Annu. Rev. Astron. Astrophys. 1982. 20:
547-85 Copyright © 1982 by Annual Reviews. All rights reserved |
4.2 Virgo - An Unevolved Cluster with a Dominant Galaxy
Virgo appears to be a cluster in the early stages of dynamical evolution. Its general appearance is irregular (Abell 1975), and it is classified as Bautz-Morgan type III (Bautz & Morgan 1970). The optical location of the cluster center is poorly defined (Sandage & Tammann 1976, de Vaucouleurs 1961, Bahcall 1974). The velocity dispersion, 673-20+45 km s-1 (Danese et al. 1980), is lower than that of evolved rich clusters, indicating a shallow gravitational potential. The spiral fraction (55%) is the highest measured by Bahcall (1977b) for X-ray-emitting clusters.
Figure 5 shows a composite Einstein X-ray image of the cluster core. The X-ray emission is dominated by thermal emission from a gas of temperature ~ 2 keV associated with M87. Several less luminous sources of smaller angular size are observed to be associated with galaxies, along with numerous background and foreground objects (e.g. QSOs and galactic stars). The total Virgo X-ray luminosity is 3 x 1043 erg s-1 (2-10 keV; Mushotzky & Smith 1980).
Figure 5. Fifteen IPC images centered on M87 in the Virgo cluster are merged. The fields are normalized to the same exposure time and corrected for telescope vignetting. Increasing brightness corresponds to increasing surface brightness. The tic-tac-toe pattern results from the shadowing of the detector supports and is most apparent in the bright central region dominated by M87. M86 is the bright extended region to the northwest (upper right). Numerous other galaxies as well as background and foreground objects can be seen. The scale of the figure is given by the size of each IPC field-of-view of 1° x 1°. |
The Virgo cluster core gas. excluding M87, has a density of ~ 5 x 10-4 cm-3 and a temperature of ~ 2-3 keV (see discussion of M87 below). This gas plays a role in the evolution of the galaxies within the core, where the timescales for thermal evaporation and ram-pressure stripping are shortest. While the proximity of the Virgo cluster allows the study of individual galaxies, the cluster's great extent complicates the study of the diffuse cluster gas. Therefore this discussion is primarily restricted to the galaxies, and it is divided into two sections - The first dealing with M87 and the second with the other cluster galaxies.
4.2.1 M87 The X-ray emission surrounding M87 is primarily thermal bremsstrahlung from hot gas with nearly solar abundances (Lea et al. 1982). The jet, nucleus, and radio lobes produce X-rays but account for only ~ 1% of the X-ray flux in the Einstein energy range (Schreier et al., in preparation). Fabricant et al. (1980) analyzed the IPC observations of M87 using techniques similar to those developed by Bahcall & Sarazin (1977) and Mathews (1978).
They used the X-ray surface brightness distribution to derive the total mass as a function of radius, assuming that the X-ray-emitting gas is in quasi-hydrostatic equilibrium. The equations of hydrostatic equilibrium and the ideal gas law give
where M(r) = total mass within radius r,
= X-ray-emitting gas
density, and TkeV temperature of the X-ray gas. The steepest
temperature gradient allowed by the spatially resolved IPC spectra
gives a minimum mass of 1.7 x 1013
M within a
radius of 50 arcmin (~ 290 kpc). As shown in
Figure 6, this large total mass dominates both
the mass in gas around M87 and the visible galaxy mass by more than an
order of magnitude.
Figure 6. The approximate density profiles
of the halo, the
X-ray-emitting gas, and the visible matter in M87 are plotted against
radius (taken from
Fabricant et
al. 1980).
The dark halo dominates the mass distribution beyond 5 arcmin.
Detailed calculations have been made for the X-ray gas surrounding
M87. Accretion models in which the gas near slowly
moving galaxies is
sufficiently dense to cool in less than a Hubble time have been described by
Cowie & Binney (1977),
Fabian & Nulsen
(1977), and
Mathews & Bregman
(1978).
An alternative model in which conduction
plays a major role was discussed by
Takahara & Takahara
(1979).
Einstein spectroscopic observations place stringent limits on
these models. Canizares (1981) summarized the FPCS observations of
the inner few arcmin of M87, which show evidence for gas with
temperatures ranging from 4 x 106 to 4 x 107 K
based on the detection
of emission lines. These data constrain the mass accretion rate to
3-4 M
yr-1
(Canizares et
al. 1982).
The SSS observations centered on
M87 confirm the presence of multiple components of cool gas
(Lea et al. 1982).
These spectral observations are in general agreement with
models invoking radiative cooling and accretion, but conflict with the
Takahara & Takahara
(1979)
model, which does not allow for the
existence of cooler gas in the innermost few arcmin.
Binney & Cowie (1981)
developed a model to explain the cool gas
components and also to reduce the total mass attributed to M87 by
Fabricant et
al. (1980).
They accomplished this by (a) relaxing the
temperature constraints derived by Fabricant et al. and introducing a
steeper temperature gradient, (b) confining the M87 gas by a hot
external cluster medium suggested from observations by
Lawrence (1978)
and Davison (1978),
and (c) introducing a cluster potential. However,
in comparing the available spectroscopic observations,
Lea et al. (1982)
showed that the hard component of the Virgo cluster X-ray
spectrum is probably produced by the nuclear synchrotron source of M87
(detected in the HRI observations by Schreier et al.) and that the
diffuse cluster emission beyond M87 must have a temperature of ~ 2.5
keV, comparable to M87's circum-galactic gas. Thus the basic
assumption of a hot intracluster medium required in the Binney & Cowie
model is probably incorrect, and we conclude that M87 is as massive as
Fabricant et al. and previous authors had suggested.
Although M87 is unique in the Virgo cluster, the bright, central
elliptical galaxies in Abell 262 and Abell 1060 are similar in X-rays
(Forman et al. 1981b).
The X-ray luminosities and gas temperatures of
Abell 262 and Abell 1060 are comparable to those of Virgo
(Mushotzky & Smith
1980).
Optically all three clusters are spiral-rich
(Bahcall 1977b)
and irregular with low velocity dispersions
(Danese et al. 1980).
The X-ray surface brightness profiles for these galaxies
also suggest the presence of massive halos.
4.2.2 THE VIRGO GALAXIES M84 AND M86
In addition to M87, more than a dozen Virgo galaxies have been
observed to be X-ray sources with luminosities from ~ 1040 to
~ 1041
erg s-1 in the Einstein 0.5-3.0 keV band
(Forman et al. 1979,
Palumbo et al. 1981).
For most Virgo galaxies, the X-ray observations measure
the luminosity but have inadequate resolution to differentiate between
small X-ray halos and compact nuclear sources. However, for the galaxy
M86, the X-ray isointensity contours
(Figure 7)
show an X-ray plume
northwest of the galaxy that is embedded in an extensive region of
emission. Also seen in Figure 7 is M84, which
was resolved only when
viewed with the HRI. The IPC spectral information for the central
enhancement on M86, the plume, and M84 all suggest gas temperatures of
~ 0.5-3.0 keV, which are consistent with or slightly cooler than that
of M87
(Forman et al. 1979).
Parameters of the gas distribution about
M86 (derived from the radial distribution to the
south) and M84 are
given in Table 2. The luminosity of the plume
region is ~ 5 x 1040 erg
s-1, and that of the entire 100 kpc region about M86 is 2 x
1041 erg
s-1. The emission seen around M84 and M86 is comparable to that
reported for the enhancements in Abell 1367
(Bechtold et al. 1983).
Figure 7. X-ray isointensity contours of a
smoothed 0.5-3.0 keV IPC
image of M86 and M84 (to the west) are superposed on an optical
photograph. The extended emission surrounding M86 is seen along with
the plume to the north. The scale of the image is given by the 17'
separation of M86 and M84.
A model for the M86 and M84 X-ray emission is suggested by the
galaxies disparate radial velocities. M86 has a velocity of approach
of ~ 1500 km s-1 with respect to the Virgo mean, while that
of M84 is within 100 km s-1 of the mean
(Tammann 1972).
Forman et
al. (1979) and
Fabian et al. (1980)
proposed that the gas associated with M86
originated within the galaxy, because its high velocity, the long
cooling time of the intracluster gas, and the inability of the galaxy
to enhance the local density prevent accretion of the intracluster
gas. They suggested that M86 travels on a radial path crossing the
cluster core once every 5 x 109 yr. Given a mass of ~
1012
M and a
specific mass-loss rate of 10-12 yr-1
(Gisler 1979),
M86 approaches the
cluster core with a gaseous halo of 5 x 109
M, which is
roughly what
is observed. The X-ray plume is gas that has been ram-pressure
stripped from M86 but not yet incorporated into the intracluster
gas. The extended emission centered on M86 originates from gas still
bound by the galaxy.
Several researchers have studied the interaction of gas in galaxies
with an intracluster medium
(Ruderman & Spiegel
1971,
Schipper 1974,
Gunn & Gott 1972,
Lea & De Young 1976,
Gisler 1976,
Nulsen 1982).
None of the present models consider all the effects important in the
evolution of the gas around M86. First, the galaxy velocity and the
external intracluster gas density change dramatically every few
billion years as M86 moves from the dense cluster core at high
velocity to the cluster outskirts at low velocity. Furtheremore, the
effects of radiative cooling and conduction have not been fully explored.
The small velocity of M84 relative to the Virgo mean suggests it is
a permanent resident of the cluster core. The mass of the
X-ray-emitting gas around M84 is about 10% of that around M86,
consistent with the suggestion that M84 resides in the cluster core
and experiences more ram-pressure stripping and evaporation than M86.
M86 d M84
Core radius a
3' (18 kpc) 0.5' (3 kpc)
X-ray luminosity b
2 x 1041 erg s-1 5 x 1040
erg s-1
Central density a
4 x 10-3 cm 8 x 10-2 cm
Mass of X-ray gas a
6 x 109
M
5 x 108
M
Cooling time c
4 x 108 yr 2 x 107 yr
a Assuming a King approximation to
an isothermal sphere.
b 0.5-3.0 keV
c Assuming no additional heat input.
d The southern half of M86 out to 60 kpc.