Forum Springer Astron. Astrophys.
Forum Whats New Search Orders

Astron. Astrophys. 351, 487-494 (1999)

Previous Section Next Section Title Page Table of Contents

5. Discussion

The results obtained here indicate that the two families of early type galaxies recently singled out have also a distinct X-ray behavior in the (0.2-2) keV band. The trend between [FORMULA] and the shape of the central profile is similar to the trends between [FORMULA] and the isophotal shape distortion, or the degree of rotational support. However, it is much sharper than the other two, and it could be at the origin of the difference in the X-ray properties.

What are the possible causes of this segregation of power law galaxies to [FORMULA] [FORMULA][FORMULA], and how do core ones evade it? Can the change in the shape of the density profile in the central regions be responsible for such a disparate behavior of [FORMULA]? Or are some other galaxy properties, to which the change in the shape of the surface brightness profile is linked, responsible for this difference? For example, besides isophotal shape and rotation, distinct evolutions, where an important role is played also by massive black holes, could be responsible for the settling of the two types of profiles (Sect. 1) and also have an effect on [FORMULA].

In the following I speculate on a few conceivable causes of the different [FORMULA] behavior in the two families of early type galaxies. I will focus first on the direct effect of the inner profile shape on [FORMULA]; then on that of a massive black hole (MBH); finally on those expected from possible differences in the environment. I assume that the scatter in [FORMULA] is related to a largely varying quantity of hot gas within galaxies, as demonstrated observationally by considering their spectral properties (e.g., Matsumoto et al. 1997).

5.1. Direct effects of the basic properties of power law and core galaxies on their hot gas flows

Ciotti et al. (1991) simulated the evolution of the hot gas behavior over the galaxy lifetime for spherical mass distributions having King profiles (i.e., with flat cores), including type Ia supernova heating. They showed that the large [FORMULA] range observed at fixed [FORMULA] can be produced by small differences in the gravitational potential energy of the gas, which cause the gas flows to be in the wind, outflow or inflow stages. Later Pellegrini & Ciotti (1998) produced a set of hydrodynamical simulations where the spatial luminosity distribution at small radii has a power law form ([FORMULA] as in the Jaffe law; this corresponds to the central luminosity density profile of power law galaxies, Gebhardt et al. 1996). The resulting gas flows are mostly partial winds: an inflow develops in the central region, while the external parts are still degassing. A large spread in [FORMULA] at fixed [FORMULA] is again produced by varying the model input parameters in their observed ranges, and corresponds to variations in the size of the inflow region. In particular, for [FORMULA]log [FORMULA] [FORMULA], [FORMULA] values as high as [FORMULA] erg s-1 can be easily reached. So, in the X-ray luminosity values of a large sample of galaxies there should be no trace of the precise shape of the mass profile in the very central regions. This is supported also by the observations that show how all power law galaxies have low [FORMULA] independently of the [FORMULA] values (Fig. 2), while core galaxies have a large range of [FORMULA] with little variation of [FORMULA].

Galactic rotation is an important property of power law galaxies (Sect. 1), and so one might suspect that the simulations mentioned above are not realistic for these galaxies. However, as shown by Ciotti & Pellegrini (1996) and by D'Ercole & Ciotti (1998) with simulations, rotation has minor effects on [FORMULA]. If no type Ia supernova's heating is assumed (so that the flows are always cooling flows) the effect is larger: a reduction of [FORMULA] by a factor of six is obtained when [FORMULA], and a very massive cold disk forms (of mass [FORMULA]; Brighenti & Mathews 1996). Even though in this case the effect of rotation goes in the right direction, the observations cannot be fully explained, because the needed reduction in [FORMULA] is at least of one order of magnitude. Moreover, for [FORMULA] there are both power law and X-ray bright core galaxies (see Fig. 4), and so the amount of rotation cannot be taken as the general explanation.

A flatter global galaxy shape is another property known to be associated on average to lower [FORMULA] values, from observations (Eskridge et al. 1995). The hot gas is less bound when the mass distribution is flatter, for fixed [FORMULA], so that an outflow is favoured (D'Ercole & Ciotti 1998). This is the case of galaxies where a disk is present, and might apply preferentially to power law galaxies. Those in the present sample, for log [FORMULA] [FORMULA], include four S0s, but also three roundish (E0-E2) galaxies. So, again the presence of a disk has an effect that goes in the right direction, but cannot be taken as a general explanation (unless all the above mentioned power law galaxies with rounder shape turned out to be disk galaxies viewed face on).

In conclusion, large systematic variations of the hot gas content resulting from differences of the inner profile shape are not expected, based on numerical simulations; galactic rotation and flattening of the mass distribution are expected to have an effect, but seem to be unable to offer a general explanation. This is consistent with the [FORMULA]-[FORMULA] relation being sharper than the [FORMULA]-[FORMULA] one (and the [FORMULA]-[FORMULA] one, where [FORMULA] is the projected galaxy ellipticity, Pellegrini et al. 1997).

5.2. Effect of the presence of a massive black hole

When dealing with properties connected with the very central regions of early type galaxies, one is naturally brought to consider the effects produced by a central MBH. In fact, on the basis of various kinds of observational evidence, in the past few years it has become commonly accepted that early type galaxies host central massive black holes, possibly remnants of dead or quiescent QSOs (Kormendy & Richstone 1995, Richstone et al. 1998). Why the X-ray emission in early type galaxies is not as high as in classical AGNs, due to accretion of hot gas by the central MBH, is a problem that has been addressed by Binney & Tabor (1995), Fabian & Rees (1995) and by Ciotti & Ostriker (1997). Fabian & Rees suggest that accretion of gas from the surrounding cooling flow has a very low radiative efficiency, and that early type galaxies should typically host low or very low luminosity AGNs. Ciotti & Ostriker suggest that the hot gas can be expelled from the galaxies when a central MBH is present, because the gas flows are found to be unstable due to Compton heating. So, most of the time the galaxies are in a low or medium [FORMULA] state, they are in a high [FORMULA] state only during global inflows, and the large observed scatter in [FORMULA] is reproduced.

In both frameworks it is not clear why power law galaxies are found only in the low or medium state, while core galaxies span the whole observed scatter in [FORMULA]. A few solutions, where other ingredients besides the presence of a MBH enter, can be imagined: the shape of the galaxies (triaxial versus axisymmetric) can have an effect on the way accretion proceeds. Or the very MBH properties (such as its mass) could play a role. In any case if a MBH is important in determining [FORMULA], the conventional approach to the problem of interpreting the X-ray properties of early type galaxies requires revision; the input ingredients are currently just stellar mass loss, supernova heating, and gas potential energy (see, e.g., Sarazin & White 1988, David et al. 1990, Ciotti et al. 1991). More precisely, since a different amount of hot gas is at the origin of the observed large scatter in [FORMULA] at fixed [FORMULA] (e.g., Matsumoto et al. 1997), one should focus on the relation between central MBH presence and amount of hot gas.

5.3. Effect of the environment

F97 present a preliminary evidence that core galaxies tend to be found in dense groups and clusters, while power law ones are preferentially found in the field. This is in line with the fact that core galaxies are on average also boxy/irregular (which has been argued to be a merger signature, or a sign of past accretion and interaction events; Nieto 1989, Barnes 1992) and show more frequently indications of accretion of small satellites such as cores within cores, central counter-rotation, etc. (Lauer et al. 1995). In this scenario, what could be the effects on [FORMULA]? More generally, what is the effect of the environment on the hot gas content?

An environment denser of intragroup/intracluster gas and of galaxies should act in the sense of producing both extremely X-ray luminous galaxies and galaxies with medium or low [FORMULA]. In fact accretion of external gas is needed to explain the X-ray brightest galaxies of the [FORMULA]-[FORMULA] diagram (Renzini et al. 1993, Mathews & Brighenti 1998) while in the X-ray faint ones the hot gaseous halos could have been stripped by ambient gas, if it is sufficiently dense, or in encounters with other galaxies (White & Sarazin 1991). This picture would be consistent with the finding that the larger spread in [FORMULA] is shown by core galaxies, that are preferentially found in high density environments (and this is so even for the sample studied here). Also it has been suggested that MBH presence and galaxy environment might be connected (F97); in this view cores might result from ejection of stars by a coalescing black hole binary during a merging event (Quinlan & Hernquist 1997, Nakano & Makino 1999). So, the association of cores and a large spread in X-ray emission could come from a combination of environmental effects and MBH presence; for example, merging produces triaxiality, and this facilitates the gas to reach the nucleus, and feed the central MBH. Note that just the variety of merging and/or tidal interaction conditions (time of the event during the galaxy history, progenitors' masses and orbits, and so on) might produce a variety of effects on the hot gas content. Numerical simulations are needed to derive more firm conclusions concerning this aspect; the first results show that the effect of perturbing the hot gas flows is in the direction of producing a spread in [FORMULA] (D'Ercole et al. 1999).

While various solutions can be imagined to explain the large range of [FORMULA] shown by core galaxies, again a major problem is why power law galaxies do not reach high X-ray luminosities, even when they are in a range of [FORMULA] values where core galaxies can be very X-ray luminous. Certainly, they do not seem to experience accretion from outside, because they do not even show [FORMULA] values as high as the numerical simulations for the hot gas evolution predict for them when they are isolated (Sect. 5.1). So, they should reside in the field 4, or never be the central dominant galaxies in clusters, subclusters and groups. But this seems not to be the case: for example NGC1553, NGC4697 and NGC5198 are also the brightest members of their groups. So, the problem remains of why power law galaxies do not become very X-ray bright, even when they are optically dominant.

Previous Section Next Section Title Page Table of Contents

© European Southern Observatory (ESO) 1999

Online publication: November 3, 1999