Astron. Astrophys. 363, 455-475 (2000)

1. Introduction

Basic ingredients of the AGN complex phenomenology are universally accepted to be a central supermassive black hole and mass accretion on it. In fact, another component seems to be a general property of the physics of these objects, that is the existence of mass and energy outflows from the central region (Laing 1996, Blandford 1993, Ulrich 1988). They appear as strongly collimated, high velocity (even relativistic in many cases) jets in radio-loud objects, where they can represent the most important element both from the observational point of view and for what regards the theoretical explanation of the nuclear activity (Blandford 1994, Zensus 1996, Laing 1996), typically with non-thermal mechanisms dominating. On the other hand, radio-quiet AGNs also show clear indications for the presence of generally sub-relativistic outflows from the central zone (Ulrich 1988, Stocke et al. 1994, Crenshaw et al. 1999, Turnshek 1988, Turnshek 1995, Weymann et al. 1997), both for Seyfert 1 objects, and for more luminous and distant QSOs. In particular, UV observations, from the International Ultraviolet Explorer (IUE) (Ulrich 1988), the Hopkins Ultraviolet Telescope (HUT) and, in the last years, the Hubble Space Telescope (HST), have shown the presence of UV absorption lines blue-shifted with respect to the corresponding broad emission lines, in the spectra of a number of Seyfert 1s, indicating outflow of the absorbing gas (Crenshaw et al. 1999). This looks like the analogue of the much more remarkable and observationally well known phenomenon of broad absorption lines observed in 10-15% of QSOs (BAL QSOs), characterized by a conspicuous blue-shift with respect to the emission lines, implying much larger outflow velocity (Weymann et al. 1997, Stocke et al. 1994, Turnshek 1995). Indeed, UV absorbers seem to be a more general property of lower luminosity objects, such as Seyfert 1s, since blue-shifted UV absorption features are observed in of the AGNs of this class. Also, recent work (Crenshaw et al. 1999) suggests that, at least for Seyfert nuclei, there is a one-to-one correspondence between the objects that show intrinsic UV-absorption from outflowing material and X-ray "warm absorbers", thus showing that these two phenomena could be correlated. There have been suggestions and attempts to relate UV and X-ray absorbers, possibly identifying them with a single component of the AGN structure (see for example Mathur 1997; Mathur et al. 1995; Mathur et al. 1997). Although we could still reasonably imagine a connection between UV and X-ray absorption, recent observations have shown that, with high spectral resolution, multiple velocity components appear in UV absorption (Mathur et al. 1999; Crenshaw et al. 1999), thus indicating that the physical picture of AGN absorbers may be more complicated and the search for the identification of a single absorbing structural component can be an oversimplification of the real structure (George et al. 1998).

UV absorbers are in general modeled with outflowing clumpy material embedded in a background medium, whose relationship with the absorbing clumps is not yet unequivocally determined. We refer to de Kool (1997) for a review of dynamical models of UV absorbing gas and the problems to be solved in both Seyferts and QSOs. Indeed, a series of physical reasons requires the clumpy UV absorbing material to be basically comoving with a surrounding medium, although it is a matter of debate whether the absorbers can be thermally confined by this background outflow and what is the origin of the clumps (instabilities developing in the background wind or injected and dragged along clouds). It is not our present aim to model the details of the relationship between the UV absorbing material and the surrounding medium, however the presence of these outflowing absorbers in radio-quiet AGNs, together with the physical necessity that they are comoving with ambient material, is an indication for the existence of global outflows in these AGN classes, and it motivates our present research for a better understanding of the possible physical structure and characteristics of a global outflow, presumably originating in the very central regions and expanding out to large distances as a kind of background/connection for the various more observationally noticeable components of the AGN structure. This wind could have an important role in defining a physical connection underlying the various components of the AGN structure that can be identified by interpreting observations.

To this regard, it will be necessary to account for the relation of the wind with another phenomenologically quite distinctive component of AGNs, namely the BLR, whose modeling also is a matter of debate; in fact, the presence of a wind-type outflow as a surrounding medium for the broad line region cloud-like structures can have an influence on their physical modeling, as we briefly outline in the following. Two scenarios that could be explored are the following: a) BLR interpreted like a region in which a large number of small clouds emits the observed lines, broadened by the cloud motions; these clouds could be continuously created by the development of local condensations in a thermal-type instability of the wind, localized in the region of interest; b) broad lines emitted by winds or expanding envelopes of giant stars ("bloated star" model, see Alexander & Netzer 1994): in this case, the interaction between the AGN wind from the central region and the expanding envelopes must be analysed. This topic will be treated extensively in a forthcoming paper, whereas here we just briefly mention the main issues. Recent observational work (see Wandel et al. 1999and references therein) as shown that broad line emitting cloud-like structures should be preferentially characterized by keplerian motion, so that the interaction of such motions with the surrounding wind gas must be studied. Indeed, if the background medium is outflowing and the "clouds" have keplerian motions, a "drag force problem" does exist (Korista 1999, Mathews & Ferland 1987) and the "clouds" must be somehow continuously replenished: this would point to the choice of a "bloated star"-like scenario for BLR modeling. As a consequence, the analysis of nuclear outflows can help to define some limitations to BLR models.

Moreover, within the accepted scenario for AGN description, in which an accretion disk is thought to be feeding the central black hole, the existence of a hot, tenuous "corona" around the central part of the disk is expected, as an effect of dissipation of accretion energy in the rarefied, "peripheral" regions, more distant from the equatorial plane of the disk itself. The hot coronal gas is believed to be substantially responsible for the X-ray emission from these AGNs reproducing the main properties of the observed hard-X-ray continuum (see, for example, Haardt & Maraschi 1993, Haardt et al. 1994). The specific physics of this coronal region is not very well established. However, allowing for a sort of analogy with the solar corona, the existence of a hot plasma wind, possibly emerging from the coronal region itself, seems to be justified from a theoretical point of view as well (see Liang & Price 1977); in fact, within this framework, some of the coronal plasma could be accelerated in a hot wind-type flow or, still within the analogy with solar-stellar coronae, the origin of wind-type outflowing gas could be expected to be related to the expansion of part of the hot coronal gas, as first suggested for accretion disk coronae by Liang & Price (1977).

Our specific aim in the present paper is then to investigate and identify the physical hypotheses and conditions that allow a wind-type solution to exist, by solving the appropriate hydrodynamical equations in which we take into account the heating and cooling processes relevant to the problem, and include Compton interaction of the wind plasma with the ionizing UV-X-ray radiation field centrally generated; we look for an outflow solution whose physical properties are "acceptable" within the general AGN scenario and range of expected physical parameters for plasma in Active Nuclei.

Various authors have already analyzed some aspects of wind-type outflows in AGNs, but in general referring to very specific regions or distance scales; Weymann et al. (1982) have proposed a hot plasma wind aiming to confine BLR clouds, but they were interested in radii  cm, and also they did not take Compton interactions into account directly. Raine & O'Reilly (1993) also discussed a hydrodynamic wind model, but, again they did not take into account the regions closer to the central black hole. Several other models have been put forward in the last years, trying to connect with the BLR (see for example the magnetically supported model of Emmering et al. (1992) or Chiang & Murray 1996) or the BAL region (Murray et al. 1995). Also, recently some authors have started working on solutions envisaging both an advection-dominated disk-like inflow and an associated outflow, carrying away mass, angular momentum, and energy (see Blandford & Begelman 1999).

Indeed, the AGN context offers quite a complex situation to be studied, and in general special limits have been studied. It is our present purpose to try to solve the wind problem for the specific physical environment of the central region of a radio quiet AGN, accounting for the known physical mechanisms at work in a simplified but as complete as possible way. This also implies that we have to study this wind-type outflow on a very extended range of distances from the central black hole, which can be rather difficult in principle.

As a matter of fact, we restrict our attention to a range of luminosities identifying typical Seyfert 1 parameters, so that we look for a model wind endowed with terminal velocities , typically  km s^-1, as it is inferred from blueshifted UV absorption lines observed in Seyferts (Mathur et al. 1995, Mathur et al. 1997, Crenshaw & Kraemer 1999, Crenshaw et al. 1999), and we are not trying at present to specifically model BAL QSOs outflows, that can be much faster.

Finally, we would like to stress that the aim of the present work is essentially to construct a consistent hydrodynamical model for a basic outflow that can in principle be applied to the AGN context and can represent a kind of physical connection for the various phenomenological components of the AGN structure, identified from the observations. We defer a detailed exploration and analysis of the relation with these other components of the AGN to forthcoming papers.

The outline of the paper is the following. We summarize the characteristics we require to be fulfilled for the wind to be both physically consistent and viable to describe a global outflow in the central regions of radio-quiet AGNs in Sect. 2. The basic hydrodynamical equations governing the outflow are given in Sect. 3. Sect. 4 is devoted to the detailed description of the various heating and cooling terms we account for in the energy balance equation. We collect in Sect. 5 some qualitative considerations on AGN winds, and in Sect. 6 the problem is set in the form of wind equations and critical point definition. Since relativistic corrections can be significant for the energy equation of a sub-relativistically outflowing thermal plasma so hot that electron temperatures in its internal region can be trans-relativistic or definitely relativistic, such corrections are taken into account and described in Sect. 7. In Sect. 8 we summarize the results of our computations and the properties of our model wind-type outflow, both for models characterized by constant mass flux and for mass entraining winds; finally, Sect. 9 is devoted to the discussion of these properties and a few considerations and physical speculations referring to the specific AGN context, namely to the relation between the wind and the BLR and the wind and the UV absorbers, to be more extensively studied in future work.

© European Southern Observatory (ESO) 2000

Online publication: December 11, 2000
helpdesk.link@springer.de