3. EXTRAGALACTIC TEV SOURCES

Arguably the most exciting results to come from the search for TeV sources of gamma rays has been the discovery that many of the nearby blazars are detectable sources. This was certainly not anticipated although some of the earliest TeV observations were directed towards radio galaxies (10) and quasars (11). The detection of 3C273 by COS B should perhaps have alerted TeV observers to the possibility that AGN might be denizens of the TeV cosmic zoo. However the spectrum was rather soft and did not suggest that even this, the closest quasar, would be detectable at TeV energies. By and large, the AGN community was equally unexcited by this detection and the discovery by EGRET that the 100 MeV sky was dominated by blazars came as a real surprise to the larger high energy astrophysics community.

The fact that these EGRET AGN all had flat spectra with little sign of a flattening above 10 GeV was auspicious for TeV astronomy. However no firm predictions were made of the fluxes to be expected and early observations of a selection of these sources by ground-based observers had disappointing results (12). It was not until the detection of Markarian 421 (Mrk421) (13) that it became apparent that AGN might be an important TeV phenomenon.

By the time AGN had been established as MeV-GeV sources by EGRET, the sensitive atmospheric Cherenkov imaging technique had been developed. Given the high degree of variability now detected in TeV sources it is perhaps fortunate that the first source seen at VHE energies was the Crab Nebula (14), a notoriously steady source. One can only speculate at the controversy that would have ensued if the highly variable Mrk421 had been the first source reported with the new VHE techniques!

In all, ten extragalactic objects have been reported as sources of VHE gamma rays and their properties are summarized in Table 2. The integral fluxes that were reported in the detection papers (referenced in the second column), are quoted above the peak response energy (E_peak) at which they were detected (as given in the last column). Their positions are shown in Table 3. It should be noted that only six of the AGN entries have been independently confirmed at the 5 level and hence are classified as A sources. All of the VHE blazars detected to date are relatively close-by with redshifts ranging from 0.031 to 0.129, and perhaps to 0.444; they are all members of the BL Lacertae blazar subclass (BL Lacs). These objects are characterized by few or no emission lines and are labeled Low frequency peaked (LBL) or High frequency peaked (HBL) depending upon the waveband, radio or X-ray, in which their detected synchrotron emission peaks.

Accurate derivation of the VHE spectrum is important for many reasons. Since these are the most energetic photons detected from blazars, the shape of their spectrum is an important input parameter to emission models and can therefore impose severe constraints on them. Multi-wavelength campaigns involving TeV observations which provide information on how the TeV spectrum varies with flux constrain key model input parameters. By comparing such variations to those at longer wavelengths, especially when the observations are simultaneous, much can be inferred about the location and the nature of the progenitor particles. Spectral features, such as breaks or cut-offs, can indicate changes in the primary particle distribution or absorption of the gamma rays via pair-production with low energy photons at the source or in intergalactic space.

The high flux VHE emission from Markarian 501 (Mrk501) in 1997 (25); (26); (27); (28) and Mrk421 in 2001 (29); (30); (31) has permitted detailed spectra to be extracted. Measurements are possible over nearly two decades of energy. As many as 25,000 photons were detected in these outbursts so that the spectra were derived with high statistical accuracy. Unlike the HE sources where the photon-limited blazar measurements are consistent with a simple power law, there is definite structure seen in the VHE measurements (27); (32) with evidence for an exponential cut-off in the spectra of some blazars. For Mrk421, this cut-off is 4 TeV and for Mrk501 it is 3-6 TeV. The coincidence of these two values suggests a common origin, i.e., a cut-off in the acceleration mechanisms in the blazars or perhaps the effect of the infra-red absorption in extragalactic space. Attenuation of the VHE gamma rays by pair-production with background infra-red photons could produce a cut-off that is approximately exponential. Indeed, consistent with this expectation, the spectrum of the more distant blazar H1426+428 shows evidence for spectral flattening at energies above 1 TeV (33).

Table 2: Extragalactic TeV Sources: Gamma-ray Properties

Source (Det. Paper)		Class	F (mean)	F (Det.)	Epeak
			> 100 MeV	> Epeak	(Det.)
			10-8cm-2s-1	10-12cm-2s-1	TeV
NGC 253	(15)	Starburst Gal.	U.L.	7.8	0.52
3C66A	(16)	BL Lac(LBL)	18.7	30.0	0.90
Mrk421	(13)	BL Lac(HBL)	13.9	15.0	0.50
M87	(17)	Radio Galaxy	U.L.	1.0	0.73
H1426+428	(18)	BL Lac(HBL)	U.L.	20.4	0.28
Mrk501	(19)	BL Lac(HBL)	U.L.	8.1	0.30
1ES1959+650	(20) †	BL Lac(HBL)	U.L.	29.4	0.60
PKS2155-304	(21)	BL Lac(HBL)	13.2	42.0	0.30
BL Lacertae	(22)	BL Lac(LBL)	11.1	21.0	1.00
1ES2344+514	(23)	BL Lac(HBL)	U.L.	11.0	0.35
† No flux was quoted in the initial detection paper (20); the flux from (24) is quoted here.