Blazars  represent a quite small but remarkably interesting fraction of the entire population of active galactic nuclei (AGN). Their defining phenomenology includes the presence of a compact unresolved radio core, with flat or even inverted spectrum, extreme (both in timescale and in amplitude) variability at all frequencies (but generally being more extreme at the highest frequencies), high degree of optical and radio polarization. The most distinctive feature, however, is the intense emission in the γ-ray band, often dominating the bolometric radiative power output.
Indeed, blazars are the most numerous extragalactic γ-ray sources, both at GeV and TeV energies. The 4-years catalogue of Fermi-LAT (3FGL, ) reports more than 1600 blazars, to be compared with about 30 non-blazar extragalactic sources. Similar is the situation at higher energies. Cherenkov telescopes detected 62 blazars and only 6 non blazar sources (4 radiogalaxies and 2 starburst galaxies). Blazars are further divided in two subgroups, namely BL Lacertae objects, characterized by extremely weak or even absent emission lines in their optical spectra, and Flat Spectrum Radio Quasars (FSRQ), showing broad emission lines typical of quasars. FSRQs are generally more powerful than BL Lacs and their radiative output tend to be more dominated by the high-energy emission. BL Lacs, on the other hand, are characterized, on average, by larger energies of the emitted γ-ray photons and, in fact, are the large majority of the blazars detected at VHE.
The peculiarities of blazars are explained assuming that these sources are AGN hosting a jet of plasma expelled at relativistic speeds (bulk Lorentz factors Γ ≈ 10−20), whose axis points almost toward the Earth . In this geometry, the luminosity of the non-thermal continuum produced within the jet appears amplified by 3-4 orders of magnitude because of relativistic beaming effects and it easily outshines any other isotropic emission component associated to the active nucleus (accretion disk, emission lines, dust) or to the host galaxy. In the so-called unified model for radio-loud AGN, BL Lacs and FSRQ are the aligned counterpart of FRI (low power) and FRII (high power) radiogalaxies, respectively .
The remarkably smooth SED of blazars, extending over the entire electromagnetic spectrum, from the radio band to γ-ray energies, is characterized by a typical “double humped” shape. The low energy component – peaking, depending on the source, between the IR and the UV-soft X-ray band – is understood as beamed synchrotron radiation of relativistic electrons (or, more generally, e± pairs), while for the origin of the second component – reaching its maximum in the γ-ray band – there is no complete consensus. In leptonic models (see e.g., ), the emission is explained as inverse Compton (IC) radiation from the same leptons producing the low-energy component. In hadronic scenarios , instead, γ rays are thought to originate from high-energy hadrons (protons) loosing energy through synchrotron emission (if the magnetic field is large enough, ) or photo-meson reactions . In the latter case neutrino emission from the decaying charged pions is also foreseen.
Despite huge efforts, several crucial aspects of the blazar phenomenology remains unclear and poorly understood. Even the physical process(es) responsible for the acceleration of the emitting relativistic particles is (are) not very clear. While until few years ago it was widely accepted that the main acceleration mechanism is the Fermi I-like process acting at shock fronts in the flow (diffusive shock acceleration), this view is now strongly debated. Another point subject to lively discussions is the location of the regions from which a large fraction of the observed radiation originates. This problem is particularly acute for FSRQ, in which the radiative environment surrounding the jet (influencing the IC process and causing the absorption of the γ-ray photons) varies with the distance from the central supermassive black hole. All these discussions have been recently triggered or revitalized by important observational results obtained in the past years by space (AGILE, Fermi) and ground-based (Cherenkov arrays) γ-ray detectors, in particular the evidence for ultra-fast (≈ minutes) variability events and the detection of FSRQ at high (E > 10 GeV) and very-high (E > 100 GeV) energies. A quite interesting discussion concerns also the nature of the so-called extreme BL Lacs, showing extremely hard TeV continua and limited (if any) variability at high energy (e.g.  and references therein). It is also important to remind that, besides the astrophysical issues, the intense high-energy photon beams of blazars are ideal probes for the cosmological fields permeating the Universe (the extragalactic background light and the extragalactic magnetic field) or even to search for new particles (axion-like particles) or look for violation of the Lorentz invariance at high energies. For a discussion of these topics see De Angelis (these proceedings).
In the following, after a sketch of the general framework, I will review in particular the observational status concerning the ultra-rapid variability and the VHE emission of FSRQ and the impact on our understanding of the functioning of blazars.
1.1. The General Framework and its Epicycles
Blazar jets are hosted by an active nucleus comprising a central supermassive black hole (MBH = 108 − 109 M⊙) accreting matter from the surroundings. The phenomenological division between FSRQ and BL Lac objects can be interpreted as reflecting a more fundamental difference in the nature of the accretion flow in the two kind of sources, ultimately regulated by the accretion rate of the infalling material, e.g. [9, 10]. In FSRQ, which show bright thermal features (optical lines) and, in some cases, a bump at optical-UV frequencies (thought to mark the direct emission from the hot accreting gas), the accretion likely occurs through a radiatively efficient (optically thick) accretion disk. The luminous UV continuum emitted by the disk is responsible for the photoionization of the gas confined in “clouds” rapidly orbiting the black hole and occupying the so-called broad line region (BLR). Various methods (in particular the reverberation mapping technique) allows us to locate the clouds at ≈0.1 parsec (the “radius” of the BLR), which displays a clear dependence on the luminosity of the disk [11}. Farther out (1-10 pc), dust grains – likely organized in the geometrical shape of a torus – intercepts a fraction ξ ≈ 0.5 of the disk continuum, reprocessing it as thermal IR emission (with temperature close to that corresponding to the sublimation of dust, T ≈ 103 K). On the other hand, the lack of strong thermal components in BL Lac optical spectra is generally interpreted as an evidence that the accretion flow present in these sources is radiatively inefficient, as expected for accretion rates much smaller than that of quasars, when the accretion flow assumes the structure of an ADAF/ADIOS .
This scheme could allows one to explain the difference between the GeV γ-ray spectra of BL Lacs (generally displaying hard spectra) and those of FSRQ (characterized by soft – photon index larger than 2 – spectra), as the effect of the different radiative losses characterizing the high-energy electrons in the two kind of sources . More generally, the interplay between radiative losses of the emitting electrons in the jet and the accretion rate onto the black hole could be at the base of the so-called “blazar sequence” , i.e. the trend between the observed luminosity (progressively increasing from BL Lacs to FSRQ) and the synchrotron and high-energy SED peak frequencies (decreasing from BL Lacs to FSRQ) displayed by the blazar population  – but see  for an alternative view.
While there is wide consensus about the general picture, there are several fundamental questions still awaiting an answer. The list of the most pressing problems includes the nature of the mechanism powering the jet, its structure, its composition (ep or e±?), the role of the magnetic field. Although not conclusive, the modelling of the emission from blazars – especially of the high-energy emission – is used as an effective tool to start to address several of these questions.
The most adopted emission models (“one-zone”) assume that a single region of the jet is responsible for the bulk of the observed emission. 1 This region could be identified with the shocked portion of the jet resulting from the collision of parts of the jet moving at different speeds (internal shocks), . One-zone models have the advantage to require a limited number of free parameters. The simplest version of the one-zone leptonic scenario is at the base of the the one-zone synchrotron-self Compton model, which assumes that the IC component is produced through the scattering of the synchrotron radiation produced by the same relativistic electrons. This framework is thought to be especially suitable to model BL Lacs, in which the possible external sources of soft target photons are negligible. The limited number of free physical parameters required by the one-zone SSC model allows us to uniquely determine them in case a well sampled SED is available [17, 18].
The application of the one-zone SSC model to the SED of the γ-ray emitting BL Lac (e.g., ) requires magnetic fields in the range 0.01-1 G, typical radius in the interval 1015 − 1016 cm, Doppler factors 2 in the interval δ = 10−30 and electron energies (at the SED peak) of the order of 0.1-1 TeV. If the jet is assumed to be conical, with typical aperture angle θj ≈ 5° the source radius can be translated into a distance from the central engine, rem ≈ 1016−1017 cm, corresponding to 102−103 gravitational radii. The results also suggest that most of BL Lac objects are quite inefficient in emitting the radiation, since the derived cooling time of the electrons is generally much longer then the dynamical timescale and thus most of the energy stored in the relativistic particles is lost [20, 21]. Somewhat paradoxically for source emitting conspicuous VHE radiation, a relatively low efficiency is also found to characterize the acceleration process itself, see . Another important point is that the emission regions appear to be matter-dominated, the magnetic field providing a negligible contribution to the power carried by the jet (more on this later). For FSRQ the situation is more complex, due to the presence of several sources of external photons potentially involved in the IC emission . The “canonical” choice is to locate the emission region within the BLR, thus exploiting for the IC emission the dense radiation field produced by the photoionized gas. Representative values of the magnetic field are larger than those of BL Lacs, in the range 1-10 G, but Doppler factors and radius are similar [24, 25]. For FSRQ we have some control on the accretion power through the observed thermal components and it is thus possible to compare the power carried by the jet with that advected by the accreting material. The comparison show that, on average, jets carry a power larger than that associated to the infalling accretion flow, suggesting that the source of the jet power is the BH spin [24, 26], as supported by recent GR-MHD simulations .
Although very attractive, the one-zone model is clearly quite a simplification of the actual, likely complex, structure of the emitting region(s). A minimal approach is to add one or more supplementary emission regions, such as in two-component models, e.g. [28, 29]. A more refined modelling within the shock-in-jet scenario includes a shock front – at which particles are accelerated – moving within a “background” plasma, in which the particles injected by the shock radiatively cool and emit [30, 31, 32]. Among all possible extensions of the one-zone framework (see below for other alternatives stemming from the interpretations of the ultra-rapid variability) I would like to mention in particular the structured jet model , which envisages a flow with two components, a faster core (the spine) surrounded by a slower sheath or layer. This kind of structure has been advanced as a possible solution for several issues related to TeV emitting BL Lacs and to unify the BL Lacs and radiogalaxy populations [34, 35]. Direct radio VLBI imaging of jets both in low-power radiogalaxies and BL Lac objects (e.g. [36, 37]), often showing a “limb brightening” transverse structure, provides a convincing observational support to this idea, further supported by MHD simulations [38, 39]. For this system we expect an increased IC γ-ray luminosity, based on the fact that for particles carried by the faster (slower) region, the radiation field produced in the layer (spine) is amplified by the relative motion between the two structures [33, 40]. The spine-layer structure could be involved in the possible production of high-energy neutrino by BL Lac jets . The structured jet scenario can also accommodate the high-energy emission (extending at VHE) observed in the misaligned jets of radiogalaxies, such as M87  and NGC 1275 , although the large opacity to γ rays caused by the intense layer radiation field could be an issue.
1 This is strictly true for frequencies above the radio band (ν ≳ 100 GHz). At lower frequencies it is expected that the source is opaque due to synchrotron-self absorption. Radio emission must therefore arise from larger, less compact, regions downstream of the blazar region . Back.
2 The relativistic Doppler factor δ, determining the apparent amplification of the emission, is defined by δ = 1 / [Γ(1 − β cosθv)], where Γ is the jet bulk Lorentz factor, β the jet speed and θv the viewing angle. Back.