Introduction

In 1967-73, the four VELA spacecraft (named after the spanish verb {velar}, to keep watch), that where originally designed for verifying whether the former Soviet Union abided by the Limited Nuclear Test Ban Treaty of 1963, observed 16 peculiarly strong events.

On the basis of arrival time differences, it was determined that they were related neither to the Earth nor to the Sun, but they were of cosmic origin. Therefore they were named cosmic Gamma-Ray Bursts (GRBs hereafter).

GRBs appear as brief flashes of cosmic high energy photons, emitting the bulk of their energy above 0.1 MeV. They are detected by instruments somewhat similar to those used by the particle physicists at their laboratories. The difference is that GRB detectors have to be placed onboard balloons, rockets or satellites.

Gamma-ray bursts and Optical Transients

It is well known that an important clue for resolving the gamma-ray burst (GRB) puzzle is the detection of transient optical emission associated with the bursts. This might be favoured by an optimum system capable of performing rapid movements towards the position of the event. After years of search, the first fading optical counterparts were found in 1997 beggining 3-20 hr after the onset of the high energy events thanks to the accurate positions provided by the satellites BeppoSAX and RossiXTE. This has led to prove their extragalactic origin.

An additional mistery surrounds the reliability of Optical Transients (OTs), as some of them could be related to GRBs. Under the assumption of the bursters being a repeating phenomena, archival plates have been used in order to look for optical transient emission in the smallest GRB error boxes, and about 50 candidates have been identified so far, but most of them were rejected as they turned out to be plate defects. Nevertheless, there are still few events (OTs) that might be associated with the GRB sources, but with the exception of the V = 9 mag optical transient discovered by Rotse simultaneously to GRB 990123, none of them can be definitively proven as optical emission related to the GRB. Three could be related to underlaying faint Active Galactic Nuclei (AGNs). In fact, a test was performed on the variable source PKS 0420-014, and was detected several times as short optical flashes lasting less than one day. Most (82%) blazars show variability on time scales of days or less; however the available observations are not dense enough to resolve the faster variations.

If the OT connection to QSO/AGNs proves solid, it will substantially extend the known amplitude of AGN variability. There are hints that rapid and sudden flares may occur, though of much smaller amplitude.

The Explosive Transient Camera (ETC), a system with 16 wide-field cameras, each with a field of view of 20 x 15 deg² recorded about 100000 flashes brighter than 11 mag. in 2.5 yr of operation in Kitt Peak. Most of the events were due to satellite glints, but about 0.5% remain unidentified This is the fraction that might be recorded with the Optical Monitoring Camera (OMC) on the European Space Agency’s INTEGRAL satellite, when launched in September 2002.

General properties

Temporal Properties

The time profiles of the bursts are very different, with some GRBs lasting a few ms and others lasting for several minutes.  In general, there was no evidence of periodicity in the time histories of GRBs.  However there was indication of a bimodal distribution of burst durations, with 25 % of bursts having durations around 0.2 s and 75 % with durations around 30 s.

 

Distribución bimodal de los GRBs presentes en el catálogo 4B de BATSE

GRB Location

The KONUS instrument on Veneras 11 and 12 gave the first indication that GRB sources were isotropically distributed in the sky.  Based on a much larger sample, this result was nicely confirmed by BATSE on board the CGRO satellite, launched in the spring of 1991, an instrument that has revolutioned the GRB field. About 800 GRBs are detected on a yearly basis, but only few of them are localized accurately. The apparent isotropy of the bursts in the sky ruled out the models dealing with neutron stars in the Galactic Plane, and it was rather interpreted in terms of GRBs arising at cosmological distances, although the possibility of a small fraction of the sources lying nearby, within a galactic disc scale of few hundred pc, or in the halo of the Galaxy, could not be discarded by that time.

Another result was that the time profiles of the bursts are very different, with some GRBs lasting a few ms and others lasting for several minutes.  In general, there was no evidence of periodicity in the time histories of GRBs.  However there was indication of a bimodal distribution of burst durations, with about 25 % of bursts having durations around 0.2 s and about 75 % with durations around 30 s.

 

Isotropic distribution of GRBs in the sky (galactic coordinates)

 

A deficiency of weak events was noticed in the log N-log S diagram, as the GRB distribution deviates from the -3/2 slope of the straight line expected for an homogeneous distribution of sources assuming an Euclidean geometry.

All these observational data led many researchers to believe that GRBs are indeed at cosmological distances. However, the origin of GRB remains unknown for more than 30 years.

 

Diagrama log N-log S

 

It was well known that an important clue for solving the GRB puzzle was going to be the detection of transient emission -at longer wavelengths- associated with the bursts.

Optical counterparts

After the first trials in the 1980’s for idfentifying optical transients in plate archives, the end of the century meant a breaktrough in the GRB puzzle.

With the advent of the X-ray satellites BeppoSAX and RossiXTE, it has been possible to carry out deep multi-wavelength observations of the counterparts associated with the long GRBs class just within a few hours of occurence, thanks to the observation of the fading X-ray emission that follows the more energetic gamma-ray photons once the GRB event has ended. The fact that this emission (the afterglow) extends at longer wavelengths, has led to the discovery optical/IR/radio counterparts in 1997-2000, greatly improving our understanding of these sources.

 

The R-band light curve of the GRB 970508 optical counterpart

BeppoSAX  made possible to detect the first X-ray afterglow following GRB 970228 whose precise localization  led to the discovery of the first optical transient (or optical afterglow, OA) associated to a GRB.

The light curve exhibited a power-law decay with the flux proportional to  F  t^(-alpha} with  alpha =  1.1. PL declines have been measured for 26 OAs in 1997-2000 yielding values in the range  0.8 <  alpha <  2.3 with  = 1.35.

GRB 970508 was the clue to the distance: optical spectroscopy obtained during the OA maximum brightness allowed a direct determination of a lower limit for the redshift (z > 0.835), implying D > 4 Gpc  and E >7 x 10^51 erg.

It was the first proof that GRB sources lie at cosmological distances. The flattening of the decay at T_0 + 100 d revealed the contribution of a constant brightness source -the host galaxy- seen in late-time imaging at T_0 + 1 yr.

The 15 GRB redshifts measured so far are in the range 0.430 < z < 4.50  with = 1.5 and they were derived either from absorption lines in the OA spectrum, from the Lyman-alpha line break, or from emission lines arising in the host galaxy.

Significant early optical emission may arise from the reverse shock, i.e. strong optical flashes accompanying gamma-ray emission should be a generic characteristic (at least for typical GRBs, with E = 10^53 erg and the density rho = 1 cm^(-3)).

Such observations will allow: i) to derive the Lorentz factor by the relative timing of optical and  gamma-ray emission, ii) to pinpoint the process by which the shells responsible for the external shock arise, and iii) to constraint the environment.

The ROTSE experiment achieved the detection simultaneously to the GRB of the bright optical emission from GRB 990123: the most luminous object ever recorded, with M_V = -36  (peaking at m_V = 8.9), implying that at least some subsets of GRBs do exhibit variable optical emission as violent as the gamma-ray variations.

The first optical/near-IR counterparts have been found for 30 precisely localized GRBs in 1997-2000.

In any case, only the population of GRBs with durations of few seconds has been explored. Short bursts lasting less than 1 s, that follow the -3/2 slope in the log N-log S diagram (in contrast to the longer bursts) remain to be detected at longer wavelengths.

Future missions should be able to address some of the issues still to be solved, i.e. prompt optical observations should be persued !

 

Contrapartida Discovery images of the GRB 980703 optical counterpart at the spanish 0.8-m IAC telescope

 

Theorical models

The most popular models fall into two broad cathegories: the explosion of a massive star and the coalescence of a compact binary system.

The “collapsar” (or hypernova) model deals with a rotating massive star with a Fe core that collapses forming a rotaing  black hole (Kerr BH) and a 0.1-1 solar mass torus.

The matter is accreted at a very high rate and the energy is released amounting up 10^{54} erg. A “dirty fireball”, is produced reaching a luminosity  300 times larger that than of a normal SN.

This would happen every  10$^6 yr on average. In this scenario, GRBs would be produced in dense enviroments near star forming regions and GRBs might be used for deriving the star-forming rate in the Universe.

The merging of a neutron star binary sistem giving rise to a GRB

The coalescence of neutron stars in a binary system has been also proposed (Narayan et al. 1992): lifetimes of such systems are of the order of  10^9 years, and large escape velocities are usual, putting them far away from the regions where their progenitors were born. The likely result is a Kerr BH, and the energy released energy during the merger process is 10^54 erg. It is also possible that a 0.1 solar mass accretion disk forms around the black hole and is accreted within a few dozen seconds, then producing internal shocks leading to the GRB.

In this scenario, GRBs would be produced far away from their host galaxies, and this could account for the 40 % of bursts not located in the optical window.

There are variations of this latter model where one or two components are substituted for black holes, white dwarfs or He stars.

A statitiscal study of the offsets of 20 long-duration GRBs from their apparent host galaxies centers favours the explosion of a massive star rather than the binary merger model.

It has been suggested that the short duration (< 1 s) bursts could be due to compact star mergers, whereas the longer ones are caused by the collapse of massive stars.