An important clue for resolving the GRB puzzle will be the detection of transient optical emission associated with the events. Optical emission has been predicted by several different theories [44, 47-50].
The first search for optical emission arising simultaneously with GRBs was performed in 1974 [51]. Later five events were identified for which time-correlated plates were available [52-53]. Very recent results were concerning GRBs detected by BATSE. Plates were found corresponding to more than 30 bursts, but none revealed any interesting brightening in the error box [54]. Thus, there is no positive detections so far. For GRB 910814, a rather deep plate (photographic limiting magnitude of 4.5 for a 1-s flash) was taken by chance. It implied a limit for F(gamma)/F(opt)=42. F(gamma) is the peak gamma-ray flux.
Some authors have assumed that the GRBs could be a repeating phenomenon, and they have used all the available plates in the major plate archives as to look for optical emission in some of the smaller GRB error boxes. Around 50 objects have been found, although most of them have later been rejected as they turned out to be plate defects. There are still few events or Optical Transients (OTs) that could indeed be associated with the GRB sources, but none can be definitively proven as optical emission related to the GRB. See the excellent discussion in [55]. Other OTs could be indeed real and not related to GRBs, like OT 910114, seen by two observers and found in a photographic plate [56] or KY Cep, that appeared in an objective-prism plate taken in 1970 [57]. Researchers have always to be very cautious in order to exclude other effects, like satellite glints, aircraft flashes, double exposures, head-on meteors and asteroids.
Most of them have been performed by wide field instruments. For one GRB observed by WATCH, a Schmidt plate was taken at La Silla 12 h after the event [58]. Other searches have been performed 1-3 d after the burst, but no concluding candidate has yet found [59]. Another result concerns the observation of the error box of GRB 941014, a strong burst observed by BATSE. A CCD image was obtained only 150 s after the burst onset. Again, the result was negative, setting a limit of V = 10.2 to any optical emission arising from this region [60]. A recent observation has been carried by some of us for GRB 970111, detected by SAX with a 10' accuracy. A deep exposure covering the central 16' was obtained by some of us 19 h after the burst. Preliminarly results indicates that no variation below 0.5 mag is seen down to B = 22.7.
Deep searches in small GRB error boxes have been carried out in all wavelenghts, but no quiescent counterpart in the optical, infrared, radio or X-range has been revealed [61- 66]. Only in one case, GRB 920501, was an X-ray source found by ROSAT 14 d after the burst. The X-ray source has been recently observed, but nothing can de concluded [67]. Optical studies of the error box are in progress. Other searches have been carried out in smaller error boxes, like the GRB 790418 error box, down to the detection threshold of B = 25.5 [68]. No definitive candidate has yet been found.
It is clear that one of the problems is the large size of GRB error boxes. Fast determination of GRB positions is very important for optical astronomers as to get CCD images of the field by means of wide-field CCD systems.
Equally important, in order to understand the GRB mystery, may be the observation of very faint bursts, below current levels of detectability. This last goal, however, cannot be achieved from a small satellite.