A. J. Castro-Tirado¹ , J. Gorosabel¹, A. de Ugarte Postigo¹, M. Jelínek¹, S. Guziy^1,2, S. B. Pandey^1,3, J. M. Castro Cerón⁴ and M. D. Pérez-Ramírez⁵

(1)   Instituto de  Astrofísica de Andalucía, CSIC,   Apt  3004, E-18080 Granada,  Spain
(2)   Nikolaev State University,    Nikolskaya 24, 54030 Nikolaev,  Ukraine
(3)   ARIES Observatory ,    Manora Peak, Nainital, 263129 India
(4)   Niels Bohr Institute, Copenhagen University,    Juliane Maries Vej 33, 2100 Copenhagen,  Denmark
(5)   Universidad de Jaén ,  Virgen de la Cabeza s/n,  Jaén, Spain
 

Introduction

Since their discovery nearly 40 years ago, gamma-ray bursts (GRBs) are still among the greatest puzzles of modern astrophysics. Important advances have been obtained over the past ~ 10 years thanks to the detection of X-ray, optical, millimetre and centimetre radio afterglows, and with the identification of several dozens of host galaxies and about 40 redshift determinations, placing them (at least those lasting longer than 2 s) in the redshift range 0.168-4.50, with the later being the record for more than 5 yr (Andersen et al. 2001).

The nature of the progenitors for this subclass of long-duration GRB is now accepted to be the so-called "collapsars": massive stars ending their lives with the formation of a black hole as a result of core collapse (MacFadyen and Woosley 1999). These very energetic supernovae (or "hypernovae") display a highly collimated emission in a form of a powerful jet. A series of internal shocks produce the observed gamma-ray emission and the resulting relativistic blastwave interacts with the interstellar medium, causing the afterglow phase of GRBs, continuously energizing the swept-up particles and thereby slowing down its own bulk velocity. The afterglow properties are only weakly dependent on the details of the event that produced the blast wave.

In this scenario, it is believed that GRBs can help determining the unobscured star-formation rate in the early universe and in fact, be detected up to very high redshifts ( z = 10-20) and be the signature of the first stellar black holes formed in the Universe following the dearth of Pop III stars (Meszaros and Rees 2003). With the launch of the SWIFT GRB mission in Nov 2004 (2.5 times more sensitive that its predecessor, the BATSE instrument aboard the CGRO ), it was clear that more distant events were going to be detected (Gorosabel et al. 2004).

Observations

On 4 Sep last year, SWIFT detected an unusually long (225 s) GRB starting at 01:51:44 UT. The slowly rising source was detected thanks to the very sensitive imaging triggering software developed at LANL, so a preliminary position could be determined onboard. This prompted inmediate follow-up observations at robotic observatories like BOOTES-1 in South Spain (Castro-Tirado et al. 2005) and TAROT in France (Boer et al. 2006). The initial results were discouraging, as no optical counterpart was detected by the robotic telescopes of BOOTES in south Spain (with R-band images starting 124 s after the onset of the event). However, about 3 hr after the event, near-IR imaging at the 4.1m SOAR telescope in Cerro Tololo (Chile) revealed a bright near-IR counterpart and it was suggested that GRB 050904 could be at high redshift.

Indeed, a set of optical and near-IR images were to confirm this early time suspicions. First, we triggered our target-of-opportunity programme at the 2.2m CAHA and got a sequence of alternating images in two filters (I-R-I-R-I). The initial anaylisis already showed that a faint object was detected with I = 21.9 +/- 0.2 at the position of the reported nIR afterglow (Fig. 1) and that faded by 0.5 mag in the 7 hr time interval during which our images were taken (Fig. 1). At the same time, only un upper limit was obtained in the R-band (> 24.1). Therefore, an indication of R - I > 2.5 already pointed out to the likely existence of the Ly-alpha break sitting on this wavelength range.

Near-IR Observations at ESO's Paranal and La Silla Observatory were also triggered by the MISTICI Collaboration simultaneously to our Calar Alto optical observations. Therefore we decided to join efforts and get the complete SED from the optical to the Ks near-IR band.

Fig. 1: The optical afterglow of GRB 050904, the most distant GRB found so far, at z = 6.29. Color composite image based on the VRI-band data obtained at Calar Alto. The afterglow is the very reddened object at the centre of the image.

Results

Haislip et al. first reported the discovery of the afterglow of GRB 050904, together with the early-time BOOTES observations and the identification of GRB 050904 as the first very high redshift GRB (Haislip et al. 2006).

From the optical and near-infrared observations of the afterglow of the gamma-ray burst GRB 050904 (including the CAHA data), a photometric redshift z = 6.3 +/- 0.1 (Fig. 2), was estimated from the presence of the Lyman break falling between the I and J filters. This is by far the most distant GRB known to date. Its isotropic-equivalent energy is 3.4x10⁵³ erg in the rest-frame 110-1100 keV energy band. Despite the high redshift, both the prompt and the afterglow emission are not peculiar with respect to other GRBs. A break in the J-band light curve was observed at t_b = 2.6 +/- 1.0 d (observer frame). See Fig. 3. If we assume this is the jet break, a beaming-corrected energy E_gamma = (4-12)x10⁵¹ erg is derived. This detection is consistent with the expected number of GRBs at z > 6 and shows that GRBs are a powerful tool to study the star formation history up to very high redshift (Tagliaferri et al. 2005).

Fig. 2: The spectral energy distribution modeling, making use of the Calar Alto, ING and VLT data. A photometric redshift z = 6.30 +/- 0.1 is derived (adapted from Tagliaferri et al. 2005).

On the basis of the internal shock model and synchrotron radiative process and under the assumption that all internal shocks are nearly equally energetic, the emission at different radii, which correspond to different observed times, can be properly modelled. The violently variated X-ray emission and the optical emission of the burst originated from internal shocks, with the energy of ejected shells from the central engine that did not decrease until the X-ray emission faded away. Hence it was suggested that GRB 050904 is a burst with super-long central engine activity (Zou et al. 2006).

The only optical spectrum of the afterglow of GRB 050904 was obtained 3.4 d after the burst (Kawai et al. 2006). The spectrum showed a clear continuum at the long wavelength end of the spectrum with a sharp cutoff at around 9000 A due to Ly-alpha absorption at z = 6.3 with a damping wing. Little flux was present in the waveband shortward of the Ly alpha break. A system of absorption lines of heavy elements at redshift z = 6.295 +/- 0.002 were also detected, yielding a precise measurement of the largest known redshift of a GRB (so far). Analysis of the Si II fine structure lines suggest a dense metal-enriched environment around the GRB progenitor, providing unique information on the properties of the gas in a galaxy when the universe was younger than one billion years.

Modeling analysis of the optical afterglow spectrum taken by the Subaru Telescope, could help to constrain the reionization history. The spectrum showed a clear damping wing at wavelengths redward of the Lyman break, and the wing shape can be fit either by a damped Ly alpha system with a column density of log N_HI ~ 21.6 at a redshift close to the detected metal absorption lines (z_metal = 6.295), or by almost neutral IGM extending to a slightly higher redshift of z_IGM,u ~ 6.36. We then show an evidence that the IGM was largely ionized already at z = 6.3, with the best-fit neutral fraction of IGM, x_HI = 0.00 +/- 0.17, and an upper limit of x_HI < 0.60 (95% C.L.). This is the first quantitative upper limit on x_HI at z > 6. To get further information on the reionization, it is important to increase the sample size of z > 6 GRBs, to find GRBs with low column densities (log N_HI < 20) within their host galaxies, and for statistical studies of Ly alpha line emission from host galaxies. (Totani et al. 2006).

Fig. 3: The optical and near-IR lightcurve of the GRB 050904 afterglow, including the VLT and Calar Alto data (adapted from Tagliaferri et al. 2005).

Conclusions

In 2000, Lamb and Reichart predicted that gamma-ray bursts (GRBs) and their afterglows occur in sufficient numbers and at sufficient brightnesses at very high redshifts (z > 5) to eventually replace quasars as the preferred probe of element formation and reionization in the early universe and to be used to characterize the star-formation history of the early universe, perhaps back to when the first stars formed (Lamb and Reichart 2000). GRB 050904 has been identified as the first very high redshift GRB. These high redshift events will likely drive a new era of study of the early universe, using GRBs as probes.

We expect that Calar Alto will continue as one of the reference astronomical observatories world-wide for the study of the most distant events in the Universe. Prior to the Swift launch in Nov 2004, a significant fraction of events were discovered (and studied) at Calar Alto. But the success rate is not so high after that time, due to the severe limitation on the number of triggers/semester in spite of Swift detecting about 50 GRBs/year now. In order for Calar Alto to continue at the leading edge, we need to significantly increase the number of allotted triggers/semester of 3 to about 10, as is the case at other larger facilities (TNG, WHT, Gemini, VLT) in this Swift Golden Age.

References

Andersen, M. et al. 2000, VLT identification of the optical afterglow of the gamma-ray burst GRB 000131, A&A 364, L54
Boer, M. et al. 2006, ApJ, in press (astro-ph/0510381)
Castro-Tirado, A. J. et al. 2005, Simultaneous and optical follow-up GRB observations by BOOTES, Il Nuovo Cimento 2005, vol. 28, 715
Gorosabel, J. et al. 2004, The potential of INTEGRAL for the detection of high redshift GRBs, A&A 427, 87
Haislip, J. et al. 2006, Discovery and identification of the very high redshift afterglow of GRB 050904, Nature, in press (astro-ph/0509660)
Kawai, N. et al. 2006, Afterglow spectrum of a gamma-ray burst with the highest known redshift z=6.295, Nature, in press (astro-ph/0512052)
Lamb, D. Q. and Reichart, D. E. 2000, Gamma-Ray Bursts as a Probe of the Very High Redshift Universe, ApJ 536, 1
MacFadyen, I. and Woosley, S. E. 1999, Collapsars: Gamma-Ray Bursts and Explosions in "Failed Supernovae", ApJ 524, 262
Meszaros, P. and Rees, M. 2003, Gamma-Ray Bursts as X-Ray Depth Gauges of the Universe, ApJ 591, L91
Tagliaferri, G. et al. 2005, GRB 050904 at redshift 6.3: observations of the oldest cosmic explosion after the Big Bang, A&A 443, L1
Totani, T. et al. 2006, Implications for the Cosmic Reionization from the Optical Afterglow Spectrum of the Gamma-Ray Burst 050904 at z = 6.3, ApJ, in press (astro-ph/0512154)
Zou Y.-C. et al. 2006, Is GRB 050904 a super-long burst?, ApJ, in press (astro-ph/0511205)