Next Contents Previous


4.1. Active Galactic Nuclei

In recent years, high-energy gamma-rays have come to play an important role in the study of AGNs. Before the launch of the CGRO in 1991, the only known extragalactic source of high-energy gamma-rays was 3C 273, which had been detected with the COS B satellite 20 years ago (Swanenburg et al. 1978). The EGRET detector on the CGRO has identified more than 65 AGNs which emit gamma-rays at energies above 100 MeV (Hartman et al. 1999), and a substantial fraction of those sources which remain unidentified in the EGRET catalog are likely to be AGNs as well. In addition, the Whipple Observatory gamma-ray telescope has discovered three AGNs which emit at energies above 300 GeV (Punch et al. 1992; Quinn et al. 1996; Catanese et al. 1998), and there are recent detections of two other AGNs with Cerenkov telescopes (Chadwick et al. 1999; Neshpor et al. 1998). During flaring episodes, the gamma-ray emission can greatly exceed the energy output of the AGNs at all other wavelengths. Thus, any attempt to understand the physics of these objects must include consideration of the gamma-ray emission.

All of the AGNs detected in high-energy gamma-rays are radio-loud sources with the radio emission arising primarily from a core region rather than from lobes. These types of AGNs are often collectively referred to as "blazars" and include BL Lacertae (BL Lac) objects, flat-spectrum radio-loud quasars (FSRQs), optically violent variables, and superluminal sources. The emission characteristics of blazars include high polarization at radio and optical wavelengths, rapid variability at all wavelengths, and predominantly nonthermal emission at most wavelengths. The emission from blazars is believed to arise from relativistic jets oriented at small angles to our line of sight. If so, the observed radiation will be strongly amplified by relativistic beaming (Blandford & Rees 1978). Direct evidence for relativistic beaming of the radio emission comes from very long baseline interferometer (VLBI) observations of apparent superluminal motion in many blazars (e.g., Vermeulen & Cohen 1994). The rapid variability and high luminosities of the detected gamma-ray sources imply that the gamma-rays are also beamed (see Section 4.1.3).

There is a growing consensus that blazars are all the same type of object, perhaps differing only in intrinsic luminosity (e.g., Fossati et al. 1998; Ghisellini et al. 1998) or some combination of luminosity and viewing angle (e.g., Georganopoulos & Marscher 1998). However, for this work we will continue the practice of referring to BL Lac objects and FSRQs as distinct objects: BL Lac objects are those blazars which have optical emission lines with equivalent width less than 5 Å, and FSRQs are the remaining blazars. As we will see, this distinction may be important in explaining why only BL Lac objects are detected at very high energies.

The spectral energy distribution of blazars appears to consist of two parts. First, a low-energy component exhibits a power per decade distribution that rises smoothly from radio wavelengths up to a broad peak in the range spanning infrared (IR) to X-ray wavelengths, depending on the specific blazar type, above which the power output rapidly drops off. Second, a distinct, high-energy component, which does not extend smoothly from the low-energy component, is often seen. It typically becomes apparent in the X-ray range and has a peak power output in the gamma-ray range between ~ 1 MeV and 1 TeV (e.g., von Montigny et al. 1995), again depending on the specific blazar type. When plotted as E2 dN / dE (or equivalently nu Fnu) the spectral energy distribution shows a two-humped shape, although some objects show evidence of a third component (e.g., Kubo et al. 1998).

Although there is no general consensus on the origin of these emission components, it is generally agreed that the low-energy component arises from incoherent synchrotron emission by relativistic electrons within the jet (e.g., Blandford & Rees 1978). This is supported most strongly by the high-level, variable polarization observed in these objects at radio and optical wavelengths. The origin of the high-energy emission is a matter of great interest. There are many variations of the models, and here we only briefly mention a few which are most often invoked to explain the gamma-ray emission. The most popular models at this time are those in which the gamma-rays are produced through inverse Compton scattering of low-energy photons by the same electrons which produce the synchrotron emission at lower energies. Synchrotron self-Compton (SSC) emission (e.g., Königl 1981; Maraschi, Ghisellini, & Celotti 1992; Bloom & Marscher 1996), in which the seed photons for the scattering are the synchrotron photons already present in the jet, must occur at some level in all blazars, but models in which the gamma-ray emission arises predominantly from inverse Compton scattering of seed photons which arise outside of the jet, either directly from an accretion disk (Dermer, Schlickeiser, & Mastichiadis 1992) or after being reprocessed in the broad-line region or scattering off thermal plasma (Sikora, Begelman, & Rees 1994), appear to fit the observations satisfactorily as well. Another set of models proposes that the gamma-rays are produced by proton-initiated cascades (e.g., Mannheim 1993). As we will show in Section 4.1.3, the gamma-ray observations strain both types of models but do not, at present, rule any out.

In the remainder of this section we discuss the status of VHE observations and the emission characteristics of the detected objects (Section 4.1.1) and the results of multiwavelength campaigns on the detected objects (Section 4.1.2) which are the best probe of the physics of blazars, and finally we briefly discuss some of the implications of these observations on our understanding of the physics of the blazars and on the models which purport to explain them.

4.1.1. Observational Status and Emission Characteristics

The BL Lac object Mrk 421 (z = 0.031) was detected as the first extragalactic source of VHE gamma-rays in 1992 using the Whipple Observatory gamma-ray telescope (Punch et al. 1992). A two-dimensional image of the emission from Mrk 421 is shown in Figure 6. Although Mrk 421 had previously been part of an active program of observing extragalactic sources (Cawley et al. 1985) by the Whipple Collaboration, the observations which led to the detection of Mrk 421 at TeV energies were initiated in response to the detection of several AGNs by the EGRET experiment. The initial detection indicated a 6 sigma excess, and the flux above 500 GeV was approximately 30% of the flux of the Crab Nebula at those energies. Mrk 421 has been confirmed as a source of VHE gamma-rays by the HEGRA Collaboration (Petry et al. 1996), the Telescope Array Project (Aiso et al. 1997), and the SHALON telescope (Sinitsyna et al. 1997), and, as discussed in Section 4.1.2 below, multiwavelength correlations have confirmed that the VHE source is indeed Mrk 421 and not some other object.

Figure 6

Figure 6. Two-dimensional plot of the VHE gamma-ray emission from the region around Mrk 421. The gray scale is proportional to the number of excess gamma-rays, and the solid contours correspond to 2 sigma levels. The dashed ellipse give the 95% confidence interval determined by EGRET (Thompson et al. 1995). The position of Mrk 421 is indicated by the cross. Figure from Buckley et al. (1996).

With the successful detection of Mrk 421, the Whipple Collaboration initiated a search for VHE emission from several other blazar-type AGNs, concentrating at first on those objects detected by EGRET, but also spending a substantial amount of time observing radio-loud blazars which were not detected by EGRET. This broad approach led to the detection of the BL Lac object Mrk 501 (z = 0.034) by the Whipple Collaboration in 1995 (Quinn et al. 1996). A two-dimensional map of the VHE emission from Mrk 501 is shown in Figure 7. Because Mrk 501 had not been detected as a significant source of gamma-rays by EGRET, this was the first object to be discovered as a gamma-ray source from the ground. Hence, VHE gamma-ray astronomy was established as a legitimate channel of astronomical investigations in its own right, not just an adjunct of high-energy observations from space. The flux of Mrk 501 during 1995 was, on average, 10% of the VHE flux of the Crab Nebula, making it the weakest detected source of VHE gamma-rays. Mrk 501 was confirmed as a source of VHE gamma-rays in 1996 by the HEGRA telescopes (Bradbury et al. 1997) the CAT telescope (Punch 1997), the Telescope Array Project (Hayashida et al. 1998), and TACTIC (Bhat 1997).

Figure 7

Figure 7. Two-dimensional plot of the VHE gamma-ray emission from the region around Mrk 501. The gray scale is proportional to the number of excess gamma-rays, and the solid contours correspond to 1 sigma levels. The position of Mrk 501 is indicated by the cross.

In addition to the confirmed detections of Mrk 421 and Mrk 501, three other objects have recently been reported as sources of VHE gamma-rays but remain to be verified by detections from independent gamma-ray telescopes. The BL Lac object 1ES 2344+514 (z = 0.044) was detected at energies about 350 GeV by the Whipple Observatory in 1995 (Catanese et al. 1998). Most of the emission comes from a single night, December 20, in which a flux of approximately half that of the Crab Nebula was detected with a significance of 6 sigma. Other observations during that year revealed an excess of 4 sigma; subsequent observations have yielded no significant signal. 1ES 2344+514 is not detected by EGRET (D. J. Thompson & the EGRET Team 1996, private communication), so if this detection is confirmed it is another instance of a gamma-ray source being first detected by a ground-based telescope. PKS 2155-304 (z = 0.117), often considered the archetypical X-ray-selected BL Lac object, was detected at energies above 300 GeV at the 7 sigma level by combining observations from 1996 and 1997 by the Durham group (Chadwick et al. 1999). The flux was approximately 40% of the VHE flux of the Crab Nebula and corresponded to an active X-ray emission period. PKS 2155-304 is an EGRET source with an average flux at E > 100 MeV comparable to that of Mrk 421. Finally, the BL Lac object 3C 66A (z = 0.444) has been reported as a source of above 900 GeV gamma-rays based on a 5 sigma excess seen in observations in 1996 by gamma-ray telescopes at the Crimean Astrophysical Observatory (Neshpor et al. 1998). The average flux during these observations was approximately 120% of the flux of the Crab Nebula at these energies. 3C 66A is an EGRET source (Hartman et al. 1999).

Extreme variability on timescales from minutes to years is the most distinctive feature of the VHE emission from these BL Lac objects. Variability in the emission is a surprising feature in some respects because it implies a small emission region. If low-energy photons (e.g., infrared, optical, and ultraviolet) are produced in the same region, the VHE photons would pair produce with these photons and would not escape. Also, if the variability occurs near the base of the jet, there is likely to be considerable ambient radiation present which can attenuate the gamma-ray signal. This opacity problem is reduced considerably if the emission is beamed toward us (e.g., Dermer & Gehrels 1995; Buckley et al. 1996), and this has been one of the main arguments for gamma-ray beaming in these objects (see Section 4.1.3).

The first clear detection of flaring activity in the VHE emission of an AGN came in 1994 observations of Mrk 421 by the Whipple Collaboration (Kerrick et al. 1995a), where a 10-fold increase in the flux, from an average level that year of approximately 15% of the Crab flux to approximately 150% of the Crab flux, was observed. Subsequent analysis of Mrk 421 data indicated evidence for less prominent episodes of variability during 1992 and 1993 as well (Schubnell et al. 1996). This suggested that variability could be present on a fairly frequent basis in the VHE emission. In order to characterize this variability, the Whipple Collaboration began systematic monitoring of Mrk 421 in 1995 which continues to the present day. The observations of Mrk 421 in 1995, shown in Figure 8, revealed several distinct episodes of flaring activity, as in previous observations but, perhaps more importantly, indicated that the VHE emission from Mrk 421 was best characterized by a succession of day-scale or shorter flares with a baseline emission level below the sensitivity limit of the Whipple detector (Buckley et al. 1996). The timescale of the flaring is derived from the fact that, for the most part, the flux levels measured each night varied fairly randomly with no evidence of a smooth pattern. Thus, although no significant intranight variability was discerned in these observations, it seemed clear that it could occur.

Figure 8

Figure 8. Daily VHE gamma-ray count rates for Mrk 421 during 1995. MJD 49,720 corresponds to 1995 January 3. Figure from Buckley et al. (1996).

The hypothesis that the VHE emission of Mrk 421 could flare on subday timescales was borne out in spectacular fashion in 1996, with the observations of two flares by the Whipple Collaboration (Fig. 9; Gaidos et al. 1996). In the first flare, observed on May 7, the flux increased monotonically during the course of ~ 2 hours of observations, beginning at a rate twice as high as any previously observed flare and reaching a counting rate approx 10 times the rate from the Crab, at which point observations had to stop because of moonrise. This flux is the highest observed from any VHE source to date. The doubling time of the flare was ~ 1 hour. Follow-up observations on May 8 showed that the flux had dropped to a flux level of approx 30% of the Crab Nebula flux, implying a decay timescale of less than 1 day. The second flare, observed on May 15, although weaker, was remarkable for its very short duration: the entire flare lasted approximately 30 minutes with a doubling and decay time of less than 15 minutes. These two flares are the fastest timescale variability, by far, seen from any blazar at any gamma-ray energy.

Figure 9

Figure 9. Light curves of two flares observed from Mrk 421 by the Whipple Collaboration on (a) 1996 May 7 and (b) May 15. The time axes are shown in coordinated universal time (UTC) in hours. For the May 7 flare, each point is a 9 minute integration; for the May 15 flare, the integration time is 4.5 minutes. Figure from Gaidos et al. (1996).

Systematic observations of Mrk 501 sensitive to day-scale flares have been conducted since 1995 with the Whipple Observatory gamma-ray telescope (Quinn et al. 1999) and since 1997 with the telescopes of the HEGRA (Aharonian et al. 1999a), CAT (Punch 1997), and Telescope Array (Hayashida et al. 1998) collaborations. The results of these observations indicate a wide range of emission levels (Fig. 10) and some very interesting similarities and differences with the VHE emission from Mrk 421. The observations in 1995 indicate a flux which is constant, with the exception of one night, MJD 49,920, when the flux was approximately 4.6 sigma above the average (approximately 5 times the flux during the remainder of the season) (Quinn et al. 1996, 1999). Observations in 1996 by the Whipple Observatory show that the average flux of Mrk 501 had increased to approximately 20% of the Crab Nebula flux above 300 GeV, indicating a twofold increase in the average flux over the 1995 observations (Quinn et al. 1999). HEGRA observations indicated an average flux of approximately 30% of the Crab flux above 1.5 TeV, perhaps indicating a harder emission spectrum than that of the Crab Nebula. The Whipple observations show no clear flaring episodes, but the probability that the average monthly flux levels are drawn from a distribution with a constant flux level is 3.6 × 10-5, clearly indicating that the emission is varying on at least month scales (Quinn et al. 1999). There is no significant evidence for day-scale variations within each month in 1996.

Figure 10

Figure 10. VHE gamma-ray light curve for Mrk 501 as observed with the Whipple Telescope between 1995 and 1998 at energies above 350 GeV. Fluxes are expressed as fractions of the Crab Nebula flux above 350 GeV. Figure adapted from Quinn et al. (1999).

In 1997, the VHE emission from Mrk 501 changed dramatically. After being the weakest known source in the VHE sky in 1995 and 1996, it became the brightest, with an average flux greater than that of the Crab Nebula (whereas previous observations had never revealed a flux greater than 50% of the Crab flux). Also, the amount of day-scale flaring increased and, for the first time, significant hour-scale variations were seen. Two clear episodes of hour-scale variability were detected with the Whipple Observatory telescope (Fig. 11), and a search for intraday variability revealed several other nights which, considered alone, would not have been considered significant but, when combined, indicated frequent intraday variability which was just below the sensitivity of the Whipple Telescope (Quinn et al. 1999). Analysis of data from the HEGRA (Aharonian et al. 1999a) and Telescope Array (Hayashida et al. 1998) projects revealed no statistically significant intranight variations, but the two nights in the HEGRA data with the smallest statistical probability of having constant emission are the same nights seen to have significant variability in the Whipple data.

Figure 11

Figure 11. Very high energy gamma-ray light curves of Mrk 501 for the two nights in 1997 which show significant intranight variability. Figure from Quinn et al. (1999).

Perhaps the most important aspect of the observations of Mrk 501 in 1997 was that, for the first time, Cerenkov telescopes other than the Whipple Telescope consistently detected a significant excess from Mrk 501 on a nightly timescale. This permitted more complete VHE light curves to be obtained (which it is expected will eventually lead to a better understanding of the VHE emission from blazars) and also provided confirmation that different VHE telescopes could obtain consistent results from a variable source (see Fig. 12).

Figure 12

Figure 12. Very high energy gamma-ray observations of Mrk 501 with the Whipple (filled circles, E > 350 GeV), CAT (open squares, E > 250 GeV), HEGRA (filled stars, E > 1 TeV), and Telescope Array (open triangles, E > 600 GeV) telescopes between 1997 February and October. Fluxes are shown in units of the Crab Nebula flux for the energy threshold of each telescope. The numbers on the horizontal axes for each plot indicate the Modified Julian Days during that month. Data are from Quinn et al. (1999), Aharonian et al. (1999a), Djannati-Atai et al. (1999), and Hayashida et al. (1998).

In addition to the establishment of the flaring itself, the Telescope Array Collaboration performed a periodicity search with their VHE observations of Mrk 501 in 1997 (Hayashida et al. 1998). They show evidence for a quasi-periodic signal in the data which has a period of approximately 12.7 days. This is disturbingly close to half the lunar cycle. If real, this would be an extraordinary result, given the very short timescale of the quasi-periodicity.

The only published results on observations of Mrk 501 in 1998 are those of the Whipple Observatory (Quinn et al. 1999). The average emission level was approximately 30% of the Crab flux, but there was considerably more variability in the emission than in 1996 or 1995. Two distinct, very high flux flares were observed, one with the highest flux (approximately 5 times the Crab flux) ever observed from Mrk 501 with the Whipple Telescope. The monthly average flux was also variable, with 3 months showing emission levels similar to the 1995 flux, approximately 10% of the Crab Nebula.

A natural question to ask about the variability in Mrk 501 is whether the degree of variability seen changes as a function of the mean flux level, that is, whether Mrk 501 is really more variable when its average flux is higher or whether it is an artifact of the telescopes being more sensitive to variations when the average flux is higher. To test this, Quinn et al. (1999) performed simulations to see if the day-scale variability observed with the Whipple Telescope in 1997 (when the average VHE flux was 1.3 times that of the Crab) would have been detectable in 1996 and 1995 (when the average VHE flux was 20% and 10% that of the Crab, respectively) and also tested whether the month-scale variations in 1997 and 1996 would have been detectable in 1995. Their simulations indicate that the day-scale flaring in 1997 would have been detectable in the 1996 data, but not the 1995 data, and the month-scale variations in 1996 would have been detectable in 1995, while the 1997 month-scale variations would not have been detectable. Thus, it appears that the higher state emission levels have different variability characteristics than the lower emission levels.

The other unconfirmed sources are, if they are indeed sources, also variable emitters of VHE gamma-rays. 1ES 2344+514 was only detected with high statistical significance on one night, but has never been detected since, with flux limits of approximately 10% of the Crab flux (Catanese et al. 1998; Aharonian et al. 1999b). PKS 2155-304 has also been claimed to be variable (Chadwick et al. 1999), although the statistical probability that the emission is constant is not quoted. Finally, 3C 66A must be variable, or else observations with the Whipple Observatory telescope (Kerrick et al. 1995b) and the HEGRA telescope (Aharonian et al. 1999b) would have easily detected this object at the flux level quoted by the Crimean group.

The high-flux VHE emission from Mrk 421 and Mrk 501 has permitted detailed spectra to be extracted. Accurate measurements of the VHE spectrum are important for a variety of reasons. First, the shape of the high-energy spectrum is a key input parameter of AGN emission models, particularly as it relates to the MeV-GeV measurements by EGRET. Second, how the spectrum varies with flux, compared with longer wavelength observations, provides further emission model tests. Third, spectral features, such as breaks or cutoffs, 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 (see Section 4.2).

For Mrk 421, the only detailed spectra published at this time come from observations of high-state emission with the Whipple Observatory telescope (Fig. 13; Zweerink et al. 1997; Krennrich et al. 1999). Analysis of the spectra obtained from observations of flares on 1996 May 7 and 15 and observations of high-state emission taken at large zenith angles in 1995 June indicate that, within the statistical uncertainties, the spectra are all consistent with a simple power-law spectrum: dN / dE propto E-2.5 (Krennrich et al. 1999). When combined, these three data sets are consistent with a simple power-law spectrum for Mrk 421 of the form (Krennrich et al. 1999)


where E is in units of TeV.

Figure 13

Figure 13. VHE gamma-ray spectra of Mrk 421 (open stars) and Mrk 501 (filled circles) as measured with the Whipple Observatory telescope (from Krennrich et al. 1999). The solid line through the Mrk 421 points indicates the best-fit spectrum to these data and the dashed line through the Mrk 501 points is the best-fit spectrum for those data.

Observations of Mrk 421 in 1997 and 1998 with the HEGRA system of Cerenkov telescopes reveal a significantly different spectrum (Aharonian et al. 1999c),


than observed with the Whipple Telescope. The emission level for the HEGRA observations was approximately 0.5 times the Crab flux, much lower than the fluxes (1-10 times the Crab flux) used in the Whipple observations. This may indicate that the spectrum in Mrk 421 becomes softer with decreasing flux. However, HEGRA observations show no evidence of variability between observations at fluxes above the Crab flux and those between one-sixth and one-half the Crab flux, and the Whipple results show no variations in spectral index despite using observations spanning a 10-fold range of fluxes. Further studies may help resolve these differences.

As with the studies of the variability of its VHE flux, the high-state emission detected from Mrk 501 in 1997 allowed detailed spectra to be derived by several experiments, permitting studies of the time dependence of the spectra and providing all-important cross-checks of the methods used to derive energy spectra. Because of the rapid variability of the emission, again the normalization of the spectra are not of fundamental importance, except perhaps in the context of multiwavelength studies which are discussed in Section 4.1.2 below.

The most detailed energy spectra published at this time come from Whipple observations between 250 GeV and 12 TeV (Samuelson et al. 1998; Krennrich et al. 1999) and HEGRA data spanning 500 GeV-20 TeV (Aharonian et al. 1999b). The Telescope Array Collaboration has also derived a spectrum over a slightly narrower energy range (600 GeV-6.5 TeV) (Hayashida et al. 1998). A search for variability in the spectrum revealed no significant changes in spectrum with flux or time (Samuelson 1999; Aharonian et al. 1999a), allowing large data sets to be combined to derive very detailed energy spectra spanning large ranges in energy. The spectra derived by Whipple and HEGRA deviate significantly from a simple power law. For Whipple, the chi2 probability that a power law is consistent with the measured spectrum is 2.5 × 10-7. This is the first significant deviation from a power law seen in any VHE gamma-ray source and any blazar at energies above 10 MeV. The Whipple spectrum is


and the HEGRA spectrum is


where E is in units of TeV. The form of the curvature term in the spectra has no physical significance as the energy resolution of the experiments is not sufficient to resolve particular spectral models. The Whipple spectral form is simply a polynomial expansion in log E versus log(dN / dE) space. The HEGRA form was chosen presumably because attenuation of the VHE gamma-rays by pair production with background IR photons could produce an exponential cutoff. In fact, the Whipple and HEGRA data are completely consistent with each other, as shown in Figure 14. The Telescope Array Collaboration derived a spectrum which is well fitted by a simple power law (dN / dE propto E-2.5±0.1). The data from this spectrum are also consistent with the Whipple and HEGRA spectra.

Figure 14

Figure 14. The energy spectrum of Mrk 501 as measured by the HEGRA array (filled circles) and the Whipple Telescope (open circles). The dashed line indicates the power law plus exponential cutoff spectrum fit to the HEGRA data. Figure from Konopelko (1999).

4.1.2. Multiwavelength Observations

Some of the most exciting results on VHE gamma-ray sources have come through observations where several telescopes operating at different wavelengths simultaneously monitor the activity in a blazar. These multiwavelength campaigns have involved the larger astronomical community in the study of VHE sources and served the subsidiary purpose of confirming the source identifications of the VHE gamma-ray-emitting blazars.

The first evidence of correlated variability between VHE gamma-rays and lower energy emission came from a multiwavelength campaign on Mrk 421 in 1994 April/May (Macomb et al. 1995, 1996). Observations were conducted with the Whipple Telescope, EGRET, ASCA in X-rays, the International Ultraviolet Explorer (IUE), the United Kingdom Infrared Telescope (UKIRT), the James Clerk Maxwell Telescope (JCMT) in the millimeter wave band, and the University of Michigan 26 m radio telescope (UMRAO) in the 4.8-14.5 GHz frequency range. The VHE gamma-ray, ASCA, and EGRET observations are shown in Figure 15. The VHE observations reveal the rising edge of a flare that developed over approximately 4 days with the peak flux detected on May 15 (UT) being approximately 9 times the mean flux measured during that observing season and approximately 1.4 times the flux of the Crab Nebula. Observations had to be halted after May 15 because the phase of the Moon precluded further observations. Observations taken with ASCA on May 16/17 (Takahashi et al. 1994, 1996) indicated a flux approximately 20 times the quiescent X-ray flux of Mrk 421, and the time coincidence between the two observations of unprecedented high states is the basis for the claim of a correlation in this campaign. Interestingly, EGRET observations taken between 1997 May 10 and 17 did not detect the strong day-scale variability seen in VHE gamma-rays. The average flux during this period was approximately twice the average flux measured in 1994 April, so there is some evidence of a higher emission state, but it is not significant enough to claim a correlation. The observations at UV, IR, millimeter, and radio wavelengths showed no evidence of variability during this period. Because of the offset in time of the observations between the VHE gamma-rays and the X-rays, detailed comparisons of the variability in those wave bands are not possible.

Figure 15

Figure 15. Light curves for observations of Mrk 421 in 1994 May by ASCA (top), Whipple (middle), and EGRET (bottom). Figure from Takahashi, Madejski, & Kubo (1999).

Spurred by this result, another multiwavelength campaign was organized in 1995 to better measure the multiwavelength properties of Mrk 421. This campaign revealed, for the first time, correlations between VHE gamma-rays and X-rays (Buckley et al. 1996). Observations were conducted between April 20 and May 5 with the Whipple Telescope, EGRET, ASCA, the Extreme Ultraviolet Explorer (EUVE), an optical telescope, an optical polarimeter, and UMRAO. Observations with EGRET did not result in a detection of Mrk 421. The 2 sigma flux upper limit for E > 100 MeV is 1.2 × 10-7 cm-2 s-1, somewhat below the level detected in 1994. The light curves for some of these observations are shown in Figure 16. The optical data have the contribution from the host galaxy of Mrk 421 subtracted off. As in the 1994 multiwavelength campaign, Mrk 421 underwent a large-amplitude flare in VHE gamma-rays during the observations. The flare is also clearly seen in the ASCA and EUVE observations. There is some evidence for correlated variability in the optical flux and polarization, but the statistics are not good enough for such an association to be claimed with confidence. The X-rays and VHE gamma-rays appear to vary together, limited to the 1 day resolution of the VHE observations, and the amplitude of the flaring is similar, ~ 400% difference between the peak flux and that at the end of the observations. The EUVE and optical data (assuming it also is correlated) are consistent with the flare being delayed by approximately 1 day relative to the X-rays and VHE gamma-rays. The amplitude of the flare also decreases with decreasing energy. The XUV flux varies by ~ 200% during the observations and the optical flux varies by about 20%. The B-band percent polarization varies by nearly a factor of 2 in the observations.

Figure 16

Figure 16. (a) Gamma-ray, (b) X-ray, (c) extreme-UV, (d) optical, and (e) optical polarization measurements of Mrk 421 taken 1995 April-May. April 26 corresponds to MJD 49,833. Figure from Buckley et al. (1996).

The observations of Mrk 421 in 1994 and 1995 were clearly undersampled, limiting the conclusions that could be drawn concerning correlations between wavelengths and emission models. Two multiwavelength campaigns organized in 1998 attempted to improve these measurements through more dense observations in X-rays and VHE gamma-rays. Improvement in the VHE measurements came about through longer VHE exposures with individual telescopes, to search for hour-scale variations, and coordination of VHE observations between CAT, HEGRA, and Whipple. Thus, light curves of 12-16 hours could in principle be achieved, allowing much more detailed measurements of the VHE emission. These observations yielded immediate improvements in the measurements of flaring activity and correlations between VHE gamma-rays and X-rays.

The first campaign, conducted in late 1998 April, was centered at X-ray wavelengths on observations with the BeppoSAX satellite and established the first hour-scale correlations between X-rays and gamma-rays in a blazar (Maraschi et al. 1999). The light curve for the observations by BeppoSAX in three X-ray bands and Whipple above 2 TeV is shown in Figure 17. The VHE threshold here is higher than typical of Whipple data because observations were taken at a wide range of elevations, causing the energy threshold and sensitivity to vary with time. These effects were corrected in the light curve through the use of simulations and analysis cuts to normalize the collection area and energy threshold. Thus, the VHE light-curve rate variations are intrinsic. As Figure 17 shows, a flare is clearly detected in X-rays and TeV gamma-rays on the first day of observations. The peaks in the light curves occur at the same time, within 1 hour, but the falloff in the X-ray flux is considerably slower than the TeV gamma-rays. Also, the TeV gamma-rays have a larger variability amplitude (~ fourfold ratio between average and peak) than the X-rays (~ twofold ratio). Both the faster VHE flux decrease and the larger amplitude variability have not been seen previously in Mrk 421. These observations provide the first clear indication that X-rays and VHE gamma-rays may not be completely correlated on all timescales.

Figure 17

Figure 17. Light curves for observations of Mrk 421 in 1998 April by Whipple and BeppoSAX. Whipple observations are for E > 2 TeV and are binned in 28 minute observing segments. All count rates are normalized to their respective averages (listed at the top of each panel) for the observations shown. Figure from Maraschi et al. (1999).

The second multiwavelength campaign started in late 1998 April, immediately after the observations discussed above, and was centered around a 7 day continuous observation of Mrk 421 with ASCA (Takahashi, Madejski, & Kubo 1999). The light curves for the X-ray observations and VHE observations by Whipple, CAT, and HEGRA are shown in Figure 18. Had these observations been conducted as in 1994 and 1995, with short X-ray exposures and just Whipple observations, the results would have shown nothing new: similar variability scale between the X-rays and VHE gamma-rays and some subday-scale X-ray variability with amplitude too low to be resolved with the Whipple observations. Instead, the X-ray observations reveal the complete cycle of about 10 flares, the first time this has been done for Mrk 421. Also, these observations seem to confirm the supposition of Buckley et al. (1996) that the VHE emission from Mrk 421 is primarily the result of flares, with little steady emission evident. Finally, the combination of VHE data from these telescopes confirms the subday-scale correlations seen in the Whipple/BeppoSAX observations. Detailed comparisons of the VHE gamma-ray and X-ray data will require more sophisticated normalization of the VHE data (e.g., to common threshold energies) and investigation of systematics in the measures of variability, but the data hold the promise of significantly advancing our study of VHE-emitting gamma-ray blazars. They also clearly demonstrate the benefits of operating multiple VHE gamma-ray installations in understanding the nature of variable sources.

Figure 18

Figure 18. Light curves for observations of Mrk 421 in 1998 April-May. (a) VHE flux observed with HEGRA, Whipple, and CAT, (b) X-ray flux observed with ASCA, and (c) X-ray hardness ratios observed with ASCA are shown. Figure from Takahashi et al. (1999).

The first multiwavelength observations of Mrk 501 which included VHE observations were conducted in 1996 (Kataoka et al. 1999). The observations were conducted with the Whipple Telescope, EGRET, ASCA, and an optical telescope. The light curve for these observations is shown in Figure 19. The observations were too undersampled, or insensitive in the case of EGRET, to clearly establish any correlations, but these observations have two very important results. First, follow-up observations in 1996 May established the first detection of Mrk 501 by EGRET, with a marginal significance of 4.0 sigma above 100 MeV but a significance of 5.2 sigma above 500 MeV, indicating a hard photon spectrum (Kataoka et al. 1999). The claim of a 3.5 sigma detection by EGRET during a small part of the multiwavelength campaign (the region between the dashed lines in Fig. 19) seems speculative, given the lack of any increased activity in other wave bands during that period. Second, the observations established a baseline spectral energy distribution for Mrk 501 during a relatively low emission state which could be compared to observations of the high-state emission in 1997 (see below for discussion of spectral energy distributions).

Figure 19

Figure 19. Light curve for observations of Mrk 501 in 1996 March. Observations in (a) 0.7-10 keV X-rays (ASCA), (b) 100 MeV-10 GeV gamma-rays (EGRET), (c) above 350 GeV gamma-rays (Whipple), and (d) R-band, 650 nm, optical are shown. Figure from Kataoka et al. (1999).

Multiwavelength observations of Mrk 501 during its high emission state in 1997 revealed, for the first time, clear correlations between its VHE gamma-ray and X-ray emission (Catanese et al. 1997). Observations were conducted with Whipple (nightly, April 7-19), EGRET and the Oriented Scintillation Spectrometer Experiment (OSSE) on CGRO (April 9-15), BeppoSAX (April 7, 11, 16), RXTE (twice nightly, April 3-16), and the Whipple Observatory 1.2 m optical telescope (nightly, April 7-15). The optical and X-ray observations were serendipitously scheduled at this time, and the CGRO observations were a public target of opportunity observation initiated in response to the high VHE emission state.

Figure 20 shows daily flux levels for the contemporaneous observations of Mrk 501. The average flux level in the U band in March is also included in the figure (Fig. 20e, dashed line) to indicate the significant (gtapprox10%) increase in flux between March and April. An 11 day rise and fall in flux is evident in the VHE and X-ray wave bands, with peaks on April 13 and 16. The 50-150 keV flux detected by OSSE also increases between April 9 and 15, with a peak on April 13. The optical data may show a correlated rise, but the variation is small (at most 6%). Subtraction of the galaxy light contribution will increase the amplitude of this variation, but it should still remain lower than in X-rays, given that the R-band contribution of the galaxy light is ~ 75% (Wurtz, Stocke, & Yee 1996) and the U-band contribution should be much less. EGRET observations indicated an excess of 1.5 sigma, not a significant detection. The ratio of the fluxes between April 13 and April 9 are 4.2, 2.6, 2.3, and 2.1 for the VHE gamma-ray, OSSE, RXTE, 15-25 keV, and RXTE 2-10 keV emission, respectively.

Figure 20

Figure 20. (a) VHE gamma-ray, (b) OSSE 50-150 keV, (c) RXTE 2-10 and 15-25 keV, (d) RXTE All-Sky Monitor 2-10 keV, and (e) U-band optical light curves of Mrk 501 for the period 1997 April 2 (MJD 50,540) to April 20 (MJD 50,558). The dashed line in (e) indicates the average U-band flux in 1997 March. Whipple, OSSE, ASM, and optical data are from Catanese et al. (1997) and RXTE data are from Catanese (1999).

The results of this campaign show that for Mrk 501, like Mrk 421, the VHE gamma-rays and the soft X-rays vary together and the variability in the synchrotron emission increases with increasing energy. However, OSSE has never detected Mrk 421 despite several observations (McNaron-Brown et al. 1995), while the Mrk 501 detection had the highest 50-150 keV flux ever detected by OSSE from a blazar. A likely explanation of the OSSE detection is that the synchrotron emission in Mrk 501 extends to 100 keV, compared with the ~ 1 keV cutoff seen in Mrk 421. This explanation was first confirmed by the observations with BeppoSAX (Pian et al. 1998). In addition, the day-scale variations for Mrk 501 are larger in gamma-rays than in X-rays, unlike Mrk 421. So, despite the similarity of Mrk 421 and Mrk 501 in some respects, these multiwavelength campaigns are beginning to reveal differences in the two objects.

In these short multiwavelength campaigns, there appears to be a correlation between the X-rays and VHE gamma-rays. A natural question to ask is whether this is always true or only during certain situations. An attempt to answer this question has been made by the HEGRA collaboration by comparing their observations of Mrk 501 above 500 GeV to those by the RXTE All-Sky Monitor (ASM), measuring 2-12 keV photons (Aharonian et al. 1999a). A cross-correlation analysis of the daily average flux measured by the ASM with the daily average flux measured by HEGRA reveals a peak in the correlation function at Deltat = 0 ± 1 day. However, the peak in the cross-correlation function is only ~ 0.4, and the significance of the peak is only 2-3 sigma. Whether this indicates that the X-ray/TeV correlation is not present is unclear because the ASM data have large statistical and significant systematic uncertainties for day-scale measurements of this relatively dim X-ray source (i.e., compared with the X-ray binaries the ASM was designed to monitor). Also, because HEGRA sits on the falling edge of the high-energy spectrum and the ASM sits (for Mrk 501 in 1997) on the rising edge of the synchrotron spectrum, it is possible that the emission detected by these two instruments will not be completely correlated, particularly for day-scale variations. Comparison of longer term variability between the measurements might help resolve such issues.

Figure 21 shows the spectral energy distributions (SEDs), expressed as power per logarithmic bandwidth, for Mrk 421 and Mrk 501 derived from contemporaneous multiwavelength observations and an average of noncontemporaneous archival measurements. Both have a peak in the synchrotron emission at X-ray frequencies, as is typical of X-ray-selected BL Lac objects, and a high-energy peak whose exact location is unknown but must lie in the 10-250 GeV range. Both the synchrotron and high-energy peak are similar in power output, unlike the EGRET-detected flat-spectrum radio quasars, which can have high-energy peaks well above the synchrotron peaks (e.g., von Montigny et al. 1995). Also, during flaring episodes, the X-ray spectrum in both objects tends to harden significantly (Takahashi et al. 1996, 1999; Pian et al. 1998) while the VHE spectrum is not observed to change.

Figure 21

Figure 21. Left: The spectral energy distribution of Mrk 421 from contemporaneous and archival observations. Dates of the observations are indicated in the figure. Figure from Buckley et al. (1997). Right: The spectral energy distribution of Mrk 501 from contemporaneous and archival observations. Dates of the observations are indicated in the figure. Data in the figure come from Kataoka et al. (1999), Catanese et al. (1997), and Catanese (1999) and references therein. The curves in the figure are meant to guide the eye to the contemporaneously measured points and do not indicate model fits to the data, nor are they an attempt to elucidate the spectral energy distribution of Mrk 501 during these observations. In both figures, the archival measurements are approximate averages of the data in the literature.

The SEDs of the two sources do, however, exhibit important differences. Most prominent among these is that the combination of contemporaneous RXTE and OSSE observations of Mrk 501 in 1997 clearly confirm the initial measurements of Pian et al. (1998) that the synchrotron spectrum extended well beyond the ~ 1 keV typical of X-ray-selected BL Lac objects. They also establish that the peak power output of the synchrotron emission occurs at ~ 100 keV. This is in contrast to the 1996 observations of Mrk 501 reported by Kataoka et al. (1999), where the synchrotron power peak is at ~ 2 keV. For Mrk 421, the X-ray spectral peak does shift to higher energies during flaring activity, but the changes are much smaller than in Mrk 501, and the peak was never observed to extend beyond ~ 1 keV. This peak is followed by a sharp cutoff which produces a deficit in the OSSE range, preventing the detection of Mrk 421 by this instrument.

Whether the shift in the location of the synchrotron peak for Mrk 501 is also accompanied by a shift in the onset of the gamma-ray emission to higher energies is not clear. Any increase in the MeV-eV flux in 1997 was not as great as at TeV energies, or EGRET would have easily detected Mrk 501. But, because the sensitivity of EGRET in 1997 was substantially poorer than in 1996, a shift in the onset of the spectrum or a flux variation that increases with increasing energy could explain the nondetection by EGRET. The fact that the VHE spectrum of Mrk 501 is harder than that of Mrk 421 below ~ 1 TeV (see Fig. 13) may also indicate that the high-energy peak of Mrk 501 is shifted to slightly higher energies. However, it could also simply indicate a slower falloff in the progenitor particle spectrum above the peak power output.

A second difference in the spectral energy distributions for these two objects is that the power output for Mrk 501 in the VHE range can be considerably less than in X-rays when it is in a low emission state. In contrast, Mrk 421 seems to maintain a similar output at X-ray and gamma-ray energies. These differences are illustrated in Table 5, which gives the ratio of contemporaneously measured fluxes for X-rays and gamma-rays for these two objects. Whether the difference in power output for Mrk 501 reflects a change of the energy at which the peak in the high-energy spectrum occurs or something related to the flaring process is not clear, because of the poor spectral measurements in the low emission states and the lack of coverage of the peak region of the spectrum. However, the lack of spectral variability in gamma-rays argues against a significant short-term shift of the gamma-ray spectral peak.

Table 5. Spectral energy ratios for VHE sources

Source Date R2 keV a R100 keV b R100 MeV c Reference

Mrk 421... 1995 2.2 ± 0.5 ... <0.38 1
  1996 0.81 ± 0.06 ... ... 2
Mrk 501... 1996 4.4 ± 2.1 ... 3.1 ± 3.4 3
  1997 1.2 ± 0.3 1.9 ± 0.5 <0.36 4

a nu Fnu (2 keV) / nu Fnu (350 GeV).
b nu Fnu (100 keV) / nu Fnu (350 GeV).
c nu Fnu (100 MeV) / nu Fnu (350 GeV).
References. (1) Buckley et al. 1996; (2) Buckley et al. 1997; (3) Kataoka et al. 1999; (4) Catanese et al. 1997; Catanese 1999.

4.1.3. Implications of the VHE Observations

The general properties of the detected extragalactic sources of VHE gamma-rays are listed in Table 6. The three objects detected by the Whipple Collaboration exhibit some interesting commonalities. They are the three closest known BL Lac objects with declination below 0°, so their gamma-ray fluxes are the least attenuated from interaction with background IR radiation. The above 100 MeV fluxes are near (Mrk 421, Mrk 501) or below (1ES 2344+514) the EGRET sensitivity limit, meaning that the gamma-ray power output does not peak in that energy range as it does for many of the EGRET-detected AGNs (von Montigny et al. 1995). Thus, VHE observations already augment the catalog of gamma-ray sources compiled by space-borne telescopes. Finally, all three of the Whipple-detected BL Lac objects are X-ray-selected BL Lac objects (XBLs). The extension of the synchrotron spectra to X-ray energies in XBLs implies that they produce high-energy electrons, making them good candidates for VHE emission if the VHE gamma-rays are produced via inverse Compton (IC) scattering of these same electrons. EGRET's tendency to detect more radio-selected BL Lac objects (RBLs) than XBLs (Lin et al. 1997) also supports this tenet because RBLs would be expected to have spectra which peak in the MeV-GeV range (Sikora et al. 1994; Marscher & Travis 1996). BL Lac objects in general have been suggested as better candidates for VHE emission than other blazars because the absence of optical emission lines in BL Lac objects may indicate less VHE-absorbing radiation near the emission region (Dermer & Schlickeiser 1994).

Table 6. Properties of the VHE BL Lac objects

Object z EGRET Flux a
(E > 100 MeV)
(10-7 cm-2 s-1)
Average Flux
(E > 300 GeV)
(10-12 cm-2 s-1)
Mv a FX a
(2 keV)
FR a
(5 GHz)

Mrk 421... 0.031 1.4 ± 0.2 40 14.4 3.9  720
Mrk 501... 0.034 3.2 ± 1.3 geq 8.1 14.4 3.7 1370
1ES 2344+514 c... 0.044 < 0.7 ltapprox 8.2 15.5 1.1  220
PKS 2155-304 c... 0.116 3.2 ± 0.8 42 13.5 5.7  310
3C 66A c... 0.444 2.0 ± 0.3 30 b 15.5 0.6  806

a Radio, optical, and X-ray data from Perlman et al. 1996. EGRET data from D. J. Thompson 1996, private communication, Mukherjee et al. 1997, and Kataoka et al. 1999.
b >1 TeV flux value.
c Unconfirmed as a VHE source.

The other two objects detected at VHE energies, PKS 2155-304 and 3C 66A, are similar in some respects to the Whipple sources. PKS 2155-304 is an XBL, so it fits the IC paradigm for VHE sources. However, 3C 66A is classified as an RBL, suggesting that protons produce the gamma-rays because in IC models, RBLs would not have high enough energy electrons to produce TeV emission. In addition, both PKS 2155-304 and 3C 66A are at much higher redshifts than the Whipple sources, implying quite low IR backgrounds. Thus, confirmation of these detections, just as for 1ES 2344+514, is essential.

VHE observations have already significantly affected our understanding of BL Lac objects. For example, the rapid variability either indicates very low accretion rates and photon densities near the nucleus (Celotti, Fabian, & Rees 1998) or, conversely, requires the gamma-ray emission region to be located relatively far from the nucleus to escape the photon fields (Protheroe & Biermann 1997). Also, the observations have helped resolve the nature of the differences between RBLs and XBLs. Based on their smaller numbers and higher luminosities, Maraschi et al. (1986) proposed that RBLs were the same as XBLs but with jets aligned more closely with our line of sight. However, the rapid variability and TeV extent of the XBL emission point to the differences between the two subclasses being more fundamental, as originally proposed by Padovani & Giommi (1994): the XBLs have higher maximum electron energies and lower intrinsic luminosities.

Simultaneous measurements of the synchrotron and VHE gamma-ray spectra also constrain the magnetic field strength (B) and Doppler factor (delta) of the jet. If the correlation between the VHE gamma-rays and optical/UV photons observed in 1995 from Mrk 421 indicates both sets of photons are produced in the same region of the jet, delta gtapprox 5 is required for the VHE photons to escape significant pair-production losses (Buckley et al. 1996). If the SSC mechanism produces the VHE gamma-rays, delta = 15-40 and B = 0.03-0.9 G for Mrk 421 (Buckley et al. 1997; Tavecchio, Maraschi, & Ghisellini 1998; Catanese 1999) and delta approx 1.5-20 and B = 0.08-0.2 G for Mrk 501 (Samuelson et al. 1998; Tavecchio et al. 1998; Hillas 1999). To match the variability timescales of the correlated emission, proton models which utilize synchrotron cooling as the primary means for proton energy losses require magnetic fields of B = 30-90 G for delta approx 10 (Mannheim 1993, 1998; Buckley 1998). The Mrk 421 values of delta and B are extreme for blazars, but they are still within allowable ranges and are consistent with the extreme variability of Mrk 421.

In addition, the VHE observations have constrained the types of models that are likely to produce the gamma-ray emission. For instance, the correlation of the X-ray and the VHE flares is consistent with IC models where the same population of electrons radiate the X-rays and gamma-rays. The absence of flaring at EGRET energies may also follow in this context (Macomb et al. 1995) because the lower energy electrons which produce the gamma-rays in the EGRET range radiate away their energy more slowly than the higher energy electrons which produce the VHE emission. The MeV-GeV emission could then be the superposition of many flare events and would therefore show little or no short-term variation.

In the mechanism of Sikora et al. (1994), which produces gamma-rays through the Comptonization of external photons, the external photons must have energies lower than 0.1 eV (in the IR band) to avoid significant attenuation of the VHE gamma-rays by pair production. Sikora et al. (1994) point out that there is little direct observational evidence of such an IR component in BL Lac objects, but the existence of such a field has been predicted as a product of accretion in AGNs (Rees et al. 1982).

Models which produce the gamma-ray emission from proton progenitors through e+e- cascades originating close to the base of the AGN jet have great difficulty explaining the TeV emission observed in Mrk 421 because the high densities of unbeamed photons near the nucleus, such as the accretion disk or the broad-line region, required to initiate the cascades cause high pair opacities to TeV gamma-rays (Coppi, Kartje, & Königl 1993). Such models also predict that the radius at which the optical depth for gamma-gamma pair production drops below unity increases with increasing gamma-ray energy (Blandford & Levinson 1995), and therefore the VHE gamma-rays should vary either later or more slowly than the MeV-GeV gamma-rays. This is in contradiction to the observations of Mrk 421 in both 1994 (Macomb et al. 1995) and 1995 (Buckley et al. 1996).

Next Contents Previous