Gamma-Ray TelescopesENCYCLOPEDIA OF ASTRONOMY AND ASTROPHYSICSGamma-Ray TelescopesGamma-rays are the highest-energy photons in theELECTROMAGNETIC SPECTRUM and their detection presentsunique challenges. On one hand it is easy to detect γ -rays.The interaction cross-sections are large and above a fewMeV the pair production interaction, the dominant γ -rayinteraction with matter, is easily recognized. Gamma-raydetectors were far advanced when the concept of ‘γ -rayastronomy’ was first raised in Phillip Morrison’s seminalpaper in 1958. Indeed it was the expected ease of detectionand the early promise of strong sources that led to thelarge concentration of effort in the field, even before thedevelopment ofX-RAY ASTRONOMY. Today the number ofknown γ -ray sources is well under a few hundred whereasthere are hundreds of thousands of x-ray sources. Whatwent wrong?The answer is simple: the detection of cosmic γ -rayswas not as easy as expected and the early predictionsof fluxes were hopelessly optimistic. Below we considerthe physical phenomenon whereby γ -rays interact withmatter, the basic detector configurations and the detectorswhich have provided the bulk of the observational results;we conclude with a description of the techniques nowunder development.First we provide a few definitions and a few caveats.The term ‘γ -ray’ is a generic one and is used to describephotons of energy from 100 keV (105eV) to >100 EeV(1020eV). A range of 15 decades is more than all the rest ofthe known electromagnetic spectrum. A wide variety ofdetection techniques is therefore necessary to cover thishuge range. We will concentrate on the telescopes inthe somewhat restricted range from 1 MeV to 100 TeV.The lower definition comes from the knowledge that thelower energies are better covered by the techniques usedto cover hard x-ray astronomy and that many authorswould define the γ -ray regime as starting at 0.51 MeV(the rest mass energy of the electron). There are nocredible detections of γ -rays at energies much beyond50 TeV and the ‘γ -ray telescope’ techniques used beyondthese energies are really the same as those used to studychargedCOSMIC RAYS and hence are best discussed underthat heading. We are left with some eight decades whichwe will, somewhat arbitrarily, define as the medium-energy (ME) range from 1 to 30 MeV, the high-energy(HE) range from 30 MeV to 100 GeV and the very-high-energy (VHE) range from 100 GeV to 100 TeV. These rangesare not defined by the physics of their production butby the interaction phenomena and techniques employedin their detection. Thus Compton scattering dominatesthe ME range necessitating the use of satellite-borneCompton telescopes. The HE and VHE ranges use thepair production interaction but in very different ways;HE telescopes identify the electron pair in balloon- orsatellite-borne detectors, whereas VHE detectors detectthe electromagnetic cascade that develops in the Earth’satmosphere.GAMMA-RAYASTRONOMY isstill an observation-dominateddiscipline and the observations have been driven not somuch by the astrophysical expectations (which have of-ten been wrong) as by the experimental techniques, whichhave permitted significant advances to be made in partic-ular energy ranges. Hence the most fruitful observationshave come at energies of 100 MeV; these were originally in-spired by the prediction of the strong bump in the spectraexpected from the decay of π0s that are created in hadroninteractions. This energy region was exploited primarilybecause the detection techniques were simpler and moresensitive. In contrast the ME region has the potential forvery interesting astrophysics with the predicted existenceof nuclear emission lines but the development of the fieldhas been slow because the techniques are so difficult.Peculiarities of γ-ray telescopesThere are several peculiarities that uniquely pertain toastronomy in the γ -ray energy regime. These factors makeγ -ray astronomy particularly difficult and have resulted inthe slow development of the discipline.In every band of the electromagnetic spectrumastronomical telescopes make use of the fact that thecosmic rain of photons can be concentrated by reflectionor refraction, so that the dimensions of the actual photondetector are a small fraction of the telescope aperture.How limited would have been our early knowledge ofthe universe if the optical astronomer had not been aidedby the simple refracting telescope which so increased thesensitivity of the human eye. The radio astronomer, theinfrared astronomer, even the x-ray astronomer, dependson the ability of a solid surface to reflect and, with suitablegeometry, to concentrate the photon beam.Above a few MeV there is no efficient way of reflectingγ -rays and hence the dimensions of the γ -ray detector areeffectively the dimensions of the γ -ray telescope. (As weshall see in a later section this is not the case for ground-based VHE telescopes.) In practice to identify the γ -rayevents from the chargedparticle background it is necessaryto use detectors whose efficiency is often quite low. Henceat any energy the effective aperture of a space-borne γ -ray telescope is seldom greater than 1 m2and often onlya few cm2, even though the physical size is much larger.For instance theCOMPTON GAMMA RAY OBSERVATORY (CGRO)was one of the largest and heaviest scientific satellitesever launched; however, its ME and HE telescopes hadeffective apertures of 1600 cm2and5cm2respectively.Beam concentration is particularly important when thebackground scales with detector area. This is the case withγ -ray detectors which must operate in an environmentdominated by charged cosmic rays.The problem of a small aperture is compounded bythe fact that the flux of cosmic γ -rays is always small. Atenergies of 100 MeV the strongest source (the Vela pulsar)gives a flux of only 1 photon min−1. With weak sourceslong exposures are necessary and one is still dealingwith the statistics of small numbers. It is small wonderthat γ -ray astronomers have been frequent pioneers inCopyright © Nature Publishing Group 2001Brunel Road, Houndmills, Basingstoke, Hampshire, RG21 6XS, UK Registered No. 785998and Institute of Physics Publishing 2001Dirac House, Temple Back, Bristol, BS1 6BE, UK1Gamma-Ray TelescopesENCYCLOPEDIA OF ASTRONOMY AND ASTROPHYSICSthe development of statistical methods and that γ -rayconferences are often dominated by arguments over realstatistical significance.As