5.8.2. Other models for M87 and other central galaxies
In Sections 5.7 and 5.8.1, a standard model for M87 (and other X-ray emitting central galaxies) has been given, in which there is a cooling flow leading to accretion by M87. The outer parts of the flow are very subsonic, and the gas is nearly hydrostatic and bound either by the massive halo of M87 or by the pressure of surrounding gas. One feature of this standard model is that thermal conduction must be suppressed in order to ensure that the losses due to cooling in the center of the galaxy are balanced by the enthalpy flux of inflowing gas rather than by heat conducted inwards. Two alternative models are now discussed in which conduction is not suppressed.
Takahara and Takahara (1979, 1981) suggested that gas was actually being thermally evaporated and flowing out from M87, rather than flowing inward. They argued that the gas in the evaporative flow was supplied by mass loss from stars in M87, at a rate of about 1 M / yr. This gas is heated by thermal conduction from surrounding, extended hot intracluster gas, in which the galaxy is assumed to be immersed.
One problem with this model is that the range of temperatures in the gas is rather small, and the gas is generally hotter than 2 × 107 K. This model cannot produce the very strong soft X-ray line emission seen in M87 (Canizares et al., 1979, 1982; Lea et al., 1982; Stewart et al., 1984a). It does not provide a very good fit to the X-ray surface brightness in the inner regions of the galaxy, and does not explain the origin of the line emitting filaments in M87 (Section 5.7.3).
Tucker and Rosner (1983) suggested that the gas in M87 (and other central dominant galaxies) is actually hydrostatic (not moving). In their model, the cooling of the gas is balanced by heating from thermal conduction in the outer regions and by heating by the relativistic electrons associated with the radio source in the inner parts. As discussed in Section 5.7.1, the extra heating by relativistic electrons actually results in thermal equilibrium being attained at a lower temperature, since cooling increases rapidly as the temperature is reduced. Such a model is necessarily thermally unstable (Stewart et al., 1984a), and might form a cooling flow due to the growth of thermal fluctuations. This model requires a fairly large rate of heating from relativistic electrons (Scott et al., 1980; Section 5.3.5), and there is considerable uncertainty about such large heating rates. The model does have the advantage that the similarity in the morphology of the X-ray and diffuse radio emission in M87 is explained.
In Section 5.7.2 the suggestion was made that the radio emission of central cluster galaxies was powered by having a small portion of the accreting gas reach the nucleus. In the Tucker and Rosner model, the gas is static. Thus there is no inflow to power the radio source. Tucker and Rosner suggest that these sources might be episodic. Initially, gas cooling in the core of the cluster might produce an accretion flow on the central galaxy, and start the radio source. The radio source would produce large numbers of relativistic electrons, which would heat the gas and turn off the accretion flow. This would stop the radio source, and the heating would decline as the relativistic electrons lost their energy. This would allow the accretion flow to restart, and the process might oscillate in this fashion indefinitely.