An important reason for the long-standing mysteries surrounding the gamma-ray burst (GRB) sources has been the lack of prompt accurate locations to look for quiescent or fading counterparts. This situation changed dramatically in 1997 with the rapid and accurate location of X-ray emission detected during GRBs using the wide-field camera on the Satellite per Astronomia X (BeppoSAX ), the All Sky Monitor on the Rossi X-ray Timing Explorer (RXTE ), and from the detection of X-ray afterglow in scans of BATSE GRB error boxes with the PCA on RXTE .
These localized X-ray counterparts have led to intense multiwavelength campaigns. Optical transients have been found to some but not all of the bursts with fading X-ray counterparts. The reason optical transients have not been discovered for the other bursts with X-ray counterparts (some of which are much brighter during the gamma-ray burst itself) is that the searches have not always been sensitive enough, particularly given the possibility of absorption local to the source (Groot et al. 1998a).
In spite of detailed searches, radio emission has only been detected for of the bursts. The first detection of a variable radio source was GRB 970508 (Frail et al. 1997; Taylor et al. 1997; Galama et al. 1998a). This mJy source had a rising spectral index, and its rapid radio variability was most likely caused by interstellar scintillation. The combined observations of GRB 970508 showed that the peak of the spectrum was in the sub-millimeter region days after the burst (Galama et al. 1998b) with the radio, millimeter, and optical emission peaking days to weeks after the burst (Frail et al. 1997; Galama et al. 1998a; Gruendl et al. 1998; Bremer et al. 1998; Pian et al. 1998; Pedersen et al. 1998; Castro-Tirado et al. 1998a). Even before any counterparts were found, two completely separate classes of models had suggested this would be the case: (1) cosmological fireball models (e.g. Paczynski & Rhoads 1993; Katz 1994; Mészáros & Rees 1997), and (2) Compton scattering models (e.g. Liang et al. 1997; Liang 1997).
For GRB 970508 there are at least two breaks in the spectrum between and Hz (Galama et al. 1998b) making this a crucial region for the models. To obtain a complete picture of the nature of the burst counterparts, it is clear that one needs to cover the entire spectrum, and sub-millimeter observations with a mJy sensitivity are needed. This is particularly important since the optical emission can be suppressed by local absorption, and the radio emission at frequencies below 20 GHz can be self absorbed as well as scrambled by interstellar scintillation (Walker 1998), while "clean" observations ought to be possible at sub-millimeter wavelengths (the radio interstellar scintillation frequency depends on the Galactic latitude of the source, and can be well below 20 GHz; D. Frail, private communication).
In this paper we discuss our ongoing program of Target of Opportunity observations using SCUBA on the James Clerk Maxwell Telescope. In Sect. 2 we discuss some of the technical features of SCUBA that make it well suited for performing counterpart searches. In Sect. 3 we present the results of our observations to date on GRBs 970508 (which was a limited trial run), 971214, 980326, 980329, 980519, 980703, 981220, and 981226. Our SCUBA observations of GRB 990123 are presented elsewhere (Galama et al. 1999). In Sect. 4 we give a brief discussion.
© European Southern Observatory (ESO) 1999
Online publication: June 18, 1999