Next Contents Previous

2.9. Afterglow

GRB observations were revolutionized on February 28, 1997 by the Italian-Dutch satellite BeppoSAX [132] that discovered an X-ray counterpart to GRB970228 [10]. GRB970228 was a double peaked GRB. The first peak which lasted ~ 15 sec was hard. It was followed, 40 seconds later, by a much softer second peak, which lasted some ~ 40 sec. The burst was detected by the GRBM (Gamma-Ray Burst Monitor) as well as by the WFC (Wide Field Camera). The WFC, which has a 40° × 40° field of view detected soft X-rays simultaneously with both peaks. Eight hours latter the NFI (Narow Field Instrument) was pointed towards the burst's directions and detected a continuous X-ray emission. The accurate position determined by BeppoSAX enabled the identification of an optical afterglow [11] - a 20 magnitude point source adjacent to a red nebulae. HST observations [133] revealed that the nebula adjacent to the source is roughly circular with a diameter of 0".8. The diameter of the nebula is comparable to the one of galaxies of similar magnitude found in the Hubble Deep Field, especially if one takes into account a possible visual extinction in the direction of GRB970228 of at least one magnitude [134].

Following X-ray detections by BeppoSAX [10, 135], ROSAT [136] and ASCA [137] revealed a decaying X-ray flux propto t-1.33±0.11 (see Fig. 5). The decaying flux can be extrapolated as a power law directly to the X-ray flux of the second peak (even though this extrapolation requires some care in determining when is t = 0).

Figure 5

Figure 5. Decay of the X-ray afterglow from GRB970228, from [10]. Shown is source flux at the 2-10KeV range. The data is fitted with a power law t-1.32.

The optical emission also depicts a decaying flux [138] (see fig. 6.) The source could not be observed from late March 97 until early September 1997. When it was observed again, on Sept. 4th by HST [139, 131] it was found that the optical nebulosity does not decay and the point source shows no proper motion, refuting earlier suggestions. The visual magnitude of the nebula on Sept. 4th was 25.7 ± 0.25 compared with V = 25.6 ± 0.25 on March 26th and April 7th. The visual magnitude of the point source on Sept. 4th was (V = 28.0 ± 0.25), which is consistent with a decay of the flux as t-1.14 ± 0.05 [131] . In spite of extensive efforts no radio emission was detected and one can set an upper limit of ~ 10µJy to the radio emission at 8.6 Ghz [140].

Figure 6

Figure 6. Decay of the optical afterglow in GRB070228, GRB9700508 and GRB971214. A clear power law decay can be seen in all cases.

Figure 7

Figure 7. Decay of the optical afterglow in GRB9700508. A clear transition from a power law decay to a constant can be seen (from [130].

GRB970508 was detected by both BATSE in gamma-rays [141] and BeppoSAX in X-rays [142] on 8 May 1997. The gamma-ray burst lasted for ~ 15 sec, with a gamma-ray fluence of ~ 3 × 10-6 ergs/cm-2. Variable emission in X-rays, optical [12, 143, 144, 145, 146, 147] and radio [13, 148] followed the gamma-rays. The spectrum of the optical transient taken by Keck revealed a set of absorption lines associated with Fe II and Mg II and O II emission line with a redshift z = 0.835 [14]. A second absorption line system with z = 0.767 is also seen. These lines reveal the existence of an underlying, dim galaxy host. HST images [149, 150] and Keck observations [130] show that this host is very faint (R = 25.72 ± 0.2 mag), compact (leq 1 arcsec) dwarf galaxy at z = 0.835 and nearly coincident on the sky with the transient.

The optical light curve peaks at around 2 days after the burst. Assuming isotropic emission (and using z = 0.835 and H = 100 km/sec/Mpc) this peak flux corresponds to a luminosity of a few × 1045 ergs/sec. The flux decline shows a continuous power law decay propto t-1.27 ± 0.02 [151, 152, 153, 154, 130]. After about 100 days the light curve begun to flatten as the transient faded and become weaker than the host [153, 155, 156, 157]. Integration of this light curve results in an overall emission of a few × 1050 ergs in the optical band. Radio emission was observed, first, one week after the burst [13] (see Fig. 8). This emission showed intensive oscillations which were interpreted as scintillation [158]. The subsequent disappearance of these oscillations after about three weeks enabled Frail et al. [13] to estimate the size of the fireball at this stage to be ~ 1017 cm. This was supported by the indication that the radio emission was initially optically thick [13], which yields a similar estimate to the size [25].

Figure 8

Figure 8. Light curve of the radio afterglow from GRB970508, from [13]

GRB970828 was a strong GRB that was detected by BATSE on August 28, 1997. Shortly afterwards RXTE [159, 160] focused on the approximate BATSE position and discovered X-ray emission. This X-ray emission determined the position of the burst to within an elliptical error box with 5' × 2'. However, in spite of enormous effort no variable optical counterpart brighter than R=23.8 that has changed by more than 0.2 magnitude was detected [161]. There was also no indication of any radio emission. Similarly X-ray afterglow was detected from several other GRBs (GRB970615, GRB970402, GRB970815, GRB980519) with no optical or radio emission.

Seventeen GRBs have been detected with arcminute positions by July 22, 1998: fourteen by the WFC of BeppoSAX and three by the All-Sky Monitor (ASM) on board the Rossi X-ray Timing Explorer (RXTE). Of these seventeen burst, thirteen were followed up within a day in X-rays and all those resulted in good candidates for X-ray afterglows. We will not discuss all those here (see table 1 for a short summary of some of the properties). Worth mentioning are however, GRB971214, GRB980425 and GRB980703.

Table 1. Observational data of several GRBs for which afterglow was detected. The two columns O and R indicate whether emission was detected in the optical and radio, respectively. The total energy of the burst is estimated through the observed fluence and redshift, assuming spherical emission and a flat Omega = 1, Lambda = 0 universe with H0 = 65 Km/sec/Mpc.

  X-ray     gamma-ray fluence   total energy
burst detection O R in [ergs/cm2] redshift in [ergs]

GRB970228 BeppoSAX + - 1 × 10-5 - -
GRB970508 BeppoSAX + + 2 × 10-6 0.835 2 × 1051
GRB970616 BeppoSAX - - 4 × 10-5 - -
GRB970815 RXTE - - 1 × 10-5 - -
GRB970828 RXTE - - 7 × 10-5 - -
GRB971214 RXTE + + 1 × 10-5 3.418 1 × 1053
GRB971227 BeppoSAX - - 9 × 10-7 - -
GRB980326 BeppoSAX - - 1 × 10-6 - -
GRB980329 BeppoSAX + + 5 × 10-5 - -
GRB980425 BeppoSAX + + 4 × 10-6 0.0085 7 × 1047
GRB980515 BeppoSAX - - 1 × 10-6 - -
GRB980519 BeppoSAX + + 3 × 10-5 - -
GRB980703 RXTE + + 5 × 10-5 0.966 1 × 1053

GRB971214 was a rather strong burst. It was detected on December 14.9 UT 1997 [162]. Its optical counterpart was observed with a magnitude 21.2 ± 0.3 on the I band by Halpern , [163] on Dec. 15.47 UT twelve hours after the burst. It was observed one day later on Dec. 16.47 with I magnitude 22.6. Kulkarni et al. [15] obtained a spectrum of the host galaxy for GRB971214 and found a redshift of z=3.418! With a total fluence of 1.09 × 10-5 ergs cm-2 [164] this large redshift implies, for isotropic emission, Omega = 1 and H0 = 65 km/sec/Mpc, an energy release of ~ 1053 ergs in gamma-rays alone 4. The familiar value of 3 × 1053 [15] is obtained for Omega = 0.3 and H0 = 0.55 km/sec/Mpc.

GRB980425 was a moderately weak burst with a peak flux of 3 ± 0.3 × 10-7 ergs cm-2 sec-1. It was a single peak burst with a rise time of 5 seconds and a decay time of about 25 seconds. The burst was detected by BeppoSAX (as well as by BATSE) whose WFC obtained a position with an error box of 8'. Inspection of an image of this error box taken by the New Technology Telescope (NTT) revealed a type Ic supernova SN1998bw that took place more or less at the same time as the GRB [151]. Since the probability for a chance association of the SN and the GRB is only 1.1 × 10-4 it is likely that this association is real. The host galaxy of this supernova (ESO 184-G82) has a redshift of z = 0.0085 ± 0.0002 putting it at a distance of 38 ± 1 Mpc for H = 67 km/sec Mpc. The corresponding gamma-ray energy is 5 × 1047 ergs. With such a low luminosity it is inevitable that if the association of this burst with the supernova is real it must correspond to a new and rare subgroup of GRBs.

GRB980703 was a very strong burst with an observed gamma-ray fluence of (4.59 ± 0.42) × 10-5 ergs cm-2 [166]. Keck observations revealed that the host galaxy has a redshift of z = 0.966. The corresponding energy release (for isotropic emission, Omega = 0.2 and H0 = 65 km sec-1 / Mpc) is ~ 1053 ergs [129].

4 This value depends also on the spectral shape of the burst. Back.

Next Contents Previous