in the right panel. At face value, the profiles are similar in terms of their equivalent widths and kinematic structures. We will now examine this quantitatively, considering also other elements and ions and their relation to the Mg ii absorption properties. We also note that there is no evidence for partial covering for the Mg ii absorbers, neither for the quasar or GRB cases. Most lines/components are saturated over a finite extent in wavelength. 35 Fig. 3.1 Observed properties of the GRB (solid black histogram) and quasar (dashed red his- togram) strong Mg ii absorption systems: (a) Distribution in Mg ii(2796) equivalent width; (b) Distribution in redshift; (c) Distribution in ∆v/c, the relative or “ejection” velocity of the ab- sorber relative to the target object; (d) Two-dimensional distribution in Mg ii(2796) rest frame equivalent width and redshift for GRB (black square) and quasar (red diamond) absorption sys- tems. All fundamental properties of the two samples exhibit a similar distribution over a similar range. 36 1 1 0 0 1 1 0 0 1 1 0 0 1 1 0 0 1 1 0 0 1 1 0 0 1 1 0 0 1 1 0 0 1 1 0 0 -600 -400 -200 0 200 400 600 -600 -400 -200 0 200 400 600 Velocity [km/s] Fig. 3.2 Mg ii λ2976 line profiles for all GRB (left) and for nine selected quasar (right) intervening systems. Each panel provides the name of the target object and the redshift of the intervening absorption system. No obvious systematic difference has been found in the absorption profiles of the two samples; in particular, both samples show saturated absorption features and multiple components with similar ranges in velocity. Quantitative metrics derived from these kinematic profiles are compared in §3.4.1. 37 Table 3.2. Systems identified Object zobj zsys ∆v/c GRB060418 (1) 1.49 0.6021 0.41 GRB060418 (2) 1.49 0.6554 0.38 GRB060418 (3) 1.49 1.1066 0.16 GRB021004 (1) 2.328 1.3800 0.32 GRB021004 (2) 2.328 1.6015 0.24 GRB050820 (1) 2.612 0.6910 0.64 GRB050820 (2) 2.612 1.4280 0.37 GRB060607 3.08 1.7996 0.36 GRB050730 3.967 2.2527 0.40 3C336 (1) 0.9273 0.6561 0.1505 3C336 (2) 0.9273 0.8913 0.0188 Q1229-021 1.038 0.3951 0.3618 Q1127-145 1.18 0.3127 0.4677 Q1629+120 1.795 0.9002 0.3677 Q0328-272 1.816 1.1228 0.2753 Q1331+170 2.084 1.7766 0.1046 HE1341-1020 2.134 1.2767 0.3091 PKS0237-23 (1) 2.224 1.3651 0.3003 PKS0237-23 (2) 2.224 1.6723 0.1855 Q0551-3637 2.318 1.9609 0.1134 Q0109-3518 2.35 1.3496 0.3405 HE1122-1648 2.4 0.6822 0.6066 HE2217 2.406 1.6922 0.2309 Q0453-423 (1) 2.66 0.7261 0.6361 Q0453-423 (2) 2.66 1.1498 0.3642 Q0100+130 (1) 2.681 0.2779 0.7848 Q0100+130 (2) 2.681 1.7969 0.2679 Q0002 (1) 2.76 0.8366 0.6147 Q0002 (2) 2.76 2.3019 0.1292 Q1151+068 2.762 1.8191 0.2808 Q0112+300 2.81 1.2452 0.4800 Q0130-4021 3.03 0.9315 0.6264 HE0940 3.083 1.7891 0.3636 CTQ0298 3.37 1.0387 0.6427 Q1418-064 3.689 1.4578 0.5689 Q1621-0042 3.7 1.1335 0.6583 Note. — List of GRB and QSO systems iden- tified, in increasing object redshift order, with rest frame EW(λ2796 > 1 A).˚ The redshift of the object (zobj) is the quasar or GRB redshift es- timated by emission lines (in the QSO case) or multiple absorption features associated with the host galaxy redshift (for the GRB). The last col- umn reports the redshift difference between the previous 2 columns assuming that the absorbing system is actually moving at a positive velocity respect the quasar or the GRB host. 38 Table 3.3 Equivalent widths 39 Table 3.4 Equivalent widths 40 Table 3.5 Equivalent widths 41 3.4.1 Mg ii Kinematics We make quantitative comparisons of the Mg ii λ2796 profiles of the quasar and GRB absorbers using some of the same statistics that Mshar et al. (2007) used to consider evolution of strong Mg ii absorbers. In addition to the equivalent widths, we describe the Mg ii λ2796 absorption profiles by several kinematic indicators, namely, the full velocity range, ∆V , kinematic spread, ωv, and D-index, D. The full velocity range of a system ∆V is the difference between the minimum and maximum velocities at which absorption was detected. As defined in detail in Appendix A of Churchill & Vogt (2001), the kinematic spread is the second velocity moment of the apparent optical depth. This kinematic indicator is particularly sensitive to weak components at high velocities from the central absorption. Finally, the D-index was defined by Ellison et al. (2006) as −1 D = 1000 ∗ (Wr(2796)[A])˚ /∆V [km s ] in order to indicate a distinction between DLAs and lower column-density absorbers. D gives an indication of the fraction of the profile that has saturated absorption in Mg ii λ2796. Figure 3.3 presents the three kinematic indicators, D, ωv, and ∆V , for quasar and GRB Mg ii λ2796 profiles as a function of redshift. There is no evoluation apparent in ωv or ∆V for the quasar population. There is a suggestion of an increase in D with decreasing z, but it is only significant at the 4% level according to the Kendall τ rank order test (Ferguson et al. 1995). If this trend is real it would indicate that low-redshift, very-strong Mg ii absorbers are more likely to have black absorption across their profiles. Figure 3.3 also shows the dependence of D on Wr(2796). There is no significant correlation between these two quantities. The main purpose of Figure 3.3, for the purpose of this study, is to compare the quasar and GRB Mg ii λ2796 profiles. We found no significant differences between the distributions of D, ωv, or ∆V between the two populations as evaluated using a Kolmogornov-Smirnov (K-S) test. In particular, a K-S test comparing the D-values between GRBs and QSOs are consistent of being drawn from the same distribution (p=13.5%). 3.4.2 Equivalent Width Comparisons Fig. 3.4 plots the rest-frame equivalent widths of selected transitions versus those of Mg ii λ2796. These particular transitions were selected for display because they were accessible to analysis in our spectra and/or because they represent important tracers of the physical conditions in the gas. Rest frame equivalent widths and equivalent width 42 Fig. 3.3 Kinematic quantities describing the GRB (open symbols) and quasar (filled symbols) strong Mg ii absorbers in our study: (a) D-index (Ellison et al. 2006) versus redshift; (b) D-index versus Mg ii(2796) rest frame equivalent width; (c) Velocity spread ωv versus redshift; and (d) Total velocity coverage ∆v versus redshift. There is no significant difference in either the range of the properties, or the distribution of properties of individual systems within the overall range, between the two samples, as confirmed in each case by a K-S test. 43 limits for other transitions are given in Tables 3.3, 3.4 and 3.5. The basic result from Fig. 3.4 is that there are no significant differences between the GRB and quasar samples. The Mg i λ2853, Fe ii λ2600, and Fe ii λ2374 transitions are almost always covered and detected, but the values for GRB absorbers span the range of values for quasar absorbers of similar Wr(2796). The detection of Mg i absorption implies that the absorbing gas cannot be located within 50 pc of the GRB afterglow (Prochaska et al. 2006). We therefore note that several of the GRB absorbers are among the lowest Mg i equivalent widths for our sample, though there is no statistically significant difference in the overall distributions. The dust tracer Mnii λ2577 is covered in many cases, though often they are not detected. Again, the GRB and quasar absorbers have a similar distribution in Fig. 3.4. Dust will be discussed further in § 3.4.4. The higher-ionization line Al iii λ1855 is only covered for the highest-redshift quasars and GRBs but it can be noted that the few GRB absorbers do not deviate significantly from the quasar absorbers in their high-ionization content. Details of the high-ionization kinematics, as traced by CIV, will be discussed in § 3.4.5. 3.4.3 Fe ii to Mg ii Ratio The ratio of Fe ii to Mg ii provides a good measure of the ionization parameters for log U values greater than -4.0, where U is the ionization potential for the particular transition. Also, if the Fe/Mg ratio is small it could suggest that the gas is α-enhanced, while if it is large it is clear that Type Ia supernovae have played a role in enriching the gas. Alternatively, a lower Fe ii to Mg ii ratio can arise in lower density gas, with a higher ionization parameter. From Fig. 3.5 we see, for the QSOs in particular, an evolution in the equivalent width ratio of Fe ii to Mg ii, with an absence of small values at z < 1.2. We interpret this as an absence of α-enhanced very-strong Mg ii absorbers at low redshifts due, presumably, to a diminishment of contributions of superwinds to this class of absorber at z < 1.2. There are not enough very strong Mg ii absorbers found toward GRB’s to confirm whether the same trend in evolution of the Fe ii to Mg ii ratio holds for the GRB absorbers. However, there are two GRBs with among the lowest Fe ii to Mg ii equivalent width ratios for low redshift absorbers (GRB060418 system at z = 0.655 and GRB050820 system at z = 0.691). If this tendency were to be supported by a larger dataset, it would suggest that the low-redshift GRB absorbers do indeed differ from the quasar population supporting an origin at high redshift in the GRB environment. However, we do not have nearly enough GRB data at this time to support this conclusion. 44 5.0 QSO 4.0 GRB 1.5 3.0 1.0 2.0 1.0 0.5 EW(MgI2853) EW(MgII2803) 0.0 0.0 -0.5 0 1 2 3 4 5 0 1 2 3 4 5 EW(MgII2796) EW(MgII2796) 4.0 2.0 3.5 3.0 1.5 2.5 1.0 2.0 1.5 0.5 1.0 EW(FeII2600) 0.5 EW(FeII2374) 0.0 0.0 -0.5 -0.5 0 1 2 3 4 5 0 1 2 3 4 5 EW(MgII2796) EW(MgII2796) 2.5 0.4 2.0 0.3 1.5 0.2 1.0 0.5 0.1 EW(AlIII1855) 0.0 EW(MnII2577) 0.0 -0.5 -0.1 0 1 2 3 4 5 0 1 2 3 4 5 EW(MgII2796) EW(MgII2796) Fig. 3.4 Rest-frame equivalent widths for different species compared to associated Mg ii λ2706 rest-frame equivalent width for GRB (open symbol) and quasar (filled symbol) absorption systems in our study. Triangles represent upper limits. No particular contrast between the GRB and quasar samples is evident. The detection of Mg i in some GRB absorbers suggests that the absorbing gas cannot be located within 50 kpc of the GRB afterglow (Prochaska et al. 2006). 45 QSO QSO 1.2 GRB GRB 1.1 1 1.0 0.9 0.9 0.8 0.7 0.8 0.6 0.7 0.5 0.5 0.4 0.5 0.3 0.4 0.2 0.3 0.1 FeII2383/MgII2796 0.2 FeII2587/MgII2796 0.1 0.25 0.5 0.75 1 1.25 1.5 1.75 2 2.25 0.25 0.5 0.75 1 1.25 1.5 1.75 2 Redshift Redshift QSO QSO 0 1 GRB 10 GRB 0.9 0.8 0.7 0.6 0.5 0.4 -1 10 0.3 FeII2600/MgII2796 FeII2374/MgII2796 0.2 0.25 0.5 0.75 1 1.25 1.5 1.75 2 1.0 1.5 2.0 Redshift Redshift Fig. 3.5 Fe ii/Mg ii equivalent width ratios for different transitions, as observed in GRB (open symbol) and quasar (filled symbol) absorption systems from our study. Errors are typically smaller than the symbol size. The quasar absorbers exhibit a possible evolution in Fe ii/Mg ii equivalent width ratio with redshift, which cannot be confirmed for the GRB sample because of limited statistics. If present, such an evolution in Fe ii/Mg ii might suggest that α-enhancement plays a dominant role at high redshift while type Ia supernovae are responsible for a greater fraction of metals at z < 1.2. 46 3.4.4 Dust To estimate the dust depletion we used the ratio Fe/Zn. This quantity is a good tracer of depletion due to the diference in these elements’ refractory properties. As shown in Savaglio (2006), dust depletion tends to be higher for a higher Zn ii column density. Specifically, we used the equivalent width ratio of Fe ii λ2374 to Zn ii λ2026 as a dust indicator, since we could directly measure this from our data. Although a column density ratio would be more physically meaningful, our goal is only to make a comparison between the GRB and quasar absorbers, so we have chosen to use the direct observable. Based on the derived values of , for exampple, Wr(2374)/Wr(2026), as shown in Fig. 3.6, we find again that the quasar and GRB populations are similar. However, the lowest data point, arising from the z = 1.106 system toward GRB 060418, deserves further comment. We confirm the finding of Ellison et al. (2006) that this absorber is particularly dusty. However, as mentioned in § 3.2.1, this amount of dust in quasar Mg ii would not be sufficient to prevent a typical quasar from being observed in large- area surveys. It might be an indication of a particularly unusual environment in that GRB absorber. 3.4.5 C iv Kinematics and Phase Structure Photoionization modeling of Mg ii absorbers has shown that C iv and Mg ii absorption does not usually arise in the same phase (gas with the same density and temperature). The C iv arises in a lower-density phase that produces generally broader lines that are often aligned in velocity with components arising in higher-density gas that gives rise to low-ionization absorption (Ding et al. 2003b,a; Masiero et al. 2005). C iv components without associated Mg ii are also often detected at other velocities. There also exists a subset of strong Mg ii systems, so-called C iv-deficient absorption, with only weak or even non-detected C iv (Churchill et al. 2000). The very strong quasar Mg ii absorbers in our sample, with C iv covered, share those general features. Churchill et al. (2000) found two different categories of very strong Mg ii absorbers: one in which the C iv absorption was also very strong (called double), interpreted as consistent with galaxy pairs and one in which the C iv was typical of that found in Wr(2796) < 1 A˚ strong Mg ii absorbers (called DLA/H i-rich), consistent with a particularly dense region producing Mg ii within a normal galaxy and its halo. Unfortunately, in only three GRB absorbers from our sample does the C iv coverage provide information. In the z = 1.7976 absorber toward GRB 060607 the C iv 47 5 3 10 10 QSO QSO 4 GRB GRB 10 3 2 10 10 2 10 1 1 10 10 FeII2383/ZnII2026 FeII2587/ZnII2026 0 10 -1 0 10 10 0.0 0.5 1.0 1.5 2.0 2.5 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 Redshift Redshift 4 4 10 10 QSO QSO GRB GRB 3 3 10 10 2 2 10 10 1 1 10 10 FeII2600/ZnII2026 0 FeII2374/ZnII2026 0 10 10 -1 -1 10 10 0.5 1.0 1.5 2.0 2.5 0.5 1.0 1.5 2.0 2.5 Redshift Redshift Fig. 3.6 Fe ii/Zn ii equivalent width ratios for different transitions, as observed in GRB (open symbol) and quasar (filled symbol) absorption systems from our study. Error bars are typically smaller than the symbol size; lower limits are shown as triangles. Low values of the various Fe ii/Zn ii ratios indicate high depletion of gas onto dust grains. Comparison of the two samples does not suggest any particular enrichment of dusty absorbers among the GRB absorber popula- tion, however, the most extreme (low-ratio) detection at z ∼ 1.1 is for one of the systems found toward GRB 060418 (Ellison et al. 2006). The distribution of ratio values, including lower limits, demonstrate that there is not an excess of dusty absorbers sufficient to bias magnitude-limited quasar surveys and thereby explain the dN/dz discrepancy between GRB and quasar lines of sight. 48 traces the Mg ii in velocity, but two phases are required to explain the relative strengths. There is also a separate C iv component in this absorber without associated detected Mg ii. The z = 1.6015 system toward GRB 021004 has quite a noisy spectrum, but it appears that C iv is only detected in the bluer components of the system. The same phenomenon is observed for the Al iii associated with that system. Finally, with the z = 1.38 system toward GRB 021004, again the spectrum is noisy, but it appears that the C iv roughly traces the Mg ii. We conclude that the GRB C iv absorber profiles appear to have similar kinematic properties to the quasar absorber C iv profiles. 3.4.6 Fine Structure and Metastable Transitions It is common to detect fine structure transitions in the highest-redshift – presumed host galaxy – absorbers of GRBs. These host galaxy absorbers are often DLAs, and are characterized by strong low-ionization absorption with accompanying high-ionization absorption. Among the six GRBs that we have studied here, GRB 050730 (Starling et al. 2005), GRB 060418 (Vreeswijk et al. 2007a) and GRB 050820 (Prochaska 2006a; Berger et al. 2005; Vreeswijk et al. 2007a) have this type of absorber at the GRB redshift. On the other hand, GRB 060607, have reported host galaxy absorbers with only high ionization absorption features (Prochaska et al. 2008b). It is possible the host galaxy absorption for GRB 060607 might actually be at a lower redshift, z = 1.799 and the detected C iv absorption at z = 3.08 might be due to material moving with positive relative velocity along the line of sight within the host. In the case of our six GRBs, fine structure transitions have been reported to be detected in GRB 060418, GRB 021004, and GRB 050730. Ordinary intervening DLAs often have detected C ii∗, but do not have the Si ii∗ and Fe ii∗ transitions that are evident in some of these GRB hosts. Two major mechanisms have been discussed for producing these fine structure lines: UV pumping or collisional excitation. The UV pumping mechanism, generally thought to be more likely (Prochaska et al. 2006), requires the host absorber to be located within the same galaxy as the GRB, but not in its immediate proximity (within tens of parsecs). If the excess of strong Mg ii absorbers in GRBs is duo to the proximity to the GRB radiation field, we might expect to observe fine structure transitions from them as well. We find that these fine structure transitions are not detected in any of the cases here. Fig. 3.7 shows an example of the regions where we would expect to detect Fe ii* and Ni ii* for the z = 1.106 system toward GRB 060418, and they are not detected. We conclude that if the excess absorption arises in material intrinsic to the GRB, it must 49 have properties distinct from the “host galaxy absorbers”, for example, it might be at a closer location but have a much higher density, as we discuss further in § 3.5. Absorption in the meta-stable levels of Fe ii∗ and Ni ii∗ were detected in the host galaxy absorption associated with GRB 060418, providing further evidence that a UV pumping mechanism is favored over collisional excitation (Vreeswijk et al. 2007a). A distance of at least 1.7 kpc was derived for the absorber in this case, using the variability of the fine structure and metastable transitions. We have also examined the metastable transitions for all of the very strong Mg ii absorbers in our GRB sample and do not detect any of these transitions. 3.5 Discussion of Intrinsic Origin of Excess GRB Very Strong Mg ii Absorbers Our conclusion is that there is no significant evidence in support of a difference between the populations of very strong Mg ii absorbers found toward GRBs as compared to quasars. Any such difference could have been used to argue for a difference in the nature of the GRB absorbers. Given the outstanding issues with explanations involving dust bias, gravitational lensing, and partial beam covering, as discussed in § 3.2.4, we would expect such a difference to arise from an origin close to the GRB. The question now is how we should interpret the negative result, that is, an absence of any difference in the observable properties of GRB and quasar absorbers. The most straightfoward explanation would be to say that material ejected from the GRB environment either does not exist or does not produce very strong Mg ii ab- sorption. Indeed, the required velocities of 0.3 or 0.4c are a challenge to models of the regions surrounding GRBs, though they are not impossible as described in § 3.2.4. However, this would leaves the mystery of the excess of Mg ii absorbers toward GRBs without a satisfactory explanation. We could also say that, in view of the small sample size of GRB absorbers and the great diversity in the population of very strong Mg ii absorbers toward quasars, that a difference would be hard to recognize. Another option is to embrace the similarity in physical conditions of the two population of Mg ii absorbers, quasars and GRB, and consider whether there could be similarities between their environments even in a model where GRB absorbers have an intrinsic origin. As described in § 3.2.4, very strong Mg ii absorbers toward quasars are thought to arise from a mix of processes, with significant contributions from dense star-forming regions and from condensations in superwinds (Rao et al. 2006; Nestor et al. 2005). GRBs also are thought to arise within star forming regions. It is intriguing to note that in the 50 z=1.4892 z=1.1066 z=0.6554 1 1 1 0 0 0 1 1 1 0 0 0 1 1 1 0 0 0 -600 -300 0 300 600 -600 -300 0 300 600 -600 -300 0 300 600 Fig. 3.7 Comparison between GRB060418 Fe ii* and Ni ii* transitions at the host galaxy redshift (z = 1.489) and at the redhsifts of two intervening absorbers (1.106, 0.655). The continuum level is shown; detected features are identified near velocity zero, as derived from several other transitions. These absorption features are present in the host galaxy absorber but not in the intervening systems, consistent with a non-intrinsic (intervening galaxy) origin for these systems. The variability of these fine structure transitions, due to UV pumping from the GRB radiation field, places constraints on the location of the host galaxy absorber relative to the GRB (Vreeswijk et al. 2007a). No other example of such fine-structure transitions has been seen in this or other GRB absorbers of our sample. 51 nearby universe, the Carina nebula, an indication of the structure of the interstellar medium near O stars, has absorption profiles very similar to those of some of the very strong Mg ii absorbers we are studying (Walborn et al. 2002). Perhaps similarities in the quasar intervening and GRB absorbers are the result of similar processes being involved. Of course, there is a significant difference in the intervening quasar absorbers and the immediate vicinity of the GRB: the intensity of the GRB radiation field. We have noted that there is an expectation that the GRB radiation field should destroy Mg i within tens of parsecs. Fox et al. (2008) proposed that high velocity components (500–5000 km s−1 relative to the GRB) detected in C iv absorption could be the result of Wolf-Rayet winds on such small scales. They suggest that only the cases without detected low-ionization absorption would be consistent with an origin in the immedi- ate GRB environment. The GRB radiation field should also lead to population of the fine structure levels of various transitions at distances up to the kiloparsec scale, thus producing absorbers consistent with the claimed host galaxy absorption of many GRBs. These host galaxy absorbers are close to the redshift of the GRB and distinct from the absorbers that would produce the excess of very strong Mg ii absorption toward GRBs. There may seem little parameter space left for absorbers near to the GRB to have the physical conditions required. However, one important analogy, already mentioned in § 3.2.4, makes us pause before dismissing an origin for the excess GRB absorbers close to the GRB. There is a population of intrinsic narrow absorption line systems observed in quasar spectra which have their origin in the accretion disk wind of the quasar. In many cases these absorbers have relativistic ejection velocities, and they have ionization parameters and kinematic structures that are virtually indistinguishable from intervening absorbers, particularly C iv absorbers. In some cases, related low-ionization gas is detected. The obvious simi- larity in properties between these intrinsic quasar absorbers and some intervening quasar absorption line systems is despite the location of the intrinsic absorbers in the imme- diate vicinity of the quasar radiation flux. In order that they have the same ionization level they have to have gas densities many orders of magnitude higher than those in intervening absorbers. Despite their different acceleration mechanism and environment, their kinematic properties are somehow similar to those of the intervening absorbers. The only way to distinguish intrinsic from intervening narrow absorption line systems is through partial covering analysis, which is observed in only a subset of the intrinsic pop- ulation (Misawa et al. 2007b). The other members of that intrinsic population remain as intruders in intervening quasar absorption line samples. 52 We conclude by considering the requirements in order that excess very strong Mg ii absorbers seen toward GRBs arise in the GRB environment/host galaxy. The general similarity between the kinematics of the Mg ii profiles of GRB and quasar ab- sorbers indicates that there are similar physical processes involved which are likely to be related to star formation. There is a requirement that the kinematic spread is large and, particularly, that the ejection velocity is relativistic. Superwinds are not adequate for the latter, but supernovae (in the same star forming region as the GRB) could be, or possibly Wolf-Rayet winds further accelerated by the GRB (van Marle et al. 2007). To achieve relativistic velocities, it seems the absorbing material must be relatively close to the GRB radiation field, yet it must still be possible for Mg i to exist. For this, it seems that high densities and shielding would be necessary, in analogy with the situation for intrinsic narrow absorption lines. Also, we need to understand why absorption is not observed in fine structure transitions in view of the proximity to the GRB. Clearly, these requirements are restrictive and lead to some skepticism about an intrinsic origin. On the other hand, there is currently no more obvious explanation for the puzzling excess of very strong Mg ii absorbers toward GRBs. In the small sample of GRB absorbers that we examined, we saw no clear differ- ence with the quasar absorber population, but given the wide range of quasar absorber properties, a larger sample of GRB absorbers might well be needed to see any such trend, if it is present. For example, perhaps some of the GRB absorbers that we have examined have particularly small Mg i equivalent widths. Given the small number of high-quality datasets in the present sample, detailed studies of the high resolution profiles of GRB absorbers are quite worthwhile to pursue for future GRBs. As here, future studies should take into account all of the observed chemical transitions and their kinematics. Through such an expanded sample, it may be finally possible to tell if the excess GRB Mg ii absorbers are telling us something about intervening absorbers or if they are telling us about some component of the environment of the GRB. 3.6 Summary of Results We compared the properties of 9 very strong Mg ii absorbers (Wr(2796) > 1 A)˚ observed along 6 GRB line of sights and compared them to the properties of absorbers at similar redshifts along 81 quasar line of sights. Our aim was to investigate if the reported excess of [dN/dz]GRB ≈ 3 × [dN/dz]QSO found by Prochter et al. (2006b) is caused by excess material in the immediate proximity of the GRB. More specifically, Prochter et al. (2006b) found 14 absorbers toward 14 GRBs, where 3.8 would be expected from quasar 53 statistics. Similarly, we would expect ∼ 2 very strong Mg ii absorbers toward the GRBs in our sample and we find 9. This is not an independent datapoint from Prochter et al. (2006b) since there is overlap between our samples, moreover the probability of us finding 3 or 4 very strong Mg ii absorbers in our GRB sample is not negligible. So, among the 9 absorbers in our GRB sample, we might expect about half should have an origin in the GRB environment if an intrinsic hypothesis is valid. From Fig. 3.1c, we see that such an interpretation of the excess would require relativistic velocities, ωp to at least v ≈ 0.3c. Our challenge, then, was to find differences between an unknown subset of 4 or 5 of the 9 GRB absorbers and the 27 quasar absorbers which themselves constitute a varied sample of intervening environments and processes. Nonetheless, in the intense radiation field of a GRB, and in regions that are directly associated with active star formation, we might expect some differences. If the material is directly related to the GRB, and ejected relativistically from its immediate environment, we might expect strong kinematic and ionization differences compared to ordinary intervening absorbers. In our analysis of 9 GRB and 27 quasar Mg ii absorption systems we have found no statistically significant differences: 1. The kinematics of the Mg ii λ2796 profiles do not differ between the two samples. This is apparent by visual comparison in Figure 3.2, and confirmed by comparison of the kinematic spread, full velocity range, and D-index quantitative measures of kinematics, as shown in Figure 3.3; 2. The equivalent widths of low-ionization transitions (Mg i and Fe ii), dust tracers (Cr ii and Mnii), and intermediate/high-ionization transitions (Al iii and C iv) cover the same range of values for GRB absorbers as for quasar absorbers. The equivalent width distributions of Mg ii λ2796 were not different between the two samples (among the Wr(2796) > 1 A˚ absorbers), so it is meaningful to compare the absolute values of the equivalent widths of other transitions, rather than ratios. Many of the other transitions were only covered for the higher-redshift GRBs in our sample, and so the statistics for this comparison were more limited. We do note that several GRBs have particularly weak Mg i compared to quasar absorbers. If this is found to be common in a larger sample it would be of great interest, since it is expected that Mg i should not survive within 50 pc of a GRB (Prochaska et al. 2006); 3. The distributions of the ratios of Fe ii to Mg ii equivalent widths did not differ significantly between the two samples. This statement applies to saturated Fe ii transitions as well as to Fe ii λ2374, which was rarely saturated for these systems. There is a possible evolutionary trend found in this ratio for quasar very strong Mg ii absorber, with few large ratios of Fe ii to Mg ii at z < 1.2; the GRB sample is not large enough to determine wether a similar trend holds; 54 4. We used the ratio of the equivalent widths of various Fe ii transitions to that of Zn ii λ2026 as an indication of the Fe/Zn ratio, which indicates dust depletion since zinc depletes more readily than iron. There is no statistical difference in the distributions, although one GRB absorber according to this metric (z = 1.106 toward GRB 060418) is confirmed to be particularly dusty(Ellison et al. 2006); 5. The C iv doublet is covered for three of the GRB absorbers in our sample, and that region of the spectrum is noisy for two of the three. In these cases we find that kinematics and strength of the C iv absorption is not exceptional, compared to the range of these properties among the quasar absorbers; 6. Fine structure and metastable transitions were not detected for any of the GRB absorbers in our sample. The fine structure lines are typically detected in asso- ciation with the absorption systems that are believed to represent the GRB host galaxies. Such host galaxy absorption systems are observed and interpret as occur- ring in regions tens of parsecs to several kiloparsecs from the GRB, where gas has been excited by UV pumping (Prochaska et al. 2006; Th¨one et al. 2008; Vreeswijk et al. 2007a). The absence of such fine structure lines in our sample suggests that a different region, or gas with a different density, must be responsible for the intervening very strong Mg ii absorption seen in many GRB afterglow spectra. 55 Chapter 4 Gemini Spectroscopic Catalog of Swift GRB Afterglows In this chapter I present a comprehensive analysis of afterglow spectra collected by our group over the last five years of GRB studies. I focus on the spectroscopic follow- up of GRBs detected by the Swift satellite; I have personally taken the lead in triggering and analyzing a significant fraction of these observations during my time here at Penn State. The amount and quality of data collected allows us, even with a mid-resolution spectrograph such as the Gemini GMOS instruments, to place interesting constraints on the properties of the absorbing gas found along the line of sight in GRB host galaxies at redshifts 0.49 4.1 Introduction The Swift satellite has dramatically reshaped and enhanced our understanding of GRBs (?). With its sensitive Burst Alert Telescope (BAT) providing few-arcmin positions for ≈100 GRBs per year, and most of these almost immediately (δt ∼< 100 s) observed to deep flux limits with the co-aligned X-ray Telescope (XRT) and UV/Optical Telescope (UVOT) instruments, Swift has yielded a flood of insights into GRB physics, progenitor diversity across the full burst population, and the evolution of the burst rate with cosmic time. In addition, the prompt positions have enabled the exploitation of young, bright afterglows as “cosmic lighthouses” illuminating the gas and dust along the line of sight, possibly from distances as near as ∼<1 pc out to cosmological (Gpc) scales. The best case in this respect, probably, was provided by the “naked-eye” GRB 080319B. Thanks to the prompt Swift alert and rapid-response UVOT observations – not to men- tion a fortuitous position in the night sky over the Western hemisphere – this burst was caught at V < 5 mag by small-aperture ground-based telescopes (Racusin et al. 2008), and subjected to a high-resolution (R ≈ 45, 000) spectroscopic sequence beginning just 15 min after the burst (D’Elia et al. 2009). Naturally, the great majority of Swift afterglows have been observed spectroscopi- cally at fainter magnitudes and at later epochs compared to the burst; moreover, most of these observations have been gathered with low- and medium-resolution spectrographs. 56 These data are gathered primarily for the sake of redshift determination. The burst redshift, of course, is a fundamental observable of GRB studies as it sets the intrinsic energy and time scales for all burst phenomena. At the same time, however, such data have the potential – with careful analysis – to be used for afterglow absorber studies as well. During the Swift era, through the end of 2009 (the cut-off date for this study), our group used the Gemini telescopes (North and South) and their GMOS spectrographs to observe 34 long-duration GRB afterglows. For 22 of these, extending over the redshift interval 0.49 4.2 Data sample and analysis The two Gemini telescopes are located in Mauna Kea, Hawaii (North) and Cerro Pachon, Chile (South). Thanks to these strategic locations it is possible, barring poor weather, to observe almost any part of the sky within less than 24 hours. Each telescope is equipped with one Gemini Multiobject Spectrograph (GMOS; Hook et al. (2004)), a standard long-slit / slitmask spectrograph offering a range of filters for imaging, and a range gratings for spectroscopy across the classical optical bandpass. For the present study we restrict our attention to Swift-discovered GRBs detected from the start of the mission through the end of 2009, which were observed with GMOS (North or South) in a spectroscopic mode, and whose data were available to us either through our own programs or via the public Gemini Science Archive. In Table 4.1 we list this sample, indicating the grating configuration, resolution, central wavelength, and total exposure time for each GRB. Gratings efficiency plots are shown in Fig. 4.1. GRB afterglow observations were generally collected within the first day following the GRB. Bursts are initially detected and localized by the Swift BAT, and followed in most cases by prompt (δt . 100 s) X-ray localization of the burst, with a position accuracy of . 2 arcsec (90% of the cases; Evans et al. 2009). In some cases an optical counterpart position was available thanks to prompt UVOT observations, or reports from ground-based facilities. Our group adopted different observational strategies depending on several factors: (1) GRB prompt phase duration (T90) – while short bursts were typically subject to more 57 ˚ A) Exposure(s) Note scopes during the period 2004- December 2009 for ˚ A) Rest-frame range ( Redshift are often determined by absorption features, see ratings used and the total exposure time are reported. ,170 and 68 km/s for the R400, B600 and R831 respectively. Grating Observed range ( Table 4.1. GRB Afterglow Spectral Sample z GRB Telescope Note. — Sample of all the 22 GRBs observed with the Gemini tele GRB060510BGRB060210 G-NorthGRB081029GRB050908 4.922 G-NorthGRB070529 G-South 3.912GRB080413A R400 G-North 3.847GRB080804 G-North R400 G-South 3.339GRB081008 R400 2.498GRB080319C 2.433 G-South 6000-10000 B600GRB080928 G-South B600 G-North 2.205GRB070125 R400 4000-8000 1.967GRB060502 1.95 G-South 4000-8000 R400GRB060418 G-North 5000-7930 B600 1.690GRB080604 G-North B600 5000-7930 1.547 4000-8000GRB071122 1013-1688 G-South R400 1.515GRB091024 G-North 4000-8000 R400 1.489GRB091208B G-North R400 3870-6690 812-1629 1.416GRB071010B 3873-6662 G-North R400 G-North 825-1650GRB080319B 1.14 1152-1827 3934-8087 R400 G-North 1.092GRB080710 1.063 4x1000 5903-10116 1429-2267 G-South 1165-2330GRB060729 R400 0.947 3946-8064 R400GRB081007 R400 0.937 G-South 1248-2496 3946-8064 R400 2x1500 G-South 1304-2254 3946-8064 0.845 R831 2x900 1312-2259 G-South 2x1200 5923-10140 0.543 5954-10182 1462-3006 B600 2x1200 2x900 4000-8100 0.529 2317-3971 B600 4000-8100 1569-3206 B600 4x900 5950-8050 1585-3239 4x900 1633-3338 4x900 3970-6773 2767-4738 3266-6083 2x900 2x2400 2846-4867 4576-7411 1939-3926 4x1800 2054-4160 8x900 3071-4156 4x450 2152-3671 2x900 4x900 2117-3943 2x900 2993-4847 1x1800 2x1800 2x1200 3x400 2x900 which we performed theSection spectroscopic 4.3 study for in more this details. chapter. The spectral resolution is 250 Also, the central wavelength of the collected spectra, the g 58 rt of spectral t t t rt st S/N S/N S/N S/N S/N S/N ort ort r hort or oor oor ernovae (2) the gratings used poor pernovae (1) NTEGRAL SS Galaxy emission ˚ A) Exposure(s) Note period or simply related to host galaxies observations. The scopes during the period 2004- December 2009 which are not pa ˚ A) Rest-frame range ( espectively.Central wavelength of the collected spectra, 2006); (2) Soderberg et al. (2006) pre-Swift , or belongs to S/N Grating Observed range ( Table 4.2. Excluded GRB Afterglow Spectral Sample z GRB Telescope Note. — Sample of all the 21 GRBs observed with the Gemini tele GRB060908GRB080603AGRB050408 G-South G-NorthGRB070208GRB080413B G-North 1.688 2.43 G-South G-South 1.236 R400 R400 1.165 1.10 R831 R400 B600 3924-8086 3939-8125 6293-8363 3959-8100 3885-6660 1460-3008 1148-2369 2814-3740 1829-3741 3532-6055 4x1200 2x900 3x1800 4x1800 1x900 I poor Ho Sh poo GRB070714BGRB070318 G-South G-South 0.92 0.836 R400 R400 5200-7900 5936-10143 2708-4110 3233-5525 10x600s 1x1800 S p GRB060123 G-NorthGRB071112C 1.099?? G-North R400 0.823 B600 3970-8104 3836-6642 1891-3861 2104-3643 2x1500 4x900 po GRB070612GRB050525A G-North G-North 0.617 0.606 R400 R400 3943-8106 4900-8900 2438-5012 3051-5541 2x600 5x1800 p Hos GRB051221GRB090424GRB091127 G-NorthGRB070724 G-SouthGRB090417B 0.547 G-North 0.544 G-South G-NorthGRB050709 R400 0.49GRB060614 0.457 R400 0.345GRB060218 G-North R400 G-South R400 R150 4050-8050 G-South 0.16 0.125 0.033 6000 3942-8100 R400 4946-9027 R400 3600,7600 R400 2617-5203 3350-7550 4000-8050 3600-7600 2646-5237 3395-6196 2676-5651 - 1200 2887-6509 3556-7156 2x900 3485-7358 2x900 4x900 SD 4x900 4x1200 Shor 1x1200 2x900 Host Sho Su Propriety Sup Shor GRB070209 G-South 0.314 R400 4432-8568 3373-6521 2x1200 Sh and the total exposure time are reported.(1) Gal-Yam et al. ( this study because they do not have good resolution is 250,170 and 68 km/s for the R400, B600 and R831 r 59 Fig. 4.1 Efficiency of the different gratings available for the GMOS spectrographs. vigorous attempts at detection and characterization, these were not successful during this period; (2) Observed X-ray and optical afterglow temporal decay index – allowing us to configure the observations as appropriate for the expected afterglow magnitude at the time of observation; (3) UV/optical/infrared counterpart detection and colors, as reported either from UVOT observations or from other facilities. We wish to acknowledge the valuable role that was played, in this context, by the several ground-based robotic telescopes devoted in part to rapid-response GRB observations. Based on these considerations, we performed our observations in order to have the best wavelength coverage (λmin −λmax), signal-to-noise ratio (S/N), and (in some cases) spectral resolution. GMOS is a mid-resolution instrument at best (the best configuration, R831 provide 2.5 A˚ per resolution element, but only in a small wavelength coverage) and, therefore, the majority of our collected spectra have R ∼ 1200 at the central wavelength (R400 grism configuration, corresponding to 5A),˚ and covers ≈4000A˚ simultaneously, typically from the range from 4000A˚ to 8000A.˚ In case of the detection of an X-ray afterglow our approach can be summarized as follows: • If a UV/Optical counterpart is detected: we usually expect a low-redshift (z . 2.5) GRB so we try to cover a blue part of the spectrum (usually R400 or B600 grating) with a 6000A˚ central wavelength; 60 • If upper limits are provided in the optical and only an infrared counterpart is detected (and no lower energy detection has been made) we focus on red end of the spectrum, using the R400 grating, and centering our observation around λ = 8000A.˚ In general, these cases are related to prompt ground-based observations due to robotic telescopes; • If deep and prompt upper limits are provided in red optical filters (R-band or longer-wavelength), we try to detect an infrared counterpart in order to constrain the redshift of the GRB based on photometric redshift estimates. In these case we performed multiband observations in i′ and z′ band with GMOS, and/or in the YJHK bands using NIRI (Hodapp et al. 2003). This strategy has been quite successful, allowing us to gather spectroscopy of GRBs out to z ≈ 5 (Price et al. 2007a), and multiband imaging of the most distant con- firmed GRB (GRB 090423 Tanvir et al. 2009b) and several other high-redshift candidates (e.g., Chapter 2). In this work we present results from our Gemini programs, including also data from other programs when these were publicly accessible in the Gemini Archive1 as of December 31, 2009; some of these data have never previously been published. In over 5 years of observations, from 2005 until December 2009, the Gemini telescopes observed and collected spectral data for 43 GRBs (including both long and short events, as well as host galaxy observations). 4.2.1 Data reduction procedure The data format and the characteristic of the two GMOS instruments (North and South) can be found at the Gemini website2 and the raw data for most of our sample are made publicly available through the Gemini Archive. Calibration data, like flat field images and arc-lamp files, are available as well. As part of our active program, calibration images are taken right after the science frames in order to allow an immediate reduction of the data. Most of the observations were made possible via target of opportunity (ToO) programs, which allow us to interrupt the current queue in order to obtain the most suitable data for each specific GRB. 1http://www1.cadc-ccda.hia-iha.nrc-cnrc.gc.ca/gsa/ 2www.gemini.edu 61 Each single spectrum has been reduced using the standard IRAF3 Gemini and GMOS tasks. The Gemini telescopes optics and site conditions are stable enough to make multiple flat fielding unnecessary, so only one “combined” flat image is acquired and used. After flat fielding, cosmic ray rejection is performed using the lacos spec IRAF routine (van Dokkum 2001). In order to make the reduction procedure faster, we trim the flat-fielded, cosmic ray-cleaned images to a smaller size, making sure that the afterglow spectrum trace is included. Wavelength calibration is then applied to the two-dimensional images using the dedicated tasks GSWAVELENGTH and GSTRANSFORM. Typically two frames were acquired, with the object shifted by ≈79 pixels along the slit between exposures. The two resulting images thus exhibit a shifted afterglow trace, which permits sky subtraction simply by subtraction of the first frame from the second and vice-versa. The resulting images are then co-added using the imcombine task. Finally, using the IRAF task APALL, a one-dimensional spectrum is extracted from the co-added image (or the single available image, if only one was taken), along with a corresponding variance spectrum. Images of our 23 extracted spectra are shown in Appendix C. 4.3 Equivalent widths measure and uncertainties Absorption spectroscopy has been used for decades in order to investigate element abundances of absorbing gas, as well as ionization conditions and gas kinematic studies. While emission lines are sometimes difficult to detect and require calibration to serve as (e.g.) metallicity metrics, absorption lines give a direct, quantitative measure of the number of atoms of a particular species along the GRB line of sight (see Bowen et al. (2005) for comparison between abundances derived from absorption and emission lines). This material is often associated with the host galaxy or, in some cases, with intervening systems (like the Mg ii systems of Chapter 3). In our present application, the presence of multiple absorbers is a challenge for line identification. At our Gemini- GMOS resolution (∼ 5A)˚ most lines are unresolved; strong lines are likely saturated and, if different components are present, these will be blended (see Section 4.3.1 for more detailed discussion). 3IRAF is distributed by the National Optical Astronomy Observatories, which are operated by AURA, Inc., under cooperative agreement with NSF 62 The equivalent width (EW) of individual absorption features is determined by the column density and velocity parameter for each ion. EW measurements are often made by line-fitting but this approach is less appropriate for mid-resolution data as we analyze here. In order to make a higher-quality measurement and take into account systematic effects, at an appropriate level, we decide not to fit the continuum over the full spectral range, but only locally to each feature, within roughly ±5A˚ of the feature wings. Note that figures in Appendix C are flux-normalized spectra, smoothed with a 3-pixel boxcar, but that we work with the unsmoothed, non-flux calibrated data in calculating equivalent widths. In the following section we present a detailed description of our EW measurement procedure, highlighting the caveats involved in making a side-by-side comparison with the more accurate results obtained with high-resolution data. For this purpose we use two datasets gathered from the afterglow of GRB 060418, one from our Gemini sample and the other gathered with the Ultraviolet and Visible Echelle Spectrograph (UVES) instrument on the Very Large Telescope (VLT), which has resolution R ≈ 40, 000. 4.3.1 Low-resolution spectra Absorption spectroscopy represents astronomers’ primary tool for studying the chemical contents of absorbing gas along the line of sight to luminous astronomical objects. Interaction of photons emitted from the source with atoms along the line of sight excites these atoms and molecules, resulting in absorption features in the source spectrum. These absorption features thus encode the ionization state and, collectively, abundance of the absorbing atoms and molecules. In this work we focus on the chemical contents of the GRB environment. Since we use mid-resolution spectra, the Doppler parameter of the absorbers beff is not resolved, and thus, the standard approach of Voight profile fitting (e.g., Savaglio et al. 2003, 2004) is not appropriate. Moreover, while a curve-of-growth analysis (COG; Uns¨old et al. 1930; Wilson 1939; Spitzer & Jenkins 1975) can be used to extract abundances from the EWs of an ensemble of (mostly) unsaturated features, the COG approach is not appropriate for the study of GRB host galaxies, which routinely show multi-component, saturated absorption features in high-resolution spectra. The necessary condition for COG analysis is, in fact, that every species share the same (mass-corrected) Doppler parameter and that they are part of a single “cloud” of material; in the case of the GRB host absorber, these conditions are not expected to be met. In particular, Jenkins et al. (1986) and Prochaska (2006b) have explored the effect of resolution on column density calculations 63 from quasar and afterglow spectra, demonstrating that a composite COG technique, applied to low- or mid-resolution data, systematically underestimates column densities. In order to explore the implications of these challenges for our present project, we tested two alternative fitting procedures on our Gemini GMOS spectrum of GRB 060418, comparing it to a contemporaneous spectrum from VLT UVES (R ≈ 45, 000). In Figure 4.2 we present four different parts of this afterglow spectrum. The UVES spectrum is shown in black and the GMOS spectrum in red; absorption features of particular interest are labeled. It is clear that saturated transitions like Fe ii (2334A)˚ and Fe ii (2600A),˚ clear in the UVES spectrum, are also detected in the GMOS data, while weaker transitions are not always detectable at lower resolution. We calculated equivalent widths for the GMOS spectrum using Voight profile fitting as well as direct summation of pixel values; in both cases the continuum is fit locally. Results of our analysis are presented in Table 4.3. 1.5 1.5 Gemini UVES 1.0 1.0 0.5 0.5 Normalized flux 0.0 0.0 ZnII2026 CrII2062 5010 5020 5030 5040 5050 5060 5070 5110 5120 5130 5140 5150 5160 1.5 1.5 1.0 1.0 0.5 0.5 Normalized flux 0.0 0.0 -0.5 FeII2344 -0.5 FeII2600 5820 5825 5830 5835 5840 5845 5850 6440 6450 6460 6470 6480 6490 6500 Wavelength Wavelength Fig. 4.2 Absorption features detected in the VLT-UVES spectra of GRB 060418 in comparison with our Gemini-GMOS spectra. As is apparent by reference to Table 4.3, the Voigt profile fitting procedure sys- tematically overestimates the “true” (UVES) value of the EW by 5% to 20%; more importantly, the error estimate from SPLOT fitting method is unrealistically small; it is 64 Table 4.3. Observed EW comparison based on GRB 060418 data Ion UVES Gemini Gemini Exact Exact Voigtfit Zn ii 2026 0.70 ± 0.32 < 0.61 < 0.43 Cr ii & Zn ii (2062) < 0.20 < 0.25 < 0.50 Fe ii 2344 1.92 ± 0.84 2.13 ± 0.84 2.76 ± 0.12 Fe ii 2382 2.74 ± 1.55 2.38 ± 0.87 2.86 ± 0.10 Fe ii 2600 3.21 ± 1.55 2.93 ± 1.09 3.31 ± 0.10 Mg ii 2796 4.76 ± 2.29 5.24 ± 2.25 5.86 ± 0.05 Mg ii 2803 4.42 ± 2.09 4.43 ± 1.67 5.20 ± 0.07 Mg i 2853 1.40 ± 0.86 1.24 ± 0.86 1.53 ± 0.02 Note. — Comparison of EW measurements for some features detected in the afterglow spectra of GRB 060418 at high and low resolution. not clear that the substnatial systematic uncertainties are being fully taken into account by this procedure. We therefore determined to estimate the EWs of all features using direct pixel summation, with EW uncertainties derived directly from our variance spec- tra, over the same wavelength range as that used to measure the EW of the absorption features. Tables 4.4, 4.5, 4.6, and 4.7 present our EW measurements for lines detected at > 3-sigma confidence among our 22-spectrum sample. Nondetections of specified features are quoted as upper limits at 3σ confidence. In cases of line-blending, lower limits are indicated; the true EW values in these cases are expected to be of the same order as our quoted limit. In this section we briefly describe the data collected for each GRB, reporting also details of the observations as well as references in the literature, whenever they have been published. GRB 060510B at z = 4.922: This dataset has not previously been published, although the redshift was reported in Price et al. (2007b). The single spectra is noisy, and no absorption lines can be clearly identified above 9000A.˚ A Lyα break is present at 7120A˚ which, in combination with other absorption lines, allows the redshift mea- surement, z = 4.922 ± 0.010. A second DLA seems appear centered at 7106A,˚ which corresponds to a z = 4.844. Some additional metal features have been found confirming this redshift. GRB 060210 at 3.912: This dataset has been published in Fynbo et al. (2009) (FJP09 hereafter). Two systems of absorption features are observed. The first one, 65 which we associate with the GRB host galaxy, is at z = 3.912. A DLA is present in our spectrum, and several other strong metal lines have been identified at similar redshift. A second system is present at z = 3.817. Our EW estimates differ from those of FJP09 at the ≈20% level; differences in normalization and continuum fitting procedures might explain this discrepancy. GRB 081029 at z = 3.847: This spectrum has not been previously published. Detailed study of the high-resolution spectrum of this burst, obtained by VLT + UVES, can be found in D’Elia et al. (2008). A host DLA is observed; metal features from our Gemini data allow us to determine the burst redshift, z = 3.847. There might be an intervening system at z = 2.023 based on detection of Fe ii (2374A)˚ and Fe ii (2382A).˚ GRB 050908 at z = 3.339: This spectrum was first published by FJP09. A sub-DLA system has been identified (log(NHI)=17.60) directly and via associated metal featurs. Also, an intervening system at z = 2.62, with strong metal lines, is present. GRB 070529 z = 2.498: These data have not been published before. The reduction has been complicated by a second source in the slit, close to the GRB afterglow position. Preliminary results were reported via GCN (Berger et al. 2007). A possible C iv (1548,1550 A)˚ doublet is observed, suggesting an intervening system at z = 2.4743 may be present. GRB 080413A at z = 2.432: This spectrum has not been previously published. Some detected features, which led to the redshift determination, were reported in Thoene et al. (2008b) and Cucchiara et al. (2008a). GRB 080804 at z = 2.205: This spectrum has not been previously published. Higher-resolution spectra have led to a detailed analysis of this afterglow (Thoene et al. 2008a; Fynbo et al. 2009). GRB 081008 at z = 1.967: These data have not been previously published. Preliminary analysis was reported in Cucchiara et al. (2008b) and with other results from VLT (D’Avanzo et al. 2008). GRB 080319C at z = 1.95: These data were published in FJP09 and reported in a GCN (Wiersema et al. 2008b). GRB 080928 at z = 1.692: The redshift was originally reported from VLT observations (Vreeswijk et al. 2008). Our Gemini spectrum shows hints of Cr ii and Zn ii lines as well as Fe ii and Mnii lines. We also identified Ca H&K features due to a possible intervening system at z = 0.735. GRB 070125 at z = 1.55: For this event 4 consecutive spectra were taken, and only the Mg ii(2796,2803) doublet is detected. These results were presented by Cenko 66 et al. (2008), who suggest that this GRB likely happened in the halo environment of its host galaxy. GRB 060502 at z = 1.515: These data have not been previously published. Preliminary results and redshift estimate were reported by Chen et al. (2006). GRB 060418 at z = 1.49: This was the first event for which temporal variability in equivalent widths of Fe ii fine-structure transitions was observed, thanks to time-series high-resolution data (Vreeswijk et al. 2007b). In our eight spectra no variability of these lines or others has been observed, probably due to a late acquisition of these data set. Four different systems were present at z = 1.49 (Host), and z = 1.106, 0.651, 0.605. Most of these were identified to be very strong Mg ii absorbers (see also Cucchiara et al. (2009c)). GRB 080604 at z = 1.416: The data for this burst were presented in Wiersema et al. (2008a) and FJP09. GRB 071122 at z = 1.14: These data were reported as a GCN by our group (Cucchiara et al. 2007), and presented recently by FJP09. GRB 091024 at z = 1.092: This GRB was observed by our group (Cucchiara et al. 2009a) and never published before. GRB 091208B at z = 1.063: This Gemini data have not been published. Preliminary results were reported in Wiersema et al. (2009). GRB 071010B at z = 0.947: The Gemini dataset is published here for the first time after a preliminary analysis was reported by Cenko et al. (2007). Also, Keck observed this event, obtaining useful information from the host galaxy emission (Stern et al. 2007; Prochaska et al. 2007). GRB 080319B at z = 0.94: The dataset for the naked-eye burst (Racusin et al. 2008) was also published by FJP09. GRB 080710 at z = 0.8445: Preliminary analysis on this object was reported in Perley et al. (2008). Recently, FJP09 performed a detailed EWs measures for several metal absorption lines. GRB 060729 at z = 0.54: The Gemini dataset are published by FJP09 and results from other facilities appeared in Thoene et al. (2006). GRB 081007 at z = 0.529: Gemini dataset shows emission line from the host galaxy (Berger et al. 2008) and CaII H&K in absorption. 4.3.2 Metallicity and dust tracers In this section, following the fundmental theory of absorption features (Spitzer & Jenkins 1975), we estimate constraints on the column density of some key species which 67 Table 4.4. Equivalent widths of the transitions 060510B 060210 081029 050908 070529 080413A Redshift 4.922 3.912 3.847 3.339 2.498 2.432 DLy-alpha ...... NV 1239 ... 6.67 (2.09) 1.34 (0.75) ...... NV 1243 ... 7.13 (1.72) 0.57 (0.45) ...... SII 1251 < 2.96 2.77 (1.24) 0.59 (0.16) ...... SII1254 5.42(1.25) 7.08(1.67) < 0.41 ...... SiII 1260 ... < 28.11 2.77 (1.01) 1.25 (0.38) ...... SiII* 1265/1266 11.89 (1.87) < 28.11 2.19 (0.93) ...... OI 1302 6.05 (1.19) ... 1.35 (0.82) ...... SiII 1304 6.05 (1.19) ... 2.08 (0.74) ...... SiII* 1309 ...... 0.95 (0.53) < 4.44 ...... CII1334 2.00(0.8) < 20.73 3.34 (0.69) 1.71 (0.93) ...... CII*1336 ... < 20.73 3.34 (0.69) 1.71 (0.93) ...... NiII 1370 < 1.39 ...... SiIV 1394 ... 14.58 (3.3) 1.42 (0.91) 6.97 (1.65) ...... SiIV1403 1.58(0.69) 16.52(3.11) 1.23(0.69) 2.86(1.11) ...... NiII 1502 < 1.62 < 0.69 5.02 (0.66) 1.67 (0.63) ...... SiII 1527 ... 15.52 (3.26) 3.88 (0.96) 2.71 (1.37) < 1.51 4.67 (1.24) SiII* 1533 ... 4.97 (1.78) 5.49 (0.56) < 0.38 1.78(0.51) 0.99(0.55) CIV 1548 ...... 10.99 (1.98) ... 1.37 (0.24) 9.69 (0.74) CIV 1551 ...... 10.99 (1.98) ... 1.11 (0.6) 9.69 (0.74) CI 1560 ...... FeII 1608 ...... < 0.51 ... AlII 1671 ...... < 0.71 4.53 (1.29) NiII 1709 ...... NiII 1742 ...... < 0.44 3.41 (0.63) SiII 1808 ...... 2.54 (0.85) 1.44 (0.93) ... AlIII 1855 ...... < 0.81 2.55 (0.94) AlIII 1863 ...... 1.22 (0.53) 0.93 (0.66) ZnII 2026 ...... < 1.72 2.98 (1.45) CrII 2056 ...... 1.32 (0.90) 1.57 (0.81) CrII2062 ...... > 0.63 > 0.97 ZnII 2063 ...... > 0.63 > 0.97 CrII 2066 ...... < 2.51 2.16 (0.92) Note. — Equivalent widths or limits for all species searched in this study. Equivalent widths are reported at 3σ confidence level. Error values, in A,˚ are quoted in parenthesis. Upper and lower limits are also reported, where appropriate; see text for details. 68 Table 4.5. Equivalent widths of the transitions (continue) 080804 081008 080319C 080928 070125 060502 Redshift 2.205 1.9675 1.950 1.690 1.55 1.51 SiII* 1309 ... < 1.03 ...... CII 1334 ... 5.51 (1.09) ...... CII* 1336 ... 5.51 (1.09) ...... NiII 1370 ...... SiIV 1394 ... 2.37 (1.14) ...... SiIV 1403 ... > 2.16 ...... NiII 1502 ... 1.09 (0.78) 5.22 (0.5) ...... SiII 1527 3.88 (0.87) 3.27 (1.04) 7.63 (0.51) ...... SiII* 1533 3.88 (0.87) > 1.37 5.53 (0.68) ...... CIV1548 6.72(0.79) ... < 25.94 ...... CIV1551 6.72(0.79) ... < 25.94 ...... CI 1560 ... < 4.35 9.51 (0.73) ...... FeII 1608 ...... 8.95 (0.83) ...... AlII 1671 4.56 (0.89) 2.77 (1.29) 13.25 (0.6) ...... NiII 1709 ... 0.94 (0.61) ...... NiII 1742 ... < 0.48 < 3.57 ...... SiII 1808 2.46 (0.61) 2.35 (0.66) 3.34 (0.76) ...... < 5.52 AlIII 1855 ... 1.18 (0.96) 4.56 (0.64) ...... < 4.25 AlIII 1863 ... 0.77 (0.7) 6.72 (0.65) ...... < 1.77 ZnII 2026 ... 2.31 (0.91) 3.27 (0.56) 2.13 (1.14) ... < 1.32 CrII 2056 ... 1.11 (0.64) 4.45 (0.58) ...... 3.59 (0.93) CrII 2062 > 0.87 < 1.34 < 3.64 < 1.57 ... < 3.9 ZnII 2063 > 0.87 < 1.34 < 3.64 < 1.57 ... < 3.9 CrII2066 ... < 1.21 ... 3.67 (0.54) ... 5.32 (0.66) FeII 2260 ...... 2.87 (0.69) ... < 2.54 NiII** 2316 ...... FeII* 2333 ...... FeII 2344 ...... < 6.07 FeII 2374 ...... 7.90 (1.40) FeII 2383 ...... 5.22 (1.17) MnII 2577 ...... 2.59 (0.92) FeII 2587 ...... 5.15 (1.00) MnII 2594 ...... 1.51 (0.42) FeII 2600 ...... 6.26 (1.48) MnII 2606 ...... MgII 2796 ...... > 13.95 0.46 (0.05) > 7.06 MgII 2803 ...... > 13.95 0.20 (0.03) > 3.36 MgI 2853 ...... < 11.46 Note. — Equivalent widths or limits for all species searched in this study. Equivalent widths are reported at 3σ confidence level. Error values, in A,˚ are quoted in parenthesis. Upper and lower limits are also reported, where appropriate; see text for details. 69 Table 4.6. Equivalent widths of the transitions (continue) 060418 080604 071122 091024 091208 071010B Redshift 1.489 1.416 0.140 1.092 1.063 0.9465 ZnII 2026 < 0.94 ...... CrII 2056 ...... CrII 2062 < 0.51 ...... ZnII 2063 ...... CrII 2066 < 0.17 ...... FeII 2260 < 1.13 < 3.11 ...... NiII** 2316 ...... FeII* 2333 ...... FeII 2344 2.24 (0.88) 3.00 (0.46) ...... FeII 2374 4.52 (1.32) 4.30 (0.89) ...... FeII 2383 2.72 (0.96) 8.54 (0.93) ...... MnII 2577 1.76 (0.8) 4.72 (0.46) ...... FeII 2587 2.09 (0.89) 8.68 (1.61) ...... 2.68 (0.52) ... MnII 2594 0.64 (0.4) > 4.93 ...... 1.39 (0.3) ... FeII 2600 3.20 (1.08) < 8.75 ...... 3.03 (0.88) ... MnII 2606 0.68 (0.54) < 2.44 ...... MgII 2796 > 5.16 12.06 (1.33) 10.78 (1.36) ... > 3.48 > 6.02 MgII 2803 > 4.79 7.28 (1.16) 7.89 (1.02) ... > 3.04 > 4.21 MgI 2853 ...... 4.26 (0.79) 4.04 (0.99) 1.79 (0.82) 1.94 (0.75) FeI 3021 ...... TiII 3074 ...... > 0.61 TiII 3384 ...... CaII 3934 ...... 2.51 (0.83) 4.14 (1.34) ...... CaII 3969 ...... 2.23 (0.52) 2.46 (1.31) ...... Note. — Equivalent widths or limits for all species searched in this study. Equivalent widths are reported at 3σ confidence level. Error values, in A,˚ are quoted in parenthesis. Upper and lower limits are also reported, where appropriate; see text for details. 70 Table 4.7. Equivalent widths of the transitions (continue) 080319B 080710 060729 081007 Redshift 0.937 0.8445 0.543 0.530 FeII 2344 ... 0.81 (0.65) ...... FeII 2374 ... 0.51 (0.47) ...... FeII 2383 ... 1.00 (0.66) ...... MnII 2577 ...... FeII 2587 ... 0.67 (0.30) 1.80 (0.67) ... MnII 2594 ... < 0.15 ...... FeII 2600 ... 1.13 (0.7) 2.22 (0.81) ... MnII 2606 ... < 0.13 ...... MgII 2796 ... 2.27 (1.04) 1.63 (0.82) ... MgII 2803 ... 1.92 (0.95) 1.67 (0.75) ... MgI 2853 ... 0.77 (0.7) 1.38 (0.85) ... FeI 3021 ...... TiII 3074 ...... TiII 3384 ...... 0.91 (0.54) CaII3934 2.39(0.83) ... 1.85(0.86) 1.34(0.62) CaII 3969 < 0.73 ...... 1.53 (0.69) Note. — Equivalent widths or limits for all species searched in this study. Equivalent widths are reported at 3σ confidence level. Error values, in A,˚ are quoted in parenthesis. Upper and lower limits are also reported, where appropriate; see text for details. 71 are often used in metallicity and dust studies of GRB host galaxies. For the following discussion we make use of the most recent compilation of solar abundances by Asplund et al. (2009). First, we want to test the dust content using the properties of non-refractory ele- ments, which are not depleted in dust grains (such as Zn ii) and compare that with other elements which we are expected to be depleted (e.g., Cr ii). We follow the methodol- ogy presented in Wolfe et al. (2005) for QSO-DLAs absorbers in order to establish the presence of dust. We therefore make the implicit assumption that most of the GRB host galaxies are indeed DLAs. Using weak lines of these two elements and the reasonable assumption that they lie on the linear part of the curve-of-growth, we can base our anal- ysis on the ratio of EWs. In particular, we can reasonably consider the presence of a dusty host if: EW log ZnII > −0.426 EWCrII In Fig. 4.3 we show our data in combination with data from FJP09. As we can see, for GRB afterglows where the Zn ii and Cr ii lines are detected, the provided ratio shows that most likely these objects are indeed “dusty”, based on the depletion of Cr ii into dust grains. One GRB present in the FJP09 sample, GRB 070802, shows in its continuum spectrum a “2175 A”˚ bump, clearly indicating a Milky Way (evolved galaxy) extinction law for its z = 2.45 host galaxy (El´ıasd´ottir et al. 2009). Regarding the metallicity content of the GRB environment, we again construct a cautious analysis given the limitations of our dataset. As with our investigation of dust depletion, we need to identify good metallicity tracers (e.g., Si, S, O, Zn). As shown by Stanek et al. (2006b), oxygen column densities inferred from host galaxiy emission lines suggest that, at low redshift (z < 0.5), GRBs arise preferentially in galaxies with sub-solar metallicity, hZi ≈ 0.1Z⊙. For our absorption line analysis, we make use of Si ii 1256, which has been considered an accurate metallicity tracer (Fynbo et al. 2009; Prochaska et al. 2008a). Roughly speaking, an equivalent width of EWSiII 1256 = 1A˚ means a metallicity of 0.1 solar. The average value for our sample is hEWSiII 1256i = 1.43 which implies Z ∼ 0.19Z⊙, in agreement with previous studies. 4.4 Conclusion We have presented a carefully-constructed and comprehensive list of equivalent width measurements (EWs) for spectral absorption features identified in the spectra of 72 0.8 0.6 ) 0.4 Cr2056 W 0.2 / 0.0 ZnII2026 (W -0.2 log log -0.4 -0.6 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 WZnII2026 Fig. 4.3 Equivalent width ratio of Zn ii and Cr ii lines in our sample (blue points) and FJP09 (red points). The horizontal line represents the limit above which the absorber is considered dusty, based on depletion of Cr ii into dust grains. 22 GRB afterglows observed at low- to medium-resolution with the GMOS spectrographs on Gemini North and South. Having determined that Voight profile-fitting approaches underestimate EW un- certainties, we pursued instead a direct pixel summation approach, with local continuum fits, which allows us to make reasonable estimates of the statistical and systematic un- certainties associated with each EW measurement. We also use direct pixel summation of pixel values and variances in the calculation of upper and lower limits on feature EWs, where appropriate. Using our results in combination with other published work, we investigated the dust content of the host galaxies of some GRBs, using observed Zn ii:Cr ii EW ratios. We verified that the GRB host galaxies are observed to be dustier than the typical pop- ulation of QSO-DLAs. Separately, based on EW values for Si ii (1256A),˚ we calculated an expected metallicity for some hosts. Consistent with earlier studies, we find evidence for sub-Solar metallicity of the GRB host galaxy absorbers. Based on our analysis, and assuming our sample absorbers have typical line-of-sight neutral hydrogen column densities, the average metallicity in our sample is Z ≈ 0.2Z⊙. This set of afterglow spectra, gathered over the course of five years of follow-up campaigns focused on GRB detections from the Swift satellite, thus offers a wealth of 73 information on the typical properties of the GRB host galaxy, as seen in absorption against the bright light of the young afterglow. Going forward, the commissioning of higher-resolution, broader-coverage spectrographs like X-Shooter on the VLT (D’Odorico et al. 2006) can be expected to dramatically expand the physical and astrophysical constraints that can be derived, in similar fashion, from data that are routinely collected in the course of such GRB campaigns. Outstanding questions that might be addressed in such future studies include: the detection and characterization of more extreme-redshift (z > 8) GRBs; the history of metal enrichment of star-forming regions across cosmic time; the role of DLAs and sub-DLAs as reservoirs of neutral hydrogen out to high redshift; and the history and evolution of the cosmic re-ionization from the highest redshifts down to z ≈ 5. 74 Chapter 5 Conclusions and Future work 5.1 GRB afterglow spectroscopy: where do we stand? The GRB field has been dramatically energized by the Nov 2004 launch and continuing successful operations of the Swift satellite. Studying the bursts detected and data collected by the three instruments of this observatory, an enormous number of new insights about the emission mechanisms involved in these phenomena have been derived. The amount of GRBs detected and localized in near-real time has reached almost 100 bursts per year, providing several opportunities for study this objects from the ground as well. From small robotic telescopes to world-class >8-meter aperture facilities, the GRB community has been following up nearly every burst detection in order to detect broadband afterglows, characterize their emission, and learn about the burst progenitors, environments and host galaxies, and the implications of these objects for our wider understanding of the Universe and its evolution, back to the earliest epochs. This thesis work has explored some of the incredible opportunities offered by GRB afterglows study in the Swift era, focusing on the use of these afterglows as probes of their host galaxies and the cosmos. As such, we have focused on optical absorption spec- troscopy and (in the case of GRB 090429B) photometric observations of the optical/NIR afterglows of Swift-detected GRBs. Although we have not discussed Swift-based XRT and UVOT observations of bursts other than GRB 090429B, we note that the rapid re- sponse necessary to collect high signal-to-noise afterglow spectra is crucially enabled by the prompt few-arcsecond and sub-arcsecond positions provided by these instruments, and by the prompt optical flux measurements from UVOT. All in all the amount of information embedded and accessible in these afterglow spectra goes well beyond a “simple” redshift determination. Of course, even the routine collection of GRB redshifts – an effort we are actively engaged in and consider well worthwhile – represents a nearly inconceivable advance on the state of the art in GRB studies as little as 15 years ago, as we reviewed in Chapter 1. 5.2 Afterglows as Probes In Chapter 2 we presented our attempt to constrain the redshift of GRB 090429B using multiband imaging data. This project is energized in part by our recent discovery 75 of GRB 090423 at z = 8.2, which has demonstrated that GRBs are being detected by Swift out to “extreme” (z > 8) redshifts. Determining the redshift distribution of GRBs is important not only to understand GRBs as a phenomenon, but because it sheds light on the history of star formation out to the earliest epochs. In addition, high-quality spectra of bright GRB afterglows from the early Universe can provide unique insights into the growth in metal enrichment and the reionization of the intergalactic medium through cosmic time. Although our lack of a spectrum prevented us from confirming a redshift for this event, our deep upper limits in multiple optical bands, combined with detection of a relatively bright, blue afterglow at wavelengths beyond 1 µm, make this burst an intriguing extreme-redshift candidate. Our photometric redshift modeling yields a best-fit solution placing this GRB at z ≈ 9.5, with negligible extinction in its host galaxy. Alternative lower-redshift solutions are possible if the burst is subject to host extinction: at z ∼ 5 with AV ≈ 0.8 mag, or at z ≈ 2 with AV ≈ 2 mag. Given the complications associated with time-critical spectroscopic observations, such photometric redshift analyses will likely remain an important part of the GRB observer’s toolkit for some time to come. The revelation of the cosmological nature of the GRBs, and the quantity and quality of the burst localizations from Swift, have enabled the first detailed studies of absorption systems along GRB lines of sight, and comparison to the results of decades of similar studies focused on quasar absorption systems. The comparison between GRB host galaxies and quasar damped Lyman-α (DLA) systems, for example, has been useful for placing the GRB host galaxies in a cosmological context. In a similar comparative analysis, Prochter et al. (2006a) investigated the Mg ii content of GRB lines of sight, finding that the very strong Mg ii absorbers (typically associated with DLAs, when both features are observable) are significantly over-represented along GRB lines of sight by comparison to quasars, by nearly a factor of four (dn/dzGRB & 3 × dn/dzQSO). In 3 we presented our study of high-quality spectra from 8 GRBs and 27 quasars, carrying out a uniform search for strong line-of-sight (non-host galaxy) Mg ii doublet features. We were particularly interested in the possibility that some of the GRB absorbers might be produced by high-velocity gas associated with the GRB, its progenitor, or its host galaxy. Our sample was selected deliberately to be homogeneous in resolution, instrumen- tation (VLT+UVES), and signal-to-noise. We made a comparative study of the GRB and quasar line-of-sight absorbers in terms of redshift distribution, equivalent width (EW) and EW ratios, velocity spread, and other kinematic metrics. In all cases, we failed to 76 identify a subpopulation of GRB absorbers with properties that set them apart, in any way, from the corresponding quasar distributions. Due to the relatively small number of GRB absorbers, we were not able to make a definitive conclusion, but overall an intrinsic (high-velocity) origin for the excess GRB absorbers is disfavored by our analysis. Thus, hopes for resolving the nature of the excess GRB absorbers have settled, for the most part, on the possibility that gravitational lensing of GRBs by the absorbers can explain the discrepancy (Vergani et al. 2009; Wyithe et al. 2010b). In our final Chapter 4, we presented an extensive catalog of GRB afterglow spectra obtained mainly by our group with the two Gemini telescopes and their mid-resolution GMOS spectrographs, located on Mauna Kea, Hawaii and Cerro Pachon, Chile. These spectra do not permit the detailed kinematic and abundance (curve of growth) analyses allowed by high-resolution spectroscopy, mainly due to blending of multiple components, and hence, inaccurate kinematic property estimates. Nonetheless, they encode extensive information about the chemical contents of the GRB host galaxy, and as has been shown in extensive quasar studies at similar resolution (mainly using Sloan Digital Sky Survey data), several carefully-chosen metallicity and dust diagnostics, as well as the neutral hydrogen column, can be accurately characterized with these data. Although we have analyzed all Gemini+GMOS afterglow spectra accessible to us, for purposes of this work we focus on the subset having continuum signal-to-noise S/N > 10, measuring equivalent widths and upper or lower limits for all host galaxy-related metal features, as appropriate given the burst galaxies redshift (the sample ranges from z = 0.033 to z = 4.9). 22 objects satisfied our selection criteria and, in combination with the similar published catalog of Fynbo et al. (2009), we obtained interesting constraints on dust and metal content in the typical GRB host. As previously found in studies of several individual GRB host galaxies, our sample host galaxies exhibit sub-solar metal content (hZi = 0.19Z⊙); moreover, some of them show a higher dust content than is seen in the quasar DLA samples. In this context, a recent suggestive result was the detection of a 2175A˚ “bump” dust absorption feature in the spectrum of GRB 070802 (El´ıasd´ottir et al. 2009). 5.3 Into the future... The last few years have thus provided important new insights into the capacity of GRB afterglows to illuminate conditions in their host galaxies and along the cosmological line of sight, out to the very highest redshifts. 77 We have successfully extended afterglow studies out to “extreme” redshifts, z > 8, deep into the “Cosmic Dark Ages” before the end of the reionization of the intergalac- tic medium. New missions, including the JANUS satellite (Roming 2008), have been proposed to capitalize on these successes. Discovering more extreme-redshift events, like GRB 090423 and (perhaps) GRB 090429B, should allow us to exploit their bright afterglows as tools to understand the chemical enrichment, star-formation history, and evolution of structure in the early Universe. In order to achieve this challenging task, further deployment of robotic telescopes, multi-band cameras (such as GROND), multi- arm spectrographs (such as X-Shooter), and enhanced NIR capabilities will be required. Such capabilities will enable rapid discrimination among the faint, extinguished, and truly extreme-redshift events, at an early time when the GRB is still bright, enabling the collection of the highest-quality spectra. As Swift continues to generate new burst and afterglow detections at the rate of ≈100 GRBs per year, it is clear that the number of spectra obtained both at low and high-resolution will increase. As we have demonstrated in this work, in addition to pro- viding burst redshifts, these data can be applied to investigate a number of outstanding questions. First, we will expect to detect more strong Mg ii absorbers, substantially increasing the statistical sample to explore the nature of the excess absorbers along GRB lines of sight. We expect that progress will also be made in direct imaging of the GRB absorbers (e.g., Chen et al. 2009), a complementary approach toward identifying systematic differences between the GRB and quasar samples. At the same time, we have shown the utility of afterglow spectroscopic data in understanding element abundances, dust depletion, and gas kinematics in the GRB host galaxy. Traditionally these data are gathered at high-resolution and high signal-to- noise; the contingencies of afterglow observing rarely allow for this best-case scenario, yet as we have shown, valuable information can nonetheless be extracted from lower- resolution, lower signal-to-noise data. We will continue to investigate the nature of GRB host galaxies and explore their metallicity, dust, and hydrogen contents, out to the highest redshifts. With greatly-increased numerical samples, we will be able to perform statistical comparisons with the vast datasets (and subsamples) of QSO spectra in order to place the GRB environment and line of sight absorbers in their proper cosmological context. Ultimately, the promise of afterglows as probes of their host galaxies and the cosmos will be realized when such studies take their place alongside the traditional galaxy survey and quasar absorption approaches as an equal, recognized for having opened a unique and complementary window on the properties and contents of the high-redshift Universe. 78 Appendix A GRB high-resolution absorption features This appendix present the sample of GRB spectra obtained with the VLT-UVES instruments. 79 GRB 060418 z=0.6021 1 0 1 0 1 0 1 0 1 0 1 0 1 0 -600 -400 -200 0 200 400 600 Velocity [km/s] Fig. A.1 Absorber at z = 0.6021 along GRB 060418 (z = 1.49) line of sight. Transitions are aligned in velocity space, with the Mg ii λ2706 transition centered at the median optical depth. 80 GRB 060418 z=0.6554 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 -600 -400 -200 0 200 400 600 Velocity [km/s] Fig. A.2 The same as Fig. A.1, but for the absorber at z = 0.6554 along the GRB 060418 (z = 1.49) line of sight. 81 GRB 050820 z=0.6910 1 0 1 0 1 0 1 0 1 0 1 0 1 0 -600 -400 -200 0 200 400 600 Velocity [km/s] Fig. A.3 The same as Fig. A.1, but for the absorber at z = 0.6910 alongthe GRB 050820 (z = 2.612) line of sight. 82 GRB 060418 z=1.1066 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 -600 -400 -200 0 200 400 600 Velocity [km/s] Fig. A.4 The same as Fig. A.1, but for the absorber at z = 1.1066 along the GRB 060418 (z = 1.49) line of sight. 83 GRB 021004 z=1.3800 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 -600 -400 -200 0 200 400 600 Velocity [km/s] Fig. A.5 The same as Fig. A.1, but for the absorber at z = 1.38 along GRB 021004 (z = 2.328) line of sight. 84 GRB 050820 z=1.4280 1 0 1 0 1 0 1 0 1 0 -600 -400 -200 0 200 400 600 Velocity [km/s] Fig. A.6 The same as Fig. A.1, but for the absorber at z = 1.428 along GRB 050820 (z = 2.612) line of sight. 85 GRB 021004 z=1.6015 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 -600 -400 -200 0 200 400 600 Velocity [km/s] Fig. A.7 The same as Fig. A.1, but for the absorber at z = 1.6015 along GRB 021004 (z = 2.328) line of sight. 86 GRB 060607 z=1.7996 1 1 0 0 1 1 0 0 1 -600 -400 -200 0 200 400 600 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 -600 -400 -200 0 200 400 600 Velocity [km/s] Fig. A.8 The same as Fig. A.1, but for the absorber at z = 1.7996 along GRB 060607 (z = 3.08) line of sight. 87 GRB 050730 z=2.2527 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 -600 -400 -200 0 200 400 600 Velocity [km/s] Fig. A.9 The same as Fig. A.1, but for the absorber at z = 2.2527 along GRB 050730 (z = 3.967) line of sight. 88 Appendix B QSO high-resolution absorption features This appendix present the sample of QSO spectra obtained with the VLT-UVES instruments. 89 Q0100+0130 z=0.2779 1 0 1 0 1 0 -600 -400 -200 0 200 400 600 Velocity [km/s] Fig. B.1 Absorber at z = 0.2779 along QSO0100+0130 (z = 2.681) line of sight.Transitions are aligned in velocity space, with the Mg ii λ2706 transition centered at the median optical depth. 90 Q1127-145 z=0.3127 1 0 1 0 1 0 1 0 1 0 1 0 -600 -400 -200 0 200 400 600 Velocity [km/s] Fig. B.2 The same as Fig. B.1, but for the absorber at z = 0.3127 along Q1127-145 (z = 1.18) line of sight. 91 Q1229-021 z=0.3951 1 0 1 0 1 0 1 0 1 0 -600 -400 -200 0 200 400 600 Velocity [km/s] Fig. B.3 The same as Fig. B.1, but for the absorber at z = 0.3951 along Q1229-021 (z = 1.038) line of sight. 92 3C336 z=0.6560 1 0 1 0 1 0 1 0 1 0 1 0 -600 -400 -200 0 200 400 600 Velocity [km/s] Fig. B.4 The same as Fig. B.1, but for the absorber at z = 0.6560 along 3C336 (z = 0.9273) line of sight. 93 HE1122-1648 z=0.6822 1 0 1 0 1 0 1 0 1 0 1 0 -600 -400 -200 0 200 400 600 Velocity [km/s] Fig. B.5 The same as Fig. B.1, but for the absorber at z = 0.6822 along HE1122-1648 (z = 2.40) line of sight. 94 Q0453-423 z=0.7261 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 -600 -400 -200 0 200 400 600 Velocity [km/s] Fig. B.6 The same as Fig. B.1, but for the absorber at z = 0.7261 along Q0453-423 (z = 2.66) line of sight. 95 Q0002-422 z=0.8366 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 -600 -400 -200 0 200 400 600 Velocity [km/s] Fig. B.7 The same as Fig. B.1, but for the absorber at z = 0.8366 along Q0002-422 (z = 2.76) line of sight. 96 3C336 z=0.8912 1 0 1 0 1 0 1 0 1 0 1 0 -600 -400 -200 0 200 400 600 Velocity [km/s] Fig. B.8 The same as Fig. B.1, but for the absorber at z = 0.8912 along 3C336 (z = 0.9273) line of sight. 97 Q1629+120 z=0.9002 1 0 1 0 1 0 1 0 1 0 1 0 1 0 -600 -400 -200 0 200 400 600 Velocity [km/s] Fig. B.9 The same as Fig. B.1, but for the absorber at z=0.9002 along Q1629+120 (z = 1.795) line of sight. 98 Q0130-4021 z=0.9315 1 0 1 0 1 0 1 0 1 0 1 0 1 0 -600 -400 -200 0 200 400 600 Velocity [km/s] Fig. B.10 The same as Fig. B.1, but for the absorber at z = 0.9315 along Q0130-4021 (z = 3.03) line of sight. 99 CTQ0298 z=1.0387 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 -600 -400 -200 0 200 400 600 Velocity [km/s] Fig. B.11 The same as Fig. B.1, but for the absorber at z = 1.0387 along CTQ0298 (z = 3.370) line of sight. 100 Q0328-272 z=1.1228 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 -600 -400 -200 0 200 400 600 Velocity [km/s] Fig. B.12 The same as Fig. B.1, but for the absorber at z = 1.1228 along Q0328-272 (z = 1.816) line of sight. 101 Q1621-0042 z=1.1334 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 -600 -400 -200 0 200 400 600 Velocity [km/s] Fig. B.13 The same as Fig. B.1, but for the absorber at z = 1.1334 along Q1621-0042 (z = 3.7) line of sight. 102 Q0453-423 z=1.1498 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 -600 -400 -200 0 200 400 600 Velocity [km/s] Fig. B.14 The same as Fig. B.1, but for the absorber at z = 1.1498 along Q0453-423 (z = 2.66) line of sight. 103 Q0112+300 z=1.2452 1 0 1 0 1 0 1 0 1 0 1 0 1 0 -600 -400 -200 0 200 400 600 Velocity [km/s] Fig. B.15 The same as Fig. B.1, but for the absorber at z = 1.2452 along Q0112+0300 (z = 2.81) line of sight. 104 HE1341-1020 z=1.2767 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 -600 -400 -200 0 200 400 600 Velocity [km/s] Fig. B.16 The same as Fig. B.1, but for the absorber at z = 1.2767 along HE1341-1020 (z = 2.134) line of sight. 105 Q0109-3518 z=1.3495 1 1 0 0 1 1 0 0 1 1 0 0 1 1 0 0 -600 -400 -200 0 200 400 600 1 0 1 0 1 0 1 0 1 0 1 0 -600 -400 -200 0 200 400 600 Velocity [km/s] Fig. B.17 The same as Fig. B.1, but for the absorber at z = 1.3495 along Q0109-3518 (z = 2.35) line of sight. 106 PKS0237-23 z=1.3650 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 -600 -400 -200 0 200 400 600 Velocity [km/s] Fig. B.18 The same as Fig. B.1, but for the absorber at z = 1.3650 along PKS0237-23 (z = 2.224) line of sight. 107 Q1418-064 z=1.4578 1 0 1 0 1 0 -600 -400 -200 0 200 400 600 Velocity [km/s] Fig. B.19 The same as Fig. B.1, but for the absorber at z = 1.4578 along Q1418-064 (z = 3.689) line of sight. 108 PKS0237-23 z=1.6723 1 1 0 0 1 1 0 0 1 1 0 0 -600 -400 -200 0 200 400 600 1 0 1 0 1 0 1 0 1 0 1 0 1 0 -600 -400 -200 0 200 400 600 Velocity [km/s] Fig. B.20 The same as Fig. B.1, but for the absorber at z = 1.6723 along PKS0237-23 (z = 2.224) line of sight. 109 HE2217-2818 z=1.6921 1 1 0 0 1 1 0 0 1 1 0 0 1 1 0 0 -600 -400 -200 0 200 400 600 1 0 1 0 1 0 1 0 1 0 1 0 -600 -400 -200 0 200 400 600 Velocity [km/s] Fig. B.21 The same as Fig. B.1, but for the absorber at z = 1.6921 along HE2217-2818 (z = 2.406) line of sight. 110 Q1331+170 z=1.7766 1 1 0 0 -600 -400 -200 0 200 400 600 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 -600 -400 -200 0 200 400 600 Velocity [km/s] Fig. B.22 The same as Fig. B.1, but for the absorber at z = 1.7766 along Q1331+170 (z = 2.084) line of sight. 111 HE0940-1050 z=1.7891 1 1 0 0 1 1 0 0 1 1 0 0 1 1 0 0 -600 -400 -200 0 200 400 600 1 0 1 0 1 0 1 0 1 0 1 0 -600 -400 -200 0 200 400 600 Velocity [km/s] Fig. B.23 The same as Fig. B.1, but for the absorber at z = 1.7891 along HE0940-1050 (z = 3.083) line of sight. 112 Q0100+0130 z=1.7969 1 1 0 0 1 1 0 0 -600 -400 -200 0 200 400 600 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 -600 -400 -200 0 200 400 600 Velocity [km/s] Fig. B.24 The same as Fig. B.1, but for the absorber at z = 1.7969 along Q0100+130 (z = 2.681) line of sight. 113 Q1151+068 z=1.8191 1 1 0 0 1 1 0 0 1 1 0 0 1 1 0 0 1 1 0 0 1 1 0 0 1 1 0 0 1 1 0 0 -600 -400 -200 0 200 400 600 1 0 1 0 -600 -400 -200 0 200 400 600 Velocity [km/s] Fig. B.25 The same as Fig. B.1, but for the absorber at z = 1.8191 along Q1151+068 (z = 2.762) line of sight. 114 Q0551-3637 z=1.9609 1 1 0 0 1 1 0 0 1 1 0 0 1 1 0 0 1 1 0 0 1 1 0 0 1 1 0 0 1 1 0 0 1 1 0 0 -600 -400 -200 0 200 400 600 1 0 -600 -400 -200 0 200 400 600 Velocity [km/s] Fig. B.26 The same as Fig. B.1, but for the absorber at z = 1.9609 along Q0551-3637 (z = 2.318) line of sight. 115 Q0002-422 z=2.3019 1 1 0 0 1 1 0 0 -600 -400 -200 0 200 400 600 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 -600 -400 -200 0 200 400 600 Velocity [km/s] Fig. B.27 The same as Fig. B.1, but for the absorber at z = 2.3019 along Q0002-422 (z = 2.76) line of sight. 116 Appendix C Catalog of GRB Gemini spectra 2.0 GRB 060510B (z = 4.922) 1.5 1.0 0.5 Normalized flux 0.0 ¡ ¡ ¡¡¡ -0.5 G G -1.0 6500 7000 7500 8000 8500 9000 9500 Wavelength (A ) Fig. C.1 GRB 060510B at z = 4.922 117 2.5 GRB 060210 (z = 3.912) 2.0 1.5 1.0 0.5 Normalized flux 0.0 -0.5 £ £ G £G -1.0 5500 6000 6500 7000 7500 8000 Wavelength (A¢) Fig. C.2 GRB 060210 at z = 3.912 2.5 GRB 081029 (z = 3.847) 2.0 1.5 1.0 0.5 Normalized flux 0.0 -0.5 ¥ ¥ G ¥G -1.0 5000 5500 6000 6500 7000 7500 8000 Wavelength (A¤) Fig. C.3 GRB 081029 at z = 3.847 118 2.0 GRB 050908 (z = 3.339) 1.5 1.0 0.5 Normalized flux 0.0 §§§ § -0.5 § GG -1.0 5000 5500 6000 6500 7000 7500 8000 Wavelength (A¦) Fig. C.4 GRB 050908 at z = 3.339 2.5 GRB 070529 (z = 2.498) 2.0 1.5 1.0 0.5 Normalized flux 0.0 -0.5 © © G © G -1.0 5000 5500 6000 6500 7000 7500 8000 Wavelength (A¨) Fig. C.5 GRB 070529 at z = 2.498 119 2.0 GRB 080413A (z = 2.432) 1.5 1.0 Normalized flux 0.5 G G 0.0 4500 5000 5500 6000 6500 7000 7500 Wavelength (A ) Fig. C.6 GRB 080413A at z = 2.432 2.0 GRB 080804 (z = 2.205) 1.5 1.0 0.5 Normalized flux 0.0 -0.5 G G -1.0 4500 5000 5500 6000 6500 7000 7500 8000 Wavelength (A ) Fig. C.7 GRB 080804 at z = 2.205 120 2.0 GRB 081008 (z = 1.9675) 1.5 1.0 Normalized flux 0.5 G G 0.0 4000 4500 5000 5500 6000 6500 Wavelength (A) Fig. C.8 GRB 081008 at z = 1.9675 2.0 GRB 080319C (z = 1.950) 1.5 1.0 Normalized flux 0.5 G G 0.0 4500 5000 5500 6000 6500 Wavelength (A) Fig. C.9 GRB 080319C at z = 1.950 121 2.5 GRB 080829 (z = 1.690) 2.0 1.5 1.0 0.5 Normalized flux 0.0 -0.5 G G -1.0 5000 5500 6000 6500 7000 7500 8000 Wavelength (A) Fig. C.10 GRB 080928 at z = 1.690 2.5 GRB 060502 (z = 1.510) 2.0 1.5 1.0 0.5 Normalized flux 0.0 -0.5 G G -1.0 5000 5500 6000 6500 7000 7500 8000 Wavelength (A) Fig. C.11 GRB 060502 at z = 1.510 2.5 GRB 060418 (z = 1.489) 2.0 1.5 1.0 0.5 Normalized flux 0.0 -0.5 G G -1.0 5000 5500 6000 6500 7000 7500 8000 Wavelength (A) Fig. C.12 GRB 060418 at z = 1.489 122 2.0 GRB 080604 (z = 1.416) 1.5 1.0 Normalized flux 0.5 G G 0.0 5000 5500 6000 6500 7000 7500 8000 Wavelength (A) Fig. C.13 GRB 080604 at z = 1.416 2.5 GRB 071122 (z = 1.14) 2.0 1.5 1.0 0.5 Normalized flux 0.0 -0.5 G G -1.0 6000 7000 8000 9000 10000 Wavelength (A) Fig. C.14 GRB 071122 at z = 1.140 123 2.5 GRB 091024 (z = 1.092) 2.0 1.5 1.0 0.5 Normalized flux 0.0 -0.5 G G -1.0 6000 7000 8000 9000 10000 Wavelength (A ) Fig. C.15 GRB 091024 at z = 1.092 2.0 GRB 091208B (z = 1.063) 1.5 1.0 0.5 Normalized flux 0.0 -0.5 G G 5000 5500 6000 6500 7000 7500 8000 Wavelength (A ) Fig. C.16 GRB 091208B at z = 1.063 2.0 GRB 071010B (z = 0.947) 1.5 1.0 0.5 Normalized flux 0.0 !!! ! -0.5 G ! G 5000 5500 6000 6500 7000 7500 8000 Wavelength (A ) Fig. C.17 GRB 071010B at z = 0.947 124 2.0 GRB 080710 (z = 0.845) 1.5 1.0 Normalized flux 0.5 ## 0.0 G #G 4000 4500 5000 5500 6000 6500 Wavelength (A") Fig. C.18 GRB 080710 at z = 0.845 125 2.0 GRB 060729 (z = 0.543) 1.5 1.0 Normalized flux 0.5 % G G % 0.0 4000 4500 5000 5500 6000 Wavelength (A$) Fig. C.19 GRB 060729 at z = 0.543 2.0 GRB 081007 (z = 0.530) 1.5 1.0 Normalized flux 0.5 ''' 'G G 5000 5500 6000 6500 7000 Wavelength (A&) Fig. C.20 GRB 081007 at z = 0.530 126 Bibliography Asplund, M., Grevesse, N., Sauval, A. J., & Scott, P. 2009, ARAA, 47, 481 Baltz, E. A. & Hui, L. 2005, ApJ, 618, 403 Band, D., Matteson, J., Ford, L., Schaefer, B., Palmer, D., Teegarden, B., Cline, T., Briggs, M., Paciesas, W., Pendleton, G., Fishman, G., Kouveliotou, C., Meegan, C., Wilson, R., & Lestrade, P. 1993, ApJ, 413, 281 Barthelmy, S. D., Barbier, L. M., Cummings, J. R., Fenimore, E. E., Gehrels, N., Hullinger, D., Krimm, H. A., Markwardt, C. B., Palmer, D. M., Parsons, A., Sato, G., Suzuki, M., Takahashi, T., Tashiro, M., & Tueller, J. 2005, Space Science Reviews, 120, 143 Becker, R. H., Fan, X., White, R. L., Strauss, M. A., Narayanan, V. K., Lupton, R. H., Gunn, J. E., Annis, J., Bahcall, N. A., Brinkmann, J., Connolly, A. J., Csabai, I., Czarapata, P. C., Doi, M., Heckman, T. M., Hennessy, G. S., Ivezi´c, Z.,ˇ Knapp, G. R., Lamb, D. Q., McKay, T. A., Munn, J. A., Nash, T., Nichol, R., Pier, J. R., Richards, G. T., Schneider, D. P., Stoughton, C., Szalay, A. S., Thakar, A. R., & York, D. G. 2001, AJ, 122, 2850 Berger, E. 2009, ApJ, 690, 231 Berger, E., Fox, D. B., & Cucchiara, A. 2007, GRB Coordinates Network, 6470, 1 Berger, E., Fox, D. B., Cucchiara, A., & Cenko, S. B. 2008, GRB Coordinates Network, 8335, 1 Berger, E., Penprase, B. E., Fox, D. B., Kulkarni, S. R., Hill, G., Schaefer, B., & Reed, M. 2005, ArXiv Astrophysics e-prints Bolzonella, M., Miralles, J.-M., & Pell´o, R. 2000, A&A, 363, 476 Bouch´e, N., Murphy, M. T., P´eroux, C., Csabai, I., & Wild, V. 2006, MNRAS, 371, 495 Bouch´e, N., Murphy, M. T., P´eroux, C., Davies, R., Eisenhauer, F., F¨orster Schreiber, N. M., & Tacconi, L. 2007, ApJL, 669, L5 Bowen, D. V., Jenkins, E. B., Pettini, M., & Tripp, T. M. 2005, ApJ, 635, 880 127 Bromm, V. & Loeb, A. 2006, ApJ, 642, 382 Burrows, D. N., Hill, J. E., Nousek, J. A., Kennea, J. A., Wells, A., Osborne, J. P., Abbey, A. F., Beardmore, A., Mukerjee, K., Short, A. D. T., Chincarini, G., Cam- pana, S., Citterio, O., Moretti, A., Pagani, C., Tagliaferri, G., Giommi, P., Capalbi, M., Tamburelli, F., Angelini, L., Cusumano, G., Br¨auninger, H. W., Burkert, W., & Hartner, G. D. 2005, Space Science Reviews, 120, 165 Cenko, S. B., Cucchiara, A., Fox, D. B., Berger, E., & Price, P. A. 2007, GRB Coordi- nates Network, 6888, 1 Cenko, S. B., Fox, D. B., Penprase, B. E., Cucchiara, A., Price, P. A., Berger, E., Kulkarni, S. R., Harrison, F. A., Gal-Yam, A., Ofek, E. O., Rau, A., Chandra, P., Frail, D. A., Kasliwal, M. M., Schmidt, B. P., Soderberg, A. M., Cameron, P. B., & Roth, K. C. 2008, ApJ, 677, 441 Chen, H., McMillan, R., Bender, A., York, D., & Lamb, D. 2006, GRB Coordinates Network, 5051, 1 Chen, H., Perley, D. A., Pollack, L. K., Prochaska, J. X., Bloom, J. S., Dessauges- Zavadsky, M., Pettini, M., Lopez, S., Dall’aglio, A., & Becker, G. D. 2009, ApJ, 691, 152 Chen, H., Prochaska, J. X., Bloom, J. S., & Thompson, I. B. 2005, ApJL, 634, L25 Churchill, C. W., Mellon, R. R., Charlton, J. C., Jannuzi, B. T., Kirhakos, S., Steidel, C. C., & Schneider, D. P. 2000, ApJ, 543, 577 Churchill, C. W. & Vogt, S. S. 2001, AJ, 122, 679 Costa, E., Feroci, M., Piro, L., Frontera, F., Zavattini, G., Nicastro, L., Palazzi, E., Gandolfi, G., Smith, M., Coletta, A., D’Andreta, G., Spoliti, G., Jager, R., & Heise, J. 1997, IAU Circ., 6533, 1 Cucchiara, A., Fox, D., & Tanvir, N. 2009a, GRB Coordinates Network, Circular Service, 65, 1 (2009), 65, 1 Cucchiara, A., Fox, D. B., & Berger, E. 2009b, GRB Coordinates Network, 9209, 1 Cucchiara, A., Fox, D. B., & Cenko, S. B. 2007, GRB Coordinates Network, 7124, 1 —. 2008a, GRB Coordinates Network, 7616, 1 128 Cucchiara, A., Fox, D. B., Cenko, S. B., & Berger, E. 2008b, GRB Coordinates Network, 8346, 1 Cucchiara, A., Jones, T., Charlton, J. C., Fox, D. B., Einsig, D., & Narayanan, A. 2009c, ApJ, 697, 345 D’Avanzo, P., D’Elia, V., & Covino, S. 2008, GRB Coordinates Network, 8345, 1 D’Elia, V., Covino, S., & D’Avanzo, P. 2008, GRB Coordinates Network, 8438, 1 D’Elia, V., Fiore, F., Perna, R., Krongold, Y., Covino, S., Fugazza, D., Lazzati, D., Nicastro, F., Antonelli, L. A., Campana, S., Chincarini, G., D’Avanzo, P., Della Valle, M., Goldoni, P., Guetta, D., Guidorzi, C., Meurs, E. J. A., Mirabel, F., Molinari, E., Norci, L., Piranomonte, S., Stella, L., Stratta, G., Tagliaferri, G., & Ward, P. 2009, ApJ, 694, 332 Ding, J., Charlton, J. C., Bond, N. A., Zonak, S. G., & Churchill, C. W. 2003a, ApJ, 587, 551 Ding, J., Charlton, J. C., Churchill, C. W., & Palma, C. 2003b, ApJ, 590, 746 D’Odorico, S., Dekker, H., Mazzoleni, R., Vernet, J., Guinouard, I., Groot, P., Hammer, F., Rasmussen, P. K., Kaper, L., Navarro, R., Pallavicini, R., Peroux, C., & Zerbi, F. M. 2006, in Presented at the Society of Photo-Optical Instrumentation Engineers (SPIE) Conference, Vol. 6269, Society of Photo-Optical Instrumentation Engineers (SPIE) Conference Series El´ıasd´ottir, A.,´ Fynbo, J. P. U., Hjorth, J., Ledoux, C., Watson, D. J., Andersen, A. C., Malesani, D., Vreeswijk, P. M., Prochaska, J. X., Sollerman, J., & Jaunsen, A. O. 2009, ApJ, 697, 1725 Ellison, S. L., Vreeswijk, P., Ledoux, C., Willis, J. P., Jaunsen, A., Wijers, R. A. M. J., Smette, A., Fynbo, J. P. U., Møller, P., Hjorth, J., & Kaufer, A. 2006, MNRAS, 372, L38 Evans, P. A., Beardmore, A. P., Page, K. L., Osborne, J. P., O’Brien, P. T., Willingale, R., Starling, R. L. C., Burrows, D. N., Godet, O., Vetere, L., Racusin, J., Goad, M. R., Wiersema, K., Angelini, L., Capalbi, M., Chincarini, G., Gehrels, N., Kennea, J. A., Margutti, R., Morris, D. C., Mountford, C. J., Pagani, C., Perri, M., Romano, P., & Tanvir, N. 2009, MNRAS, 397, 1177 129 Evans, P. A., Beardmore, A. P., Page, K. L., Tyler, L. G., Osborne, J. P., Goad, M. R., O’Brien, P. T., Vetere, L., Racusin, J., Morris, D., Burrows, D. N., Capalbi, M., Perri, M., Gehrels, N., & Romano, P. 2007, A&A, 469, 379 Ferguson, T. S., Genest, C., & Hallin, M. 1995, Kendalls tau for autocorrelation Fernandez-Soto, A., Mannucci, F., Fugazza, D., Antonelli, L. A., Campana, S., Chin- carini, G., Covino, S., D’Avanzo, P., D’Elia, V., Della Valle, M., Fiorenzano, A., Guidorzi, C., Maiorano, E., Mao, J., Margutti, R., Palazzi, E., & Thoene, C. C. 2009, GRB Coordinates Network, 9222, 1 Fox, A. J., Ledoux, C., Vreeswijk, P. M., Smette, A., & Jaunsen, A. O. 2008, A&A, 491, 189 Frail, D. A., Kulkarni, S. R., Wieringa, M. H., Taylor, G. B., Moriarty-Schieven, G. H., Shepherd, D. S., Wark, R. M., Subrahmanyan, R., McConnell, D., & Cunningham, S. J. 2000, in American Institute of Physics Conference Series, Vol. 526, Gamma-ray Bursts, 5th Huntsville Symposium, ed. R. M. Kippen, R. S. Mallozzi, & G. J. Fishman, 298–302 Frank, S., Bentz, M. C., Stanek, K. Z., Mathur, S., Dietrich, M., Peterson, B. M., & Atlee, D. W. 2007, APSS, 312, 325 Fynbo, J. P. U., Jakobsson, P., Prochaska, J. X., Malesani, D., Ledoux, C., de Ugarte Postigo, A., Nardini, M., Vreeswijk, P. M., Wiersema, K., Hjorth, J., Sollerman, J., Chen, H., Th¨one, C. C., Bj¨ornsson, G., Bloom, J. S., Castro-Tirado, A. J., Christensen, L., De Cia, A., Fruchter, A. S., Gorosabel, J., Graham, J. F., Jaunsen, A. O., Jensen, B. L., Kann, D. A., Kouveliotou, C., Levan, A. J., Maund, J., Masetti, N., Milvang- Jensen, B., Palazzi, E., Perley, D. A., Pian, E., Rol, E., Schady, P., Starling, R. L. C., Tanvir, N. R., Watson, D. J., Xu, D., Augusteijn, T., Grundahl, F., Telting, J., & Quirion, P. 2009, ApJS, 185, 526 Fynbo, J. P. U., Starling, R. L. C., Ledoux, C., Wiersema, K., Th¨one, C. C., Soller- man, J., Jakobsson, P., Hjorth, J., Watson, D., Vreeswijk, P. M., Møller, P., Rol, E., Gorosabel, J., N¨ar¨anen, J., Wijers, R. A. M. J., Bj¨ornsson, G., Castro Cer´on, J. M., Curran, P., Hartmann, D. H., Holland, S. T., Jensen, B. L., Levan, A. J., Limousin, M., Kouveliotou, C., Nelemans, G., Pedersen, K., Priddey, R. S., & Tanvir, N. R. 2006, A&A, 451, L47 Gal-Yam, A., Fox, D. B., Price, P. A., Ofek, E. O., Davis, M. R., Leonard, D. C., Soderberg, A. M., Schmidt, B. P., Lewis, K. M., Peterson, B. A., Kulkarni, S. R., Berger, E., Cenko, S. B., Sari, R., Sharon, K., Frail, D., Moon, D., Brown, P. J., Cucchiara, A., Harrison, F., Piran, T., Persson, S. E., McCarthy, P. J., Penprase, B. E., Chevalier, R. A., & MacFadyen, A. I. 2006, Nature, 444, 1053 130 Galama, T. J., Vreeswijk, P. M., van Paradijs, J., Kouveliotou, C., Augusteijn, T., B¨ohnhardt, H., Brewer, J. P., Doublier, V., Gonzalez, J., Leibundgut, B., Lidman, C., Hainaut, O. R., Patat, F., Heise, J., in’t Zand, J., Hurley, K., Groot, P. J., Strom, R. G., Mazzali, P. A., Iwamoto, K., Nomoto, K., Umeda, H., Nakamura, T., Young, T. R., Suzuki, T., Shigeyama, T., Koshut, T., Kippen, M., Robinson, C., de Wildt, P., Wijers, R. A. M. J., Tanvir, N., Greiner, J., Pian, E., Palazzi, E., Frontera, F., Masetti, N., Nicastro, L., Feroci, M., Costa, E., Piro, L., Peterson, B. A., Tinney, C., Boyle, B., Cannon, R., Stathakis, R., Sadler, E., Begam, M. C., & Ianna, P. 1998, Nature, 395, 670 Ganguly, R. & Brotherton, M. S. 2008, ApJ, 672, 102 Garnavich, P. M., Loeb, A., & Stanek, K. Z. 2000, ApJL, 544, L11 Gehrels, N., Ramirez-Ruiz, E., & Fox, D. B. 2009, ARAA, 47, 567 Greiner, J., Bornemann, W., Clemens, C., Deuter, M., Hasinger, G., Honsberg, M., Hu- ber, H., Huber, S., Krauss, M., Kr¨uhler, T., K¨upc¨uYolda¸s, A., Mayer-Hasselwander, H., Mican, B., Primak, N., Schrey, F., Steiner, I., Szokoly, G., Th¨one, C. C., Yolda¸s, A., Klose, S., Laux, U., & Winkler, J. 2008, PASP, 120, 405 Greiner, J., Kr¨uhler, T., Fynbo, J. P. U., Rossi, A., Schwarz, R., Klose, S., Savaglio, S., Tanvir, N. R., McBreen, S., Totani, T., Zhang, B. B., Wu, X. F., Watson, D., Barthelmy, S. D., Beardmore, A. P., Ferrero, P., Gehrels, N., Kann, D. A., Kawai, N., Yolda¸s, A. K., M´esz´aros, P., Milvang-Jensen, B., Oates, S. R., Pierini, D., Schady, P., Toma, K., Vreeswijk, P. M., Yolda¸s, A., Zhang, B., Afonso, P., Aoki, K., Burrows, D. N., Clemens, C., Filgas, R., Haiman, Z., Hartmann, D. H., Hasinger, G., Hjorth, J., Jehin, E., Levan, A. J., Liang, E. W., Malesani, D., Pyo, T., Schulze, S., Szokoly, G., Terada, K., & Wiersema, K. 2009, ApJ, 693, 1610 Hao, H., Stanek, K. Z., Dobrzycki, A., Matheson, T., Bentz, M. C., Kuraszkiewicz, J., Garnavich, P. M., Howk, J. C., Calkins, M. L., Worthey, G., Modjaz, M., & Serven, J. 2007, ApJL, 659, L99 Hirose, Y., Umemura, M., Yonehara, A., & Sato, J. 2006, ApJ, 650, 252 Hjorth, J., Sollerman, J., Møller, P., Fynbo, J. P. U., Woosley, S. E., Kouveliotou, C., Tanvir, N. R., Greiner, J., Andersen, M. I., Castro-Tirado, A. J., Castro Cer´on, J. M., Fruchter, A. S., Gorosabel, J., Jakobsson, P., Kaper, L., Klose, S., Masetti, N., Pedersen, H., Pedersen, K., Pian, E., Palazzi, E., Rhoads, J. E., Rol, E., van den Heuvel, E. P. J., Vreeswijk, P. M., Watson, D., & Wijers, R. A. M. J. 2003, Nature, 423, 847 131 Hodapp, K. W., Jensen, J. B., Irwin, E. M., Yamada, H., Chung, R., Fletcher, K., Robertson, L., Hora, J. L., Simons, D. A., Mays, W., Nolan, R., Bec, M., Merrill, M., & Fowler, A. M. 2003, PASP, 115, 1388 Hook, I. M., Jørgensen, I., Allington-Smith, J. R., Davies, R. L., Metcalfe, N., Murowin- ski, R. G., & Crampton, D. 2004, PASP, 116, 425 Jakobsson, P., Malesani, D., Fynbo, J. P. U., Hjorth, J., & Milvang-Jensen, B. 2009, in American Institute of Physics Conference Series, Vol. 1133, American Institute of Physics Conference Series, ed. C. Meegan, C. Kouveliotou, & N. Gehrels, 455–463 Jenkins, E. B., Savage, B. D., & Spitzer, Jr., L. 1986, ApJ, 301, 355 Kacprzak, G. G., Churchill, C. W., Steidel, C. C., & Murphy, M. T. 2008, AJ, 135, 922 Kann, D. A., Klose, S., Zhang, B., Malesani, D., Nakar, E., Pozanenko, A., Wilson, A. C., Butler, N. R., Jakobsson, P., Schulze, S., Andreev, M., Antonelli, L. A., Bikmaev, I. F., Biryukov, V., Boettcher, M., Burenin, R. A., Castro Cer´on, J. M., Castro-Tirado, A. J., Chincarini, G., Cobb, B. E., Covino, S., D’Avanzo, P., D’Elia, V., Della Valle, M., de Ugarte Postigo, A., Efimov, Y., Ferrero, P., Fugazza, D., Fynbo, J. P. U., Galfalk, M., Grundahl, F., Gorosabel, J., Gupta, S., Guziy, S., Hafizov, B., Hjorth, J., Holhjem, K., Ibrahimov, M., Im, M., Jelinek, M., Jensen, B. L., Karimov, R., Khamitov, I. M., Kiziloglu, U., Klunko, E., Kubanek, P., Kutyrev, A. S., Laursen, P., Levan, A. J., Mannucci, F., Martin, C. M., Mescheryakov, A., Mirabal, N., Norris, J. P., Ovaldsen, J., Paraficz, D., Pavlenko, E., Piranomonte, S., Rossi, A., Rumyantsev, V., Salinas, R., Sergeev, A., Sharapov, D., Sollerman, J., Stecklum, B., Stella, L., Tagliaferri, G., Tanvir, N. R., Telting, J., Testa, V., Updike, A. C., Volnova, A., Watson, D., Wiersema, K., & Xu, D. 2007, ArXiv e-prints Kawai, N., Kosugi, G., Aoki, K., Yamada, T., Totani, T., Ohta, K., Iye, M., Hattori, T., Aoki, W., Furusawa, H., Hurley, K., Kawabata, K. S., Kobayashi, N., Komiyama, Y., Mizumoto, Y., Nomoto, K., Noumaru, J., Ogasawara, R., Sato, R., Sekiguchi, K., Shirasaki, Y., Suzuki, M., Takata, T., Tamagawa, T., Terada, H., Watanabe, J., Yatsu, Y., & Yoshida, A. 2006, Nature, 440, 184 Klebesadel, R. W., Strong, I. B., & Olson, R. A. 1973, ApJL, 182, L85+ Kulkarni, S. R., Frail, D. A., Wieringa, M. H., Ekers, R. D., Sadler, E. M., Wark, R. M., Higdon, J. L., Phinney, E. S., & Bloom, J. S. 1998, Nature, 395, 663 132 Lamb, D. Q. & Reichart, D. E. 2000, ApJ, 536, 1 Levesque, E. M., Kewley, L. J., Graham, J. F., & Fruchter, A. S. 2010, ApJL, 712, L26 Loeb, A. & Perna, R. 1998, ApJ, 495, 597 Masiero, J. R., Charlton, J. C., Ding, J., Churchill, C. W., & Kacprzak, G. 2005, ApJ, 623, 57 M´enard, B., Nestor, D., Turnshek, D., Quider, A., Richards, G., Chelouche, D., & Rao, S. 2008, MNRAS, 385, 1053 Mesinger, A. 2009, ArXiv e-prints Meszaros, P. & Rees, M. J. 1997, ApJ, 476, 232 Metzger, M. R., Djorgovski, S. G., Kulkarni, S. R., Steidel, C. C., Adelberger, K. L., Frail, D. A., Costa, E., & Frontera, F. 1997, Nature, 387, 878 Misawa, T., Charlton, J. C., Eracleous, M., Ganguly, R., Tytler, D., Kirkman, D., Suzuki, N., & Lubin, D. 2007a, ApJS, 171, 1 Misawa, T., Eracleous, M., Charlton, J. C., Ganguly, R., Tytler, D., Kirkman, D., Suzuki, N., & Lubin, D. 2007b, in Astronomical Society of the Pacific Conference Series, Vol. 373, The Central Engine of Active Galactic Nuclei, ed. L. C. Ho & J.-W. Wang, 291–+ Mshar, A. C., Charlton, J. C., Lynch, R. S., Churchill, C., & Kim, T.-S. 2007, ApJ, 669, 135 Narayanan, A., Misawa, T., Charlton, J. C., & Kim, T.-S. 2007, ApJ, 660, 1093 Nardini, M., Ghisellini, G., & Ghirlanda, G. 2006, A&A, 451, 821 —. 2008, MNRAS, 386, L87 Nestor, D. B., Turnshek, D. A., & Rao, S. M. 2005, ApJ, 628, 637 Olivares, F., Kupcu Yoldas, A., Greiner, J., Kruehler, T., & Yoldas, A. 2009, GRB Coordinates Network, 9283, 1 Perley, D. A., Cenko, S. B., Bloom, J. S., Chen, H., Butler, N. R., Kocevski, D., Prochaska, J. X., Brodwin, M., Glazebrook, K., Kasliwal, M. M., Kulkarni, S. R., Lopez, S., Ofek, E. O., Pettini, M., Soderberg, A. M., & Starr, D. 2009, AJ, 138, 1690 133 Perley, D. A., Chornock, R., & Bloom, J. S. 2008, GRB Coordinates Network, 7962, 1 Piro, L., Costa, E., Feroci, M., Frontera, F., Amati, L., dal Fiume, D., Antonelli, L. A., Heise, J., in ’t Zand, J., Owens, A., Parmar, A. N., Cusumano, G., Vietri, M., & Perola, G. C. 1999, ApJL, 514, L73 Pontzen, A., Hewett, P., Carswell, R., & Wild, V. 2007, MNRAS, 381, L99 Porciani, C., Viel, M., & Lilly, S. J. 2007, ApJ, 659, 218 Price, P. A., Songaila, A., Cowie, L. L., Bell Burnell, J., Berger, E., Cucchiara, A., Fox, D. B., Hook, I., Kulkarni, S. R., Penprase, B., Roth, K. C., & Schmidt, B. 2007a, ApJL, 663, L57 —. 2007b, ApJL, 663, L57 Prochaska, J. X. 2006a, ApJ, 650, 272 —. 2006b, ApJ, 650, 272 Prochaska, J. X., Chen, H.-W., & Bloom, J. S. 2006, ApJ, 648, 95 Prochaska, J. X., Dessauges-Zavadsky, M., Ramirez-Ruiz, E., & Chen, H. 2008a, ApJ, 685, 344 Prochaska, J. X., Dessauges-Zavadsky, M., Ramirez-Ruiz, E., & Chen, H.-W. 2008b, ApJ, 685, 344 Prochaska, J. X., Perley, D. A., Modjaz, M., Bloom, J. S., & Poznanski, D. 2007, GRB Coordinates Network, 6890, 1 Prochter, G. E., Prochaska, J. X., Chen, H., Bloom, J. S., Dessauges-Zavadsky, M., Foley, R. J., Lopez, S., Pettini, M., Dupree, A. K., & Guhathakurta, P. 2006a, ApJL, 648, L93 Prochter, G. E., Prochaska, J. X., Chen, H.-W., Bloom, J. S., Dessauges-Zavadsky, M., Foley, R. J., Lopez, S., Pettini, M., Dupree, A. K., & Guhathakurta, P. 2006b, ApJL, 648, L93 Racusin, J. L., Karpov, S. V., Sokolowski, M., Granot, J., Wu, X. F., Pal’Shin, V., Covino, S., van der Horst, A. J., Oates, S. R., Schady, P., Smith, R. J., Cummings, J., Starling, R. L. C., Piotrowski, L. W., Zhang, B., Evans, P. A., Holland, S. T., Malek, K., Page, M. T., Vetere, L., Margutti, R., Guidorzi, C., Kamble, A. P., Curran, 134 P. A., Beardmore, A., Kouveliotou, C., Mankiewicz, L., Melandri, A., O’Brien, P. T., Page, K. L., Piran, T., Tanvir, N. R., Wrochna, G., Aptekar, R. L., Barthelmy, S., Bartolini, C., Beskin, G. M., Bondar, S., Bremer, M., Campana, S., Castro-Tirado, A., Cucchiara, A., Cwiok, M., D’Avanzo, P., D’Elia, V., Della Valle, M., de Ugarte Postigo, A., Dominik, W., Falcone, A., Fiore, F., Fox, D. B., Frederiks, D. D., Fruchter, A. S., Fugazza, D., Garrett, M. A., Gehrels, N., Golenetskii, S., Gomboc, A., Gorosabel, J., Greco, G., Guarnieri, A., Immler, S., Jelinek, M., Kasprowicz, G., La Parola, V., Levan, A. J., Mangano, V., Mazets, E. P., Molinari, E., Moretti, A., Nawrocki, K., Oleynik, P. P., Osborne, J. P., Pagani, C., Pandey, S. B., Paragi, Z., Perri, M., Piccioni, A., Ramirez-Ruiz, E., Roming, P. W. A., Steele, I. A., Strom, R. G., Testa, V., Tosti, G., Ulanov, M. V., Wiersema, K., Wijers, R. A. M. J., Winters, J. M., Zarnecki, A. F., Zerbi, F., M´esz´aros, P., Chincarini, G., & Burrows, D. N. 2008, Nature, 455, 183 Rao, S. M., Turnshek, D. A., & Nestor, D. B. 2006, ApJ, 636, 610 Rodriguez Hidalgo, P., Hamann, F., Nestor, D., & Shields, J. 2007, in Astronomical So- ciety of the Pacific Conference Series, Vol. 373, The Central Engine of Active Galactic Nuclei, ed. L. C. Ho & J.-W. Wang, 287–+ Roming, P. 2008, in COSPAR, Plenary Meeting, Vol. 37, 37th COSPAR Scientific As- sembly, 2645–+ Roming, P. W. A., Kennedy, T. E., Mason, K. O., Nousek, J. A., Ahr, L., Bingham, R. E., Broos, P. S., Carter, M. J., Hancock, B. K., Huckle, H. E., Hunsberger, S. D., Kawakami, H., Killough, R., Koch, T. S., McLelland, M. K., Smith, K., Smith, P. J., Soto, J. C., Boyd, P. T., Breeveld, A. A., Holland, S. T., Ivanushkina, M., Pryzby, M. S., Still, M. D., & Stock, J. 2005, Space Science Reviews, 120, 95 Salvaterra, R., Della Valle, M., Campana, S., Chincarini, G., Covino, S., D’Avanzo, P., Fern´andez-Soto, A., Guidorzi, C., Mannucci, F., Margutti, R., Th¨one, C. C., Antonelli, L. A., Barthelmy, S. D., de Pasquale, M., D’Elia, V., Fiore, F., Fugazza, D., Hunt, L. K., Maiorano, E., Marinoni, S., Marshall, F. E., Molinari, E., Nousek, J., Pian, E., Racusin, J. L., Stella, L., Amati, L., Andreuzzi, G., Cusumano, G., Fenimore, E. E., Ferrero, P., Giommi, P., Guetta, D., Holland, S. T., Hurley, K., Israel, G. L., Mao, J., Markwardt, C. B., Masetti, N., Pagani, C., Palazzi, E., Palmer, D. M., Piranomonte, S., Tagliaferri, G., & Testa, V. 2009, Nature, 461, 1258 Sari, R. & Piran, T. 1999, ApJ, 520, 641 135 Sari, R., Piran, T., & Narayan, R. 1998, ApJL, 497, L17+ Savaglio, S. 2006, New Journal of Physics, 8, 195 Savaglio, S., Fall, S. M., & Fiore, F. 2003, ApJ, 585, 638 Savaglio, S., Glazebrook, K., Abraham, R. G., Crampton, D., Chen, H., McCarthy, P. J. P., Jørgensen, I., Roth, K. C., Hook, I. M., Marzke, R. O., Murowinski, R. G., & Carlberg, R. G. 2004, ApJ, 602, 51 Savaglio, S., Glazebrook, K., & Le Borgne, D. 2009, ApJ, 691, 182 Schady, P., Mason, K. O., Page, M. J., de Pasquale, M., Morris, D. C., Romano, P., Roming, P. W. A., Immler, S., & vanden Berk, D. E. 2007, MNRAS, 377, 273 Schady, P., Page, M. J., Oates, S. R., Still, M., de Pasquale, M., Dwelly, T., Kuin, N. P. M., Holland, S. T., Marshall, F. E., & Roming, P. W. A. 2009, MNRAS, 1787 Soderberg, A. M., Kulkarni, S. R., Nakar, E., Berger, E., Cameron, P. B., Fox, D. B., Frail, D., Gal-Yam, A., Sari, R., Cenko, S. B., Kasliwal, M., Chevalier, R. A., Piran, T., Price, P. A., Schmidt, B. P., Pooley, G., Moon, D., Penprase, B. E., Ofek, E., Rau, A., Gehrels, N., Nousek, J. A., Burrows, D. N., Persson, S. E., & McCarthy, P. J. 2006, Nature, 442, 1014 Spitzer, Jr., L. & Jenkins, E. B. 1975, ARAA, 13, 133 Stanek, K. Z., Gnedin, O. Y., Beacom, J. F., Gould, A. P., Johnson, J. A., Kollmeier, J. A., Modjaz, M., Pinsonneault, M. H., Pogge, R., & Weinberg, D. H. 2006a, Acta Astronomica, 56, 333 —. 2006b, Acta Astronomica, 56, 333 Stanek, K. Z., Matheson, T., Garnavich, P. M., Martini, P., Berlind, P., Caldwell, N., Challis, P., Brown, W. R., Schild, R., Krisciunas, K., Calkins, M. L., Lee, J. C., Hathi, N., Jansen, R. A., Windhorst, R., Echevarria, L., Eisenstein, D. J., Pindor, B., Olszewski, E. W., Harding, P., Holland, S. T., & Bersier, D. 2003, ApJL, 591, L17 Starling, R. L. C., Vreeswijk, P. M., Ellison, S. L., Rol, E., Wiersema, K., Levan, A. J., Tanvir, N. R., Wijers, R. A. M. J., Tadhunter, C., Zaurin, J. R., Gonzalez Delgado, R. M., & Kouveliotou, C. 2005, A&A, 442, L21 Stern, D., Perley, D. A., Reddy, N., Prochaska, J. X., Spinrad, H., & Dickinson, M. 2007, GRB Coordinates Network, 6928, 1 136 Stocke, J. T. & Rector, T. A. 1997, ApJL, 489, L17+ Tanvir, N., Levan, A., Wiersema, K., Fynbo, J., Hjorth, J., & Jakobsson, P. 2009a, GRB Coordinates Network, 9219, 1 Tanvir, N. R., Fox, D. B., Levan, A. J., Berger, E., Wiersema, K., Fynbo, J. P. U., Cucchiara, A., Kr¨uhler, T., Gehrels, N., Bloom, J. S., Greiner, J., Evans, P. A., Rol, E., Olivares, F., Hjorth, J., Jakobsson, P., Farihi, J., Willingale, R., Starling, R. L. C., Cenko, S. B., Perley, D., Maund, J. R., Duke, J., Wijers, R. A. M. J., Adamson, A. J., Allan, A., Bremer, M. N., Burrows, D. N., Castro-Tirado, A. J., Cavanagh, B., de Ugarte Postigo, A., Dopita, M. A., Fatkhullin, T. A., Fruchter, A. S., Foley, R. J., Gorosabel, J., Kennea, J., Kerr, T., Klose, S., Krimm, H. A., Komarova, V. N., Kulkarni, S. R., Moskvitin, A. S., Mundell, C. G., Naylor, T., Page, K., Penprase, B. E., Perri, M., Podsiadlowski, P., Roth, K., Rutledge, R. E., Sakamoto, T., Schady, P., Schmidt, B. P., Soderberg, A. M., Sollerman, J., Stephens, A. W., Stratta, G., Ukwatta, T. N., Watson, D., Westra, E., Wold, T., & Wolf, C. 2009b, Nature, 461, 1254 Thoene, C., D’Avanzo, P., Fernandez-Soto, A., Campana, S., Antonelli, L. A., Chincar- ini, G., Covino, S., D’Elia, V., Malesani, D., Palazzi, E., Salvaterra, R., & Fiorenzano, A. 2009, GRB Coordinates Network, 9216, 1 Thoene, C. C., de Ugarte Postigo, A., Vreeswijk, P. M., Malesani, D., & Jakobsson, P. 2008a, GRB Coordinates Network, 8058, 1 Thoene, C. C., Levan, A., Jakobsson, P., Rol, E., Gorosabel, J., Fynbo, J. P. U., Jensen, B. L., Hjorth, J., & Vreeswijk, P. 2006, GRB Coordinates Network, 5373, 1 Thoene, C. C., Malesani, D., Vreeswijk, P. M., Fynbo, J. P. U., Jakobsson, P., Ledoux, C., & Smette, A. 2008b, GRB Coordinates Network, 7602, 1 Th¨one, C. C., Wiersema, K., Ledoux, C., Starling, R. L. C., de Ugarte Postigo, A., Levan, A. J., Fynbo, J. P. U., Curran, P. A., Gorosabel, J., van der Horst, A. J., Llorente, A., Rol, E., Tanvir, N. R., Vreeswijk, P. M., Wijers, R. A. M. J., & Kewley, L. J. 2008, A&A, 489, 37 Totani, T., Kawai, N., Kosugi, G., Aoki, K., Yamada, T., Iye, M., Ohta, K., & Hattori, T. 2006, PASJ, 58, 485 Uns¨old, A., Struve, O., & Elvey, C. T. 1930, Zeitschrift fur Astrophysik, 1, 314 137 van Dokkum, P. G. 2001, PASP, 113, 1420 van Marle, A. J., Langer, N., & Garc´ıa-Segura, G. 2007, A&A, 469, 941 van Paradijs, J., Groot, P. J., Galama, T., Kouveliotou, C., Strom, R. G., Telting, J., Rutten, R. G. M., Fishman, G. J., Meegan, C. A., Pettini, M., Tanvir, N., Bloom, J., Pedersen, H., Nørdgaard-Nielsen, H. U., Linden-Vørnle, M., Melnick, J., van der Steene, G., Bremer, M., Naber, R., Heise, J., in’t Zand, J., Costa, E., Feroci, M., Piro, L., Frontera, F., Zavattini, G., Nicastro, L., Palazzi, E., Bennett, K., Hanlon, L., & Parmar, A. 1997, Nature, 386, 686 Vergani, S. D., Petitjean, P., Ledoux, C., Vreeswijk, P., Smette, A., & Meurs, E. J. A. 2009, A&A, 503, 771 Vreeswijk, P., Malesani, D., Fynbo, J., Jakobsson, P., Thoene, C., Sollerman, J., Watson, D., & Milvang-Jensen, B. 2008, GRB Coordinates Network, 8301, 1 Vreeswijk, P. M., Ellison, S. L., Ledoux, C., Wijers, R. A. M. J., Fynbo, J. P. U., Møller, P., Henden, A., Hjorth, J., Masi, G., Rol, E., Jensen, B. L., Tanvir, N., Levan, A., Castro Cer´on, J. M., Gorosabel, J., Castro-Tirado, A. J., Fruchter, A. S., Kouveliotou, C., Burud, I., Rhoads, J., Masetti, N., Palazzi, E., Pian, E., Pedersen, H., Kaper, L., Gilmore, A., Kilmartin, P., Buckle, J. V., Seigar, M. S., Hartmann, D. H., Lindsay, K., & van den Heuvel, E. P. J. 2004, A&A, 419, 927 Vreeswijk, P. M., Ledoux, C., Smette, A., Ellison, S. L., Jaunsen, A. O., Andersen, M. I., Fruchter, A. S., Fynbo, J. P. U., Hjorth, J., Kaufer, A., Møller, P., Petitjean, P., Savaglio, S., & Wijers, R. A. M. J. 2007a, A&A, 468, 83 —. 2007b, A&A, 468, 83 Walborn, N. R., Fullerton, A. W., Crowther, P. A., Bianchi, L., Hutchings, J. B., Pellerin, A., Sonneborn, G., & Willis, A. J. 2002, ApJS, 141, 443 Wang, L., Baade, D., & H¨oflich, G. 2003, Archiv.org, astro-ph/0301266v1 Wiersema, K., Graham, J., Tanvir, N., Rol, E., Levan, A., & Fruchter, A. 2008a, GRB Coordinates Network, 7818, 1 Wiersema, K., Tanvir, N., Vreeswijk, P., Fynbo, J., Starling, R., Rol, E., & Jakobsson, P. 2008b, GRB Coordinates Network, 7517, 1 138 Wiersema, K., Tanvir, N. R., Cucchiara, A., Levan, A. J., & Fox, D. 2009, GRB Coor- dinates Network, Circular Service, 263, 1 (2009), 263, 1 Wilkins, S. M., Bunker, A. J., Ellis, R. S., Stark, D., Stanway, E. R., Chiu, K., Lorenzoni, S., & Jarvis, M. J. 2010, MNRAS, 403, 938 Wilson, O. C. 1939, ApJ, 90, 244 Wolfe, A. M., Gawiser, E., & Prochaska, J. X. 2005, ARAA, 43, 861 Wyithe, J. S. B., Hopkins, A. M., Kistler, M. D., Y¨uksel, H., & Beacom, J. F. 2010a, MNRAS, 401, 2561 Wyithe, S., Oh, S. P., & Pindor, B. 2010b, ArXiv e-prints Xue, R.-R., Fan, Y.-Z., & Wei, D.-M. 2009, A&A, 498, 671 Zitouni, H., Daigne, F., Mochkovich, R., & Zerguini, T. H. 2008, MNRAS, 386, 1597 Vita Antonino Cucchiara Education The Pennsylvania State University State College, Pennsylvania 2005–Present Ph.D. in Astronomy & Astrophysics, expected in August 2010 Area of Specialization: Gamma-ray Bursts Universita’ degli Studi di Milano Milano, (Italy) 1998–2004 M.S. in Physics with distinction in Astronomy&Astrophysics Awards and Honors Zaccheus Daniel Foundation for Astronomical Science Grant 2007, 2008 Research Experience Doctoral Research The Pennsylvania State University 2005–Present Thesis Advisor: Prof. Derek B. Fox This research involved studies of Gamma-ray Bursts Afterglows spectra. Graduate Research The Pennsylvania State University 2005–Present Thesis Advisor: Prof. Derek B. Fox This research involved studies of Gamma-ray Bursts Afterglows spectra. Undergraduate Research Universita’ degli Studi di Milano 1999–2004 Research Advisor: Prof. Stefano Covino Analysis of two GRBs for the VLT photometric archive. Teaching Experience Teaching Assistant The Pennsylvania State University 2008 I taught ASTRO 597 and ASTRO 001.