Frontiers: Gamma-Ray BurstsWe will now take a look at one of the most exciting and controversial fields in allastrophysics: gamma-ray bursts. These events may have a higher peak photon lumin ositythan any other events in the universe, and their physics is therefore extreme enough tohave motivated a number of exotic suggestions. Their spectra and brightness also maymake them great backlighting for the universe, meaning that in p rinciple absorption linesin their spectra can tell us about the composition and evolution of the z ∼ 5 − 20 universe(although currently no GRBs have been established beyond z ≈ 6.3). The history of thisfield is also an object lesson in how new evidence can shift opinions dramatically. Atone time or anoth er, a substantial majority of people in the field have believed (with afair degree of certainty) that gamma-ray bursts are (1) in the Galactic disk, (2) standardcandles, (3) caused by merger and inspiral of two compact objects, and (4) the product of aspecial type of supernova. Current opinion favors (4) for one of the major types of GRB,and probably (3) for the other, but new evidence can always change this. To me, the studyof gamma-ray bursts encapsulates much of what makes th e scientific process unique. It is asubject filled with rancor and conflict, but the emergence of n ew data has had its say in away not available with pure philosophy.Brief summary of propertiesLet’s first summarize briefly what gamma-ray bursts are. Indeed, it is somewhatdifficult, because unlike many of the sources and phenomena we’ve discussed, gamma-raybursts are rather heterogeneous in their properties. The first two properties come from theirname: the emission is primarily in gamma rays (with a spectral νFνpeak in the hundredsof keV), and the events have a limited duration (from milliseconds to about a thousandseconds, as seen so far). There seems to be a broad bimodal distribution of durations, onepeak being less than a second and the other being at 10-20 seconds. Unlike X-ray bursts,the profile of the flux with time is not universal. Many bursts have a “FRED” profile (fastrise, exponential decay), but others are more spiky, or have some emission, a long quiescentperiod, and then have more emission. Within the sensitivity of current instruments, thereappears to be of order a few bursts per day in the universe, of which perhaps 10-20% are ofthe short hard variety.It appears at this time that the distribution of locations of bursts on the sky is consistentwith isotropic, although occasional evidence for weak clustering in a subset of bursts isreported. There is also no definitive evidence that any bur st has repeated, although someevents are consistent with a repetition of up to four events. The positional uncertaintiesfor most detected bursts are large (at least several degrees), although detections fromBeppoSAX, HETE-2 and SWIFT can do much better than that. This is why statementsabout isotropy and repetition are difficult to make. The flux observed at Earth has anextremely broad range between different burst, from a maximum of about 10−3erg cm−2s−1to the flux limits of detectors, down to 10−8erg cm−2s−1. All bursts that have beenlocalized enough for pointed follow-up have X-ray afterglows lasting days (before they aretoo weak to detect), and about half have detectable optical afterglows. The spectrum andthe time development of the bursts are adequately described by power laws with a fewbreaks in them. Redshifts (or at least lower limits to the redshift) have been obtained for anumber of bursts, clearly indicating that many, perhaps all, bursts are at large cosmologicaldistances.History of detectionGamma-ray bursts were first discovered as a byproduct of the Cold War. In the late1960s there was a concern that the Soviets might test nuclear weapons in space. The USdecided that it needed to be able to detect the gamma-ray emission that would result, andit therefore launched the Vela series of satellites. They were alarmed when, starting in 1968,the satellites detected gamma-ray flashes from space! The spatial resolution of the satelliteswas poor, but eventually it was determined that the flashes came from outside the solarsystem, so in 1973 the flashes were reported publicly.In 1979 there was an apparent breakthrough in the study of gamma-ray bursts. OnMarch 5, 1979, nine separate satellites detected a remarkably strong burst (impressiveenough that this is simply known as the “March 5 event”). Many of these satellites were farenough away from the Earth that it was possible to localize the direction of this event bytiming; an aid to this localization was that the event had an extremely sharp onset. Thisevent came from the N49 supernova remnant in the Large Magellanic Cloud, and later waseven more specifically determined to come from an X-ray hot spot in the cloud. This wasexciting, because this was the first time that a GRB had been identified with a quiescentsource. Moreover, this source repeated; 16 more bursts were seen over the following months.However, it is now thought that this event was the first identified member of a separateclass, soft gamma-ray repeaters. At the time, though, this mislead people for a long time,because it appeared that this was clear evidence for a Galactic source of the bursts, and itwas so clearly established that it appeared to be a fixed point in the data.In the 1980s, other bits of evidence appeared to support the local origin of the bursts.Data from the Japanese satellite Ginga for several bursts suggested the existence ofcyclotron absorption-like features in three bursts, one that appeared very secure. Thisalso seemed to argue strongly for a relatively local origin. The point is that without anypersistent sources or direct evidence of distance, a given flux is not informative about thedistance (in the dark, a light could be a nearby firefly or a distant airplane). However, theargument was that if the distance was cosmological, the luminosity would be so high as toprevent the formation of lines near a compact object. At the end of the 1980s, virtuallythe entire community (with the notable exception of Bohdan Paczynski) was sure thatgamma-ray bursts mostly came from neutron stars in the disk of the Galaxy.In 1991, the Compton Gamma-Ray Observatory was launched, as one of NASA’sGreat Observatories program. The Burst and Transient Source Experiment (BATSE) wasparticularly well-suited for detection of GRBs,