GAMMA-RAY BURSTS, OPTICAL FLASHES BRADLEY E. SCHAEFER Gamma-ray bursters (GRBs) are mysterious objects that emit short intense bursts of gamma radiation. The nature of the source and the cause of the bursts are not known, although there are good theoretical and observational reasons to suspect that a neutron star is somehow involved. It is reasonable to expect that at least some small fraction of the burst energy will be emitted as optical light and will appear to astronomers as an optical flash. In 1981, three such flashes were identified on photographs at the Harvard College Observatory. In 1987, a fourth flash was identified on a pair of photographs from Odessa, USSR. After a burst is over, the burster will appear without any disruptive glare, so that it may be possible to discern the nature of the underlying system. Much labor has been expended to find this quiescent counterpart, but to the limits of current technology at all wavelengths, no counterpart has been found. BURSTING COUNTERPARTS A supernova eruption emits only 1/10,000 of its energy as visible light. If a bright GRB emits a similar fraction in the visible range, then this light should appear as a bright flash. This flash may be visible to anyone standing in their backyard. The trick is to be looking at the right time and in the right direction. This is the major trouble because both the times and positions of GRB events are totally unpredictable. It is also unreasonable to pick some random position on the sky and wait for a flash, because they are so rare. For example, a region the size of the full moon has a GRB once every several centuries. One stratagem is to observe a carefully selected direction where a GRB is already known to exist from gamma-ray data. This stratagem relies on the certainty that most GRBs burst repeatedly, perhaps as frequently as once every several years. Since 1985, the most sophisticated flash search system in the world (designed, built, and operated by amateur astronomers with the Santa Barbara Astronomy Group) has used this stratagem. However, observation time can be accumulated only at a rate of 0.1/year for each detector at best. An additional stratagem is to utilize past observations with a large cumulative exposure. This is usually done by examining archival photographs of a region containing a known GRB. These studies have been performed on photographs stored in many observatories worldwide with a cumulative exposure of nearly a decade. In this decade of monitoring time, four GRB optical flashes have been discovered. The first three were found in 1981 on photographs from 1901, 1928 (see Fig. 1), and 1944 in the Harvard collection. The fourth GRB flash was identified in 1987 on a pair of photographs from 1959 stored in Odessa, USSR. The images found on these photographs occur at positions inside the regions on the sky known to contain GRBs. No stars (to very faint limits) appear at the flash positions. Several of the flash photographs are accompanied by other photographs taken immediately before and/or after that show nothing out of the ordinary; hence the "new" images cannot be caused by a passing comet or asteroid. The Odessa flash was recorded on two simultaneous photographs, so plate defects are not a reasonable explanation. Three of the flash photographs are slightly trailed (a condition where normal star images are elongated because of imperfect tracking of the sky motion), yet the flash images are untrailed; hence the flashes must be shorter than several minutes long. In one case, the flash image is off axis and shows a characteristic distortion (coma) due to the telescope optics, demonstrating that the flash light originated in the sky. A study was made where a control region that contained no known GRB positions was examined. The control region (with a cumulative exposure 16 times larger than the GRB search region) had no flash images, hence providing a further connection between the flashes and the GRB phenomenon. The primary difficulty with these historic flash images is that no simultaneous gamma-ray data are available, so the relation between the optical event and the gamma-ray event is unknown. A reasonable presumption is that the optical events in 1901, 1928, 1944 and 1959 were each associated with a gamma-ray event that was comparable to the gamma-ray events later detected by satellites. In this case, the fraction of energy emerging as optical radiation is between 1/1000 and 1/10,000. The utility in finding the four GRB optical flashes is fourfold: 1. First, the fraction of burst energy that appears in the optical can be used to deduce that the GRB system contains something much larger than a neutron star, hence demonstrating that the flash must come from some other body in the GRB system. 2. It demonstrates that the GRBs do repeat on a short time scale (one decade of observations netted four flashes), a fact that is incompatible with many theoretical explanations. 3. The optical photographs yield a position for the burster that is much more accurate than deduced from gamma-ray data, so searches for quiescent counterparts (see below) can go much deeper. 4. The mere existence of the optical flashes can (in some cases severely) constrain GRB models. Two classes of theoretical explanations for the optical flashes have been advanced: The first class claims that the optical light is from the burst's gamma radiation being absorbed and reradiated by a companion star or accretion disk. The second class ascribes the visible photons to cyclotron emission from plasma in a neutron star magnetosphere that was excited by the burst. The observations merely constrain the properties of GRB flashes to be bright and of duration shorter than several minutes. We do not know if the flashes are blue or red in color or whether they last a second or a minute. Unfortunately, the night sky is full of flashes, many of which could be mistaken for GRB events. Background flash sources include fireflys, airplane strobes, head-on meteors, satellite glints, and flare stars. The rates and properties of this background are poorly known. This is illustrated by the series of flashes (the so-called Perseus flasher) seen by Canadian meteor observers from 1984 to 1985 that, only after many long investigations, were shown to be satellite glints (sunlight reflected from orbiting artificial satellites). The identification of GRB flashes is like trying to find a needle in a haystack when we do not know what the needle looks like and we have only just realized that we do not even know what the hay looks like. QUIESCENT COUNTERPARTS During a burst, the emitted light comes from a fireball generated by some unknown energy source. Unfortunately, the properties of this fireball are relatively insensitive to the energy source, so it is difficult to learn about the cause of the burst. After a burst, the GRB should fade back to its normal brightness, so that the underlying system can then be seen. If we could spot even one such quiescent counterpart, then classical astronomical techniques would reveal the nature of the system (e.g., a binary star, a lone neutron star, or a galaxy). This would immediately eliminate many of the proposed GRB models from consideration. Many groups have realized that the firm identification of even one quiescent counterpart is likely to be the break point of our current ignorance. As such, great effort has been spent on determining GRB positions as accurately as possible and then searching for counterparts at any wavelength. Early on, it was realized that GRBs are not coincident with planets, bright stars, pulsars, x-ray sources, or anything else of note. The more accurate the GRB position is known, the deeper a quiescent counterpart search can go. This is because if the GRB is only known to be within some large area on the sky, then there are likely to be many stars (some bright) within that region. The searches are usually based on the idea that the counterpart will look unusual in some way (say, it might appear very blue), but it is impractical to examine large numbers of stars for oddities. So deep searches can only be reasonably performed on small regions. The localization of GRBs on the sky is a difficult problem, and currently only seven have their position determined to within an arcminute and have had deep searches made on them. The 5 March 1979 GRB was the brightest burst ever detected; therefore, its position could be measured with unprecedented accuracy (within 15 arcsec). This direction coincides with that of N49, a supernova remnant in the Large Magellanic Cloud, and so the GRB may be physically associated with N49. If true, this identification would imply an incredibly large burst energy (10**ù* erg) as well as associate the burster with the neutron star presumably inside the supernova remnant. The theoretical and observational arguments for and against the association of this GRB with N49 are still raging. Even if the burst is proven to come from N49, the 5 March 1979 event is so highly unusual in so many ways that conclusions drawn from this event may not be applicable to normal bursters. Besides this one problematic identification, no other reasonable counterparts have been identified (see Fig. 2). At radio wavelengths, over 30 h of observing time with the Very Large Array in New Mexico have revealed no counterparts. For far infrared wavelengths, the Infrared Astronomical Satellite has identified no counterpart for the 23 best localized bursts. Similarly, at near infrared wavelenghts, several GRB locations were found to be empty even when a 4-m telescope with infrared array detectors was employed. At optical wavelengths, one GRB position was even found to be starless to a B and R magnitude of fainter than 25. At x-ray wavelengths, both the Einstein and EXOSAT satellites have spent over 40 h with only one debatable detection. In summary, current technology has been pushed to its limit and no confirmed quiescent counterpart has been found. FUTURE PROSPECTS The primary hope for advances must be based on the introduction of new technology. For the detection of bursting counterparts, an awaited new technology is the all sky flash search system made up of the explosive transient camera (ETC) and the rapidly moving telescope (RMT). The ETC consists of two widely separated arrays of 16 charge coupled device (CCD) cameras each that will record most of the sky down to magnitude 11 at 1-s intervals. When both arrays detect a flash with no parallax (hence ensuring that the flasher is far from Earth), the data will be saved and the RMT will be notified. The RMT is a 7-in. telescope that can slew to any position on the sky within 1 s of notification. The CCD camera on the RMT will measure the flash position to better than 1 arcsec and obtain a light curve when the flash is brighter than 15.2 magnitude. The GRB nature of these flashes will be established by time and directional coincidence with GRBs detected by the Gamma Ray Observatory satellite. For the detection of quiescent counterparts, the primary new technology is the Hubble Space Telescope. This telescope will not only allow much deeper searches of error boxes in visible light, but will for the first time allow deep searches in the ultraviolet. The ultraviolet search capability is important because of the possibility that GRBs are lone, hot, neutron stars and hence visible only in the ultraviolet. Other new important instruments include the burst and transient source experiment to fly on the Gamma Ray Observatory satellite and the Advanced X-ray Astrophysics Facility satellite. All these new detectors will supplement the currently existing systems (such as the Santa Barbara Astronomy Group's telescope network) and the classical astronomy techniques (such as archival photograph searches). All in all, I believe that a counterpart will be confidently identified in the next decade, with the result that the cause and nature of GRBs finally will be known. Additional Reading Hurley, K.(1986). Astronomical issues. In Gamma-Ray Bursts, E.P. Liang and V. Petrosian, eds. American Institute of Physics, New York, p. 3. Maley, P.D.(1987). Specular satellite reflection and the 1985 March 19 optical outburst in Perseus. Ap. J. Lett. 317 L39. Pedersen, H., Danziger, J., Hurley, K., Pizzichini, G., Motch, C., Ilovaisky, S., Gradmann, N., Brinkmann, W., Kanbach, G., Rieger, E., Reppin, C., Trumper, W., and Lund, N.(1984). Detection of possible optical flashes from the gamma-ray burst source GBS 0526-66. Nature 312 46. Schaefer, B.E.(1981). Probable optical counterpart of a gamma-ray burster. Nature 294 722. Schaefer, B.E.(1985). Gamma-ray bursters. Scientific American 252 (No. 2) 52. Schwartz, R.E.(1986). A hunt for flashing stars. Sky and Telescope 72 560.