5.1. X-ray luminosity and / ratio
The objects examined span a luminosity range from erg s-1 in the (0.1-2.4 keV) band. There is still some bias towards selecting the high LINERs. This does not hold for the ROSAT survey data but given the short exposure times of typically 400 s upper limits, although already meaningful, are not very restrictive concerning the low-luminosity end.
None exceeds the limit of erg s-1 which is usually taken as indicative for the presence of a `normal' AGN (e.g., Wisotzki & Bade 1997). Also, none reaches the high usually observed for ellipticals in the group/cluster environment (e.g., Brown & Bregman 1998, Irwin & Sarazin 1998, Beuing et al. 1999) which occasionally show LINER-like emission lines.
Most of the present objects fall in the intermediate / range (Fig. 4). Since the majority of LINERs are found in bulge-dominated early-type galaxies (e.g., Ho 1998) the same emission mechanisms might contribute to the observed X-ray luminosity. Among the suggestions for early-type galaxies are accumulated stellar winds, SN heating and cooling flows (see Pellegrini 1999 for a recent overview).
Further clues on the emission mechanism can be drawn from the observed spectral shapes.
5.2. X-ray spectral shapes
The LINERs analyzed here show some spectral variety. Some of them are best described by a powerlaw model of photon index similar to that observed in AGN (e.g., Schartel et al. 1996a,b, 1997), the others are best fit by Raymond-Smith (RS) emission of a very sub-solar abundances plasma (see Table 2).
We consider the RS model with heavily depleted gas-phase metal abundances to be quite unrealistic, and as previously found for other galaxies (ellipticals, AGN; e.g., Matsushita et al. 1997, Buote & Fabian 1998, Komossa & Schulz 1998) we prefer the alternative of a two-component spectrum, consisting of contributions from both, a powerlaw (or other hard component) and thermal RS emission of solar abundances gas. Such models were not fit, though, due to the quite low number of photons available per spectrum. Other possibilities (than the two-component solution) include a contribution from Fe L emission, not yet fully modeled (but see Buote & Fabian 1998 who excluded the FeL region from fitting ASCA spectra of early type galaxies, and still find similar results concerning abundances), a multi-temperature distribution of the emitting medium (e.g., Strickland & Stevens 1998) and the possibility that the X-ray emitting gas is far out of collisional-ionization equilibrium (e.g., Breitschwerdt & Schmutzler 1994, 1999, Komossa et al. 1999; see Böhringer 1998 for a recent review).
In case of the two-component interpretation - as indeed observed by ASCA with its broader X-ray energy range for several early-type galaxies (e.g., Matsushita et al. 1994, Matsumoto et al. 1997, Buote & Fabian 1998) and some LINERs (e.g., Ptak et al. 1999) - the hard component could be due to stellar sources, namely LMXBs, or a low-luminosity AGN, the soft component due to emission from hot gas (see previous Section).
In the case of NGC 6500 the X-ray emission might be related to the outflow/wind for which optical and radio evidence was reported (Unger et al. 1989, Gonzales Delgado & Perez 1996), in analogy to the X-ray emission associated with starburst-driven winds observed in several starburst galaxies (e.g., Heckman et al. 1990, Schulz et al. 1998). It is also interesting to note that in two further cases, NGC 4450 and NGC 2768, the weak extended emission appears roughly perpendicular to the galaxy's disk and could result from an outflow.
We interpret the powerlaw spectral component, in those cases where it dominates the spectrum, as arising most likely from a low-luminosity active nucleus, since the inferred luminosities are above those expected from discrete stellar sources and the powerlaw indices derived are in the range to -2.1 (Table 2) similiar to what is observed for AGNs. However, we cannot exclude a more complex situation where the superposition of several different emission components mimics a single AGN-like powerlaw.
For a detailed spatial and spectroscopic disentanglement of the contributing components (thermal RS-like emission, presumably extended, and a pointlike powerlaw-like component) and a better determination of the metal abundances high-spectral resolution observations with future X-ray observatories like XMM, AXAF or Spectrum-X-Gamma will be very useful.
5.3. X-ray variability
We do not find evidence for X-ray variability on the timescale of hours/days. This also holds for the LINERs examined by Ptak et al. (1998) and some further sources, and is not in line with a continuation of the trend of higher variability at lower luminosity seen in AGN. It is consistent with the presence of advection-dominated accretion disks (e.g., Abramowicz et al. 1988, Narayan & Yi 1994) in LINERs as suggested by Ptak et al. (1998; see also Lasota et al. 1996).
© European Southern Observatory (ESO) 1999
Online publication: August 25, 1999