Abstract

The two persistent soft gamma-ray sources in the galactic centre region are black hole candidates of stellar mass with comptonized accretion disks that radiate 1037−38 erg s−1. They appear as microquasar stellar remnants from which emanate double-sided radio jets that extend over distances of a few parsecs.

The galactic centre in soft γ-rays

Contrary to the standard X-ray band (below 20keV), where many sources have been detected at less than 5° from the galactic centre, at higher energies (35-500 keV) the field is dominated by only two persistent sources: 1E1740.7-2942 and GRS1758-258 located respectively at ∼ 50′ and ∼ 5° from Sgr A (see Fig. 1). In the 30-500 keV band these sources radiate a few ×1037 erg s−1, near the Eddington limit of collapsed objects of stellar mass. Although persistent, they are time variable. Since no γ-ray source has been detected from Sgr A, if there is a super-massive black hole at the dynamic centre of the Galaxy, at present it is in sepulchral silence.

The telescope SIGMA on board GRANAT detected a powerful burst around 420 keV from the strongest γ-ray source 1E1740.7-2942 which is interpreted as the redshifted (probably gravitationally) 511 keV line from the annihilation of e+e− pairs. Since this Einstein source can produce and annihilate 10 billion (1010) tonnes of positrons in just one second, it is now known as the “Great Annihilator”. No annihilation line has yet been detected from GRS1758-258, the second strongest soft γ-ray source in that region.

Introduction

Several authors have proposed that one of the mechanisms for fuelling active galactic nuclei (AGN) may involve the presence of a stellar bar in the galaxy (Shlosman et al., 1989 and references therein). The bar potential produces an overall instability that drives the ambient gas of the disk inwards, forming a gaseous disk sorrounding the active nucleus at a radius of several hundred parsecs. Under certain circumstances the gas can dissipate its angular momentum and fall into the centre. Numerical simulations show that the movement of the gas towards the nucleus would be preferentially along the stellar bar in two arms (Athanassoula 1992). The shocked material in the flow would be delineated by dust lanes that, in regions close to the nucleus might be seen as a ring–like structure or “inner spiral arms”. We present here evidence of such a structure surrounding the nucleus in the Seyfert 1 galaxy NGC4151.

Results

Broad–band images of the nucleus of NGC4151 were obtained in the U,B,V,I and Z filters with the 1m JKT telescope in La Palma. The images were calibrated in magnitudes and then subtracted in pairs to produce a series of colour maps. The B and V images are clearly contaminated by extended emission lines. However, a line–free colour map was obtained by combining a continuum image obtained by Pérez et al. (1989) in a filter centred at 5700Å, with the I-band image.

Due to the combination of high radiation densities in the BLR and the high column densities of clouds which are optically thick to the Lyman continuum, many lines emitted from the BLR are likely to be optically thick, and hence will be emitted anisotropically. Ferland et al. (1992) discussed the case where the amount of anisotropy is constant for all clouds, but in reality the anisotropy will vary for a spatially extended BLR, depending on the local physical conditions. To study this effect, we have computed several models for spherical BLR's populated by an ensemble of spherical clouds in pressure balance with an intercloud medium, using our BLR modelling code, PROSYN (Goad, O'Brien & Gondhalekar 1993), which utilizes the photoionization code CLOUDY (Ferland 1991). Here we show some of the results for a constant pressure BLR model, where the cloud density (N) and column density (Ncol) are constant with radius (r), and hence the ionization parameter (U) varies as r−2. The model was normalized to have N = 1010 cm−3, Ncol = 1023.75 cm−2 and U = 10−2 at a radius of 10 light-days. The inner and outer radii were set at 0.3 and 56.2 light-days respectively, giving a range in log U from 1.0 to −3.5.

The total line emissivity εtotl = εin + εout, where εin is the emissivity of the inward cloud face (towards the ionizing continuum source) and εout is the outward emissivity (away from the ionizing continuum source).

Abstract

Double radio sources provide perhaps the most graphic evidence for anisotropy in AGN, and observations of their large-scale structure can potentially give us a direct means of measuring the emission axis of the AGN itself. Attempts to unify radio sources using orientation and doppler boosting arguments alone have been relatively successful, however recent discoveries of correlated asymmetries on the kpc- to Mpc-scale in the radio, optical and IR may pose serious problems for such models.

Introduction

At first sight a paper about the large-scale (kpc – Mpc) structure in radio galaxies appears to have little direct relevance to the nature of the compact object in AGN. If we are considering evidence for anisotropy however then radio sources provide the most graphic evidence for such emission from the AGN – for example 4C74.26 (Fig. 1) shows all the principal features: an essentially linear structure with a one-sided jet linking the core with the southern lobe, the alignment of core and hotspots and a one-sided VLBI jet aligned with the large-scale jet. Potentially, radio sources could provide us with one of the best ways of determining the orientation of the emission axis of the AGN, however the recent discovery of large-scale correlated asymmetries, may not only make this a difficult task, but may also provide strong evidence against unified models of AGN phenomena in terms of orientation dependencies.

So as to limit my discussion I will consider only high-power objects with P(178 MHz) ≥ 1025.5 W Hz sr−1.

Abstract

We present a [O III] λ5007 image of the nuclear region of the Seyfert 1 galaxy NGC 4151 obtained with the Planetary Camera on the Hubble Space Telescope. The [O III] image shows a striking biconical structure centred on the bright, unresolved central source. Simple geometric arguments place our line of sight outside the cone of ionizing radiation. Since we have a nearly unobstructed view of the UV continuum and broad-line region, an optically thick molecular torus cannot be the source of the collimation. Lower column density material visible in UV spectra is largely transparent at UV and optical wavelengths, but opaque beyond the Lyman limit. It can collimate the ionizing radiation field without obscuring our view of the central engine.

The simplest unified models of Seyfert 1 and Seyfert 2 galaxies invoke a dense, optically thick, molecular torus surrounding the central engine and broad-line region which serves as both a shield which prevents us from viewing the centres of Seyfert 2's and as the source of the collimation for the ionizing radiation. The difference between Seyfert 1's and Seyfert 2's is then due entirely to the opening angle of the torus and its orientation relative to our line of sight. The observations presented here and by Evans et al. present a direct challenge to this simple view.

Figure 1 shows the raw image obtained with the Planetary Camera on the Hubble Space Telescope through the F502N filter in a 1608 s exposure.

Abstract

Here is presented a statistical test for the accretion disc model. The effect of different inclinations of the disc to the line of sight is shown. Then the possibility that the ratio R of the radio core luminosity to the radio lobe luminosity is a good orientation indicator is investigated. It is shown that R. describes the intrinsic power of the central radio source.

A statistical test of the accretion disc model

The model adopted here is that, described by Collin and Dumont (1989), of an accretion disc illuminated by the central nonthermal source. The disc emits in its inner part the UV and optical continuum, and in its outer parts the low ionization lines (LIL; e.g., the Balmer lines) as proposed by Collin (1987). A second region emits the high ionization lines (HIL; e.g., Lα and CIV); it can have a spherical geometry. Both regions are illuminated by the central source. For the disc, illumination occurs through backscattering of the nonthermal radiation by a hot medium (halo). This model has been worked out by Dumont and Joly (1992).

Using the code of Dumont and Collin (1990) we have computed the line spectrum emitted by the illuminated disc as a function of the central source luminosity, the Eddington ratio (Lbol\LEdd) and the characteristics of the scattering halo (density structure and optical thickness).

Abstract

The γ-ray emission from three of the sources detected by the COMPTON Observatory is discussed in the framework of relativistic jets emitting via the synchrotron self-Compton mechanism. The physical conditions in the three sources are derived and compared.

One of the most exciting discoveries of the COMPTON Observatory is the observation of γ-ray emission from 16 blazars. The broad band continuum (radio to X-ray) from these objects can be understood on the basis of the synchrotron self-Compton process in an inhomogeneous jet of plasma moving with relativistic speed (e.g.). In this model the synchrotron emission usually extends up to the UV and possibly to the X-ray bands. It is therefore natural to expect γ-rays produced by first order self-Compton scattering. The strength of this component depends on the ratio between the radiation and magnetic energy densities, and on the optical depth to photon-photon absorption. The latter process can be avoided if the source is beamed, and independent evidence for beaming is indeed found from radio and X-ray observations.

Nevertheless the γ-ray intensity measured from 3C 279 was a great surprise, in that the γ-ray power exceeded that in all other bands. This requires the radiation energy density to be larger than the magnetic energy density, a situation usually called the “Compton catastrophe” since at first sight it gives rise to divergent Compton energy losses.

Abstract

We present the results of a monitoring campaign of the low luminosity Seyfert 1 NGC 4593, at X-ray, ultraviolet, optical and near IR frequencies. The rapid and large amplitude fluctuations detected set interesting upper limits to the sizes of the various emitting regions. We investigate the properties of the nuclear energy distribution, its pattern of variability and the relationship between the emission processes at different frequencies. We also study the variations of the broad emission lines in relation to those of the continuum. The implications of our data are discussed. It is found that they are consistent with a very hot accretion disk illuminated by the hard X-ray source.

Observations

Repeated observations of the nucleus of NGC 4593 have been performed from the near IR to X-rays. Though the temporal density of our observations is poor, they have the advantage of a wide wavelength coverage. Besides, NGC 4593 turned out to be highly variable at all frequencies.

Results and implications

The strong and rapid variations detected in all wavebands imply that the continuum source is unusually compact: ≤ 1.1 lt-hr in the hard X-rays, ≤ 3 lt-hrs in the soft X-rays, 70 lt-hrs at 1450 Å, and 37 lt-days at 2.5 microns. The latter corresponds roughly to the dust evaporation radius (∼ 33 lt-d, Barvainis, 1987) in such a low luminosity source. Such an IR behaviour is reminiscent to that of Fairall 9, a more luminous Seyfert 1, (Clavel et al., 1989).

Abstract

Accretion disks are the preferred central engine for AGN, but theoretical progress has long been hampered by the unknown nature of the angular momentum transport mechanism. An obstacle preventing the general acceptance of turbulent disk theories has been the intransigent stability of thin Keplerian disks to hydrodynamic perturbations. This difficulty is overcome by the result discussed here, that a weak magnetic field renders a disk locally unstable whenever the angular velocity decreases outward. The salient and remarkable feature that distinguishes this instability is that the maximal growth rate (which is of order the angular velocity) is independent of the strength of the field, even as the latter vanishes. The instability can be derived straightforwardly from the equations of orbital mechanics and a spring-like magnetic interaction. We list several key questions regarding the instability's implications for realistic disks, and review the results from two-dimensional numerical simulations.

Introduction

Many of the scenarios created to explain the spectra from Active Galactic Nuclei (AGN) involve, either implicitly or explicitly, the release of gravitational energy through accretion. The classical paper of Lynden-Bell (1969) was the first to suggest that the heart of an active nucleus consisted of an accretion disk surrounding a supermassive black hole. The appeal of this model (and its direct descendents) has always been one of energetics: no other proposed mechanism is as efficient and as compact. Gravitational potential energy release is hard to beat as an energy production mechanism.

Abstract

From spectropolarimetric observations of the galaxy, IRAS 110548-1131 we report strong, broad Hα emission (FWHM 7600 km s−1) in the polarized flux spectrum. This suggests that IRAS 110548-1131 has an obscured broad line region, whose radiation is scattered into our-line-of-sight by scatterers outside the obscured region.

Introduction

Since the discovery by Antonucci & Miller (1985) and its confirmation by Bailey et al. (1988) that the type II Seyfert galaxy NGC 1068 has broad hydrogen Balmer lines in it's polarized flux spectrum, and therefore has a Seyfert I type nucleus, it has been possible to construct a physical model in which the two types of Seyfert are in fact the same. Whether an object is seen as a type I or II depends upon the orientation of the galaxy and obscuration of the broad line region, probably in the form of a dusty torus, although the universality of such a model is an open question. Radiation from the BLR can escape along the axis of the torus and then be scattered into the line-of-sight by electrons and/or dust. The scattered component of these lines should then be observable in polarized flux.

In recent years there has been considerable interest in those galaxies identified from the IRAS survey which are very luminous in the far infrared region. Some evidence that highly luminous IRAS galaxies contain obscured QSO nuclei comes from the work of Hough et al.

Introduction

X-ray emission provides ∼ 10% of the bolometric luminosity of a typical Seyfert galaxy and as X-ray photons of energy > 2 keV can penetrate column densities of > 1 × 1022 atoms cm−2, such observations provide the best observational probe available (with current instrumentation) of the active nucleus, and its immediate environment.

Variations cannot be observed faster than the light travel time, thus δt ∼ r/c gives an upper limit to the size of the emitting region. A variability timescale may be associated with the dynamical (orbital) timescale of the inner accretion disk. If the X-rays arise from r ∼ 5rs and rs = 2GM/c2 then t ∼ 50M6S (M6 are units of 106 solar mass). This is reasonable if a persistent period were found, but could be misleading otherwise.

Two other parameters of interest are efficiency η (of mass to energy conversion) and compactness ℓ. Efficiency, η ∼ 5 × 10−43dL/dt, if η > 0.1 an exotic mechanism is required (such as relativistic beaming). Compactness, ℓ = Lστ/Rmec3 (Svensson 1986; Guilbert, Fabian & Rees 1983). When ℓ > 10 the source becomes optically thick to γ-rays and pair production becomes important (assuming the spectrum extends to ∼> 1MeV), affecting the X-ray spectrum and the temporal behaviour (e.g., Mosalik & Sikora 1986; Fabian et al. 1986).

Historical

Ariel V and HEAO-1 established long term (days to years) variability as a property of AGN (e.g., Marshall, Warwick & Pounds 1981), with variations in amplitude of factors of ∼ 7 or so being common.

The high sensitivity and improved spectral resolution (ΔE/E = 0.4 at 1 keV) of ROSAT has enabled high quality spectra to be obtained in the soft X-ray (0.1−2 keV) band. When added to simultaneous Ginga data, spectra extending up to 20 keV are achieved.

3C273 was observed by ROSAT in a pointed observation in June 1990 and again in December 1990 during the survey, when it was also observed by Ginga. The December observation yielded a 3σ detection in the SI filter of the Wide Field Camera. E1821+643, benefits from being at a high galactic latitude and was observed for a total of 40 days during the ROSAT survey, with an effective exposure time of 9ks. Ginga observed the source during this period. It was also observed serendipitously during the pointed phase in February 1991.

Both sources need a two component model to fit their combined 0.1−20 keV spectrum (see Tables 1 & 2). The spectrum of 3C273 has a significant soft excess which exists below 1.2 keV (source rest frame). This may be adequately parameterised by a 200eV Bremsstrahlung or a power law of α = 3.4, however, a single temperature Blackbody or Raymond-Smith model is not a good fit to the data, neither are elementary disc models.

E1821+643 has a strong, steep soft excess which rises below 0.65 keV (source rest frame) in both ROSAT observations.

Abstract

Recent work has revealed that a number of Seyfert 2 galaxies exhibit conicallyshaped regions of gas apparently illuminated by a collimated, nuclear ionizing source. In this work, we test one model for this collimation, namely that the cones result from shadowing of a compact nuclear continuum source by a thick, dusty torus. From the emission-line ratios measured for gas within the cones, we have calculated the number of ionizing photons emitted by the compact nucleus. Then, on the assumption that the nuclear source radiates isotropically, we have found the power incident on the torus, which is expected to be reradiated in the infrared. Given the uncertainties in the calculation, and the fact that the torus may be somewhat anisotropic in the infrared, we find the observed IRAS luminosities are consistent with the torus model in 9 objects with sufficient data to perform the calculation.

Introduction

There is now considerable evidence that the ionizing photons escape from the nuclear source anisotropically in many Seyfert 2 galaxies (see Wilson 1992 for a recent review). For some objects, this anisotropy becomes evident in the elongated morphologies detected in images taken through narrow-band filters centred in high excitation lines (e.g. [OIII]λ5007) or through “ionization maps” (Pogge 1988a,b) – ratios between the continuum subtracted images in [OIII]λ5007 and in Hα+ [NII]λλ6548,6583.

In this work, we combine new with existing data for a sample of nine Seyfert 2 galaxies with known “ionization cones” in order to test whether the collimation is the result of shadowing of radiation from a small, isotropic, nuclear source by a thick dusty torus, as suggested in the “Unified Scheme” for Seyfert galaxies (Antonucci & Miller 1985; Krolik & Begelman 1986).

Abstract

The structure and emission properties of accretion discs are briefly reviewed, with special attention given to geometrically thin “standard” accretion discs. Different approximations used to compute their emission spectrum are summarized, and it is shown that the bulk of the luminosity is emitted in the EUV range (the “UV bump”). However, according to recent optical, UV and X–ray monitoring campaigns of Seyfert galaxies, it appears that a large fraction of the observed optical–UV and even X–ray continuum is reprocessed by the disc from a primary source, probably of hard X–ray continuum, sitting near the black hole and heating the farther regions of the disc. A first consequence is that the disc must also contribute some fraction of the line emission. A second consequence is that standard accretion discs drawing their energy from viscosity and radiating it locally are ruled out in AGN, and another type of model must be built.

Generalities on accretion discs

For the reader who wishes to know more than just the basic properties that are recalled below, several reviews on accretion discs can be recommended: Pringle (1981), Frank, King & Raine (1992; revised version), Begelman (1985), Shields (1990), Treves, Maraschi & Abramovicz (1988).

If the matter accreted onto the black hole possesses angular momentum, it will settle in an accretion disc, in which matter is transported inward and angular momentum outward. This may be achieved through viscous friction, or through other stresses such as produced by a magnetic field.

Abstract

We have previously reported evidence for radio-luminosity dependence of the strength of the near-infrared ‘alignment effect’ in radio galaxies (Dunlop & Peacock 1991). Here we present evidence for an associated radio-luminosity dependence in the level of optical activity found in radio galaxies at z ≃ 1. We find that this correlation, the strength of which is maximised by considering a combination of radio power and spectral index, is very similar and probably closely related to the correlation between radio-jet power and LNLR found by Rawlings & Saunders (1991); the available evidence suggests that both correlations in fact arise from an underlying correlation with environment.

These correlations along with (i) the universal shape of the supposedly stellar UV-continuum in powerful high-redshift radio galaxies; (ii) the detections of significant optical/UV polarisation in several 3CR radio galaxies; (iii) the inaccuracy of the optical-radio alignments; (iv) the close spatial correspondence between the extended UV continuum and line emission; and (v) the correlation between radio-lobe depolarisation and extended optical emission, indicate that a large fraction of the optical/UV activity and the optical alignment effect in the 3CR sample is the result of Thomson scattering of a ‘flat’ (fv ∝ v−0.2) quasar continuum emitted within a broad cone centred on the radio axis.

To investigate the relationship between the optical/UV activity in high-z radio galaxies and their radio properties, we have considered a composite sample of radio galaxies with z ≃ 1 which spans a decade in radio luminosity.

Abstract

Much effort has gone into analysing the now well-known luminosity evolution of the quasar population. Theoretical models, such as the accretion disk model, however, predict a spectral evolution which is more complicated than a simple overall luminosity shift [3], and more detailed spectral information is necessary to compare physical models with the data. We investigate the distribution and evolution of quasars on the color-flux plane in order to test for consistency with the thin accretion disk model.

Introduction

Much effort has gone into analysing the now well-known luminosity evolution of the quasar population. Theoretical models, such as the accretion disk model, however, predict a spectral evolution which is more complicated than a simple overall luminosity shift [3], and more detailed spectral information is necessary to compare physical models with the data. We investigate the distribution and evolution of quasars on the colour-flux plane in order to test for consistency with the thin accretion disk model.

Quasar Colours

It has been known since the first quasar surveys that active galactic nuclei (AGN) occupy a specific region of the two colour diagram, separate from most other galactic and extragalactic sources. Consequently, UBV photometry has been used to select AGN in numerous surveys up to a redshift z ∼ 2.2 where the Lα emission line and absorption edge are redshifted into the observed bands.

Abstract

A technique is presented to deconvolve variable Seyfert 1 profiles that consist of components that differ in radial velocity coverage and time scale of variability.

Introduction

Variable structure in the emission-line profiles of AGN can in principle be used to map out the kinematics of the broad-line region (BLR). However, profile variations can also occur if the profile consists of more than one component and each component is variable in flux on a different time scale (Peterson et al. 1990). In order to use profile variations as a probe of the BLR, it is important to devise techniques to deconvolve components that may represent physically distinct emission-line regions. One possible technique will be outlined using data obtained from the 1989 campaign to monitor the Seyfert 1 galaxy NGC 5548 with the International Ultraviolet Explorer (IUE), as well as IUE spectra obtained by Webb and Crenshaw in April, 1990, when NGC 5548 was in a particularly low state.

Figure 1 shows that the profile of C IV λ1550 is much narrower in the low state, suggesting that there is a strong narrow component that has not varied as much as the rest of the profile. In addition, it is clear that excess emission in the blue wing, which Peterson et al. (1990) claim is a distinct component that varies on a longer timescale than the majority of the broad-line flux, has nearly disappeared.