SpringerLink
Forum Springer Astron. Astrophys.
Forum Whats New Search Orders


Astron. Astrophys. 349, 88-96 (1999)

Previous Section Next Section Title Page Table of Contents

4. Notes on individual objects

Below, we first give a brief summary of what is known for the individual galaxies (only the detected ones plus NGC 1167) from the literature and then describe the results from our X-ray temporal, spectral and spatial analysis of the individual objects.

A detailed literature search revealed that some of the present galaxies were already very briefly discussed in other/previous samples with different aims. Given the inhomogeneity of the assumptions made and models fit (for details see below), we extent here the spectral analysis of these objects and also perform a spatial and temporal analysis.

4.1. NGC 404

NGC 404 is blueshifted (Stromberg 1925). Baars & Wendker (1976) noted its peculiar radio properties. Optical spectroscopy was performed by, e.g., Burbidge & Burbidge (1965), Keel (1983), and Filippenko & Sargent (1985) and revealed very narrow emission lines; for an image see Sandage (1961). Larkin et al. (1998) obtained NIR spectra and reported the detection of strong [FeII] emission in this and several further LINERs (but not in all of their sample) and suggested X-ray heating to be at work. A molecular gas ring was observed by Wiklind & Henkel (1990). The detection of a UV core with HST was presented by Maoz et al. (1995). Based on the analysis of UV spectra, Maoz et al. (1998) explained the data by the presence of a central star cluster.

A deep HRI observation of NGC 404 is available. The source is detected ([FORMULA]25 source photons) but too weak to allow a more detailed temporal or spatial analysis. Assuming a powerlaw spectral shape as described above we derive a (0.1-2.4 keV) luminosity of [FORMULA] erg s-1, the lowest [FORMULA] among the present objects, and among the lowest so far detected for a LINER. The X-ray emission of NGC 404 is consistent with originating completely from discrete stellar sources given the galaxy's blue luminosity. Using the relation between [FORMULA] and [FORMULA] of Canizares et al. (1987), we predict [FORMULA] erg s-1 in the (0.5-4.5 keV) band which compares to the observed value of [FORMULA] erg s-1, which is below the expectation but consistent within the scatter. (The intrinsic X-ray luminosity could be boosted if there is some excess absorption or the spectral shape is different from the assumed one.) It is interesting to note that Wiklind & Henkel (1990) argue for a much larger distance of NGC 404 than derived from, e.g., the Tully catalog (1988): they suggest 10 Mpc instead of 2.4 Mpc which would correspondingly increase the values of both [FORMULA] and [FORMULA].

4.2. NGC 1167

The galaxy is a well-known radio source (4C +34.09) and has been intensively studied at radio wavelengths (e.g., Long et al. 1966, Condon & Dressel 1978, Bridle & Fomalont 1978, Sanghera et al. 1995). Optical spectra were presented by, e.g., Wills (1967), Wills & Wills (1976), and Gelderman & Whittle (1994). Despite earlier suspicions, Ho et al. (1997; H97 hereafter) did not detect a broad component in H[FORMULA]. An upper limit for the X-ray luminosity derived from Einstein observations, [FORMULA] erg s-1, was reported by Dressel & Wilson (1985; see also Canizares et al. 1987, Fabbiano et al. 1992).

The source is undetected in the PSPC pointing, which might be partly traced back to the large [FORMULA] value in its direction, [FORMULA] cm-2. We estimate an upper limit for the countrate of 0.005 cts/s, from the conservative assumption of a countrate less than that of the weakest detected source in the field of view at similar off-axis angle.

4.3. NGC 2768

No broad H[FORMULA] component was detected by H97. Weak CO emission was found by Wiklind et al. (1995). The source is included in a sample of galaxies by Davis & White (1996) who fit a Raymond-Smith model and find [FORMULA] keV for metal abundances 0.2 [FORMULA] solar, and absorption of [FORMULA] = [FORMULA] cm-2, less than the Galactic value.

The source seems to be slightly variable from the first to the second pointing with a drop in countrate from 0.021[FORMULA]0.002 cts s-1 to 0.013[FORMULA]0.003 cts s-1. The short-term light curve (first pointing) shows constant source flux.

Spectral fits were performed for the deeper PSPC observation only. Neither a single powerlaw with [FORMULA] nor emission from a Raymond-Smith plasma with solar abundances provides a successful X-ray spectral fit. The fit becomes acceptable for very subsolar abundances. In that case we find a lower temperature than Davis & White (1996) (and cold absorption consistent with the Galactic value, which should not be underpredicted). This T also is more consistent with the [FORMULA] relation of Davis & White.

A comparison of the source's spatial extent with the theoretical point spread function (PSF) of a point source shows that most of the X-ray emission is consistent with arising from a point source (Fig. 3). There may be some extended emission at weak levels (Fig. 2). A nearby weak second source is detected with a countrate of 0.003[FORMULA]0.001 cts/s. It coincides with a stellar objects on a POSS plate.

[FIGURE] Fig. 2. Contour plots for the X-ray emission of the three brightest on-axis PSPC sources overlaid on optical images from the digitized POSS. NGC 2768: contours are shown for 0.8, 1.1, 1.4, 2.1, 2.5, 3.1[FORMULA] above the background; NGC 3642: contours are 1.2, 1.8, 2.5, 3.9, 4.7, 5.8, 7.4[FORMULA]; NGC 4450: contours are 1.2, 1.5, 1.9, 2.8, 4.3, 8.2, 15.9, 31[FORMULA] above the background.

[FIGURE] Fig. 3. Azimuthally averaged radial profiles of the X-ray emission from the three brightest on-axis PSPC sources compared to the PSPC's point spread function for a point source (solid line: at 1 keV, dotted line: at 0.5 keV). The bulk of the X-ray emission is consistent with arising from a point source.

4.4. NGC 3642

Some X-ray properties of this LINER were earlier examined by K95 who fit a powerlaw model to the ROSAT PSPC spectrum and found the HRI source extent to be consistent with a point source. A broad component in H[FORMULA] is present (e.g., K95). Barth et al. (1998) reported the detection of a compact nuclear UV source based on HST data, and conclude that the extrapolation of the UV continuum, assuming an AGN-like shape, would provide enough ionizing photons to power the NLR emission of this galaxy.

We neither detect short-time variability nor variability between the two PSPC observations separated by 5 months.

The spectrum is best fit by a Raymond-Smith model of heavily depleted abundances, around 0.03 [FORMULA] solar (or, alternatively, by a powerlaw model with some excess absorption, confirming K95).

A comparison with the PSF of the PSPC shows that most of the X-ray emission arises from an unresolved source (Fig. 3). There is a second source nearby with a countrate of 0.0025 [FORMULA]0.0007 cts/s. Its position falls close to two star-like knots projected onto (or in) one of the spiral arms of NGC 3642 (they could be HII regions or foreground stars). If the X-ray source is intrinsic to NGC 3642, its luminosity of [FORMULA] erg s-1 assuming a powerlaw spectral shape as described in Sect. 3 is fairly high. For instance, its exceeds the Eddington luminosity of a solar mass black hole by a factor [FORMULA]10. One possible interpretation is a powerful X-ray binary with either a super-eddington low-mass black hole or a massive black hole. We note in passing that no optical supernova was detected in NGC 3642.

4.5. NGC 3898

There have been several studies of this galaxy in the optical (e.g., Burbidge & Burbidge 1965, Barbon et al. 1978, Mizuno & Hamajima 1986, and references given in van Driel & van Woerden 1994). H97 tentatively concluded that broad H[FORMULA] is absent from the optical spectrum. 21 cm HI observations with the WRST were presented by van Driel & van Woerden (1994).

The source is quite weak with only about 50 detected photons, very close to the limits of meaningful [FORMULA] spectral fits. Therefore, we only applied a powerlaw model with fixed Galactic absorption. This results in [FORMULA] =-2.1 and gives an acceptable fit.

4.6. NGC 4450

A fairly weak broad H[FORMULA] line is probably present in the optical spectrum (Stauffer 1982, H97). For an optical image see, e.g., Sandage (1961). The HII region population of the galaxy was studied by Gonzalez Delgado et al. (1997). An Einstein IPC image is shown in Fabbiano et al. (1992). They derive an (0.2-4 keV) X-ray flux [FORMULA] erg cm- 2 s-1 under the assumption of a thermal bremsstrahlung spectrum with kT=5 keV.

Again, we do not detect short-timescale variability.

The source is quite bright and nearly 2000 photons are available for the spectral analysis (we used the deepest pointing). No Raymond-Smith fit is possible. When allowing [FORMULA] to be free, it underpredicts the Galactic value. If subsolar abundances are allowed, the best fit requires abundances less than 1/100 solar and that fit is still unsatisfactory. In contrast, a single powerlaw with [FORMULA] = -2.0 near the AGN-canonical value (e.g., Pounds et al. 1994; Svensson 1994) gives an excellent fit. If [FORMULA] is treated as free parameter, the Galactic value is recovered. We derive a soft X-ray luminosity of [FORMULA] erg s-1, the highest value found among the present galaxies. The corresponding (0.5-4.5 keV) X-ray luminosity is [FORMULA] erg s-1, a factor [FORMULA]5 above the value expected from the stellar contribution using the relation of Canizares et al. (1987). We note that Tully's catalog places NGC 4450 at the distance of the Virgo cluster. If the galaxy is instead located in the sheet of galaxies behind the Virgo cluster, the luminosities inferred above increase correspondingly.

A comparison with the PSF of the PSPC shows that most of the X-ray emission is consistent with arising from a point source. At weak emission levels there is evidence for source extent (Fig. 3; several of the structures are seen in both, the soft (0.1-0.5 keV) and hard (0.5-2.4 keV) band). Again, there is a nearby second source. Its countrate is 0.011 [FORMULA]0.002 cts/s and since the pointing is deep, a spectral analysis is possible. A powerlaw spectral fit gives a spectrum similar to NGC 4450 itself, but a bit softer with [FORMULA] =-2.4. At the distance of NGC 4450 this corresponds to a luminosity of [FORMULA] erg s-1. The source is also present in the second PSPC pointing of slightly lower exposure time (Table 1). Its countrate is constant. Inspection of the POSS plates does not reveal any optical counterpart. Neither is there any X-ray source visible in the Einstein IPC image (see Fig. 7 of Fabbiano et al. 1992). The `reality' of these nearby sources is examined in Sect. 4.9.

4.7. NGC 5371

The galaxy was classified as a LINER by Rush et al. (1993). Elfhag et al. (1996) presented CO measurements and suggested NGC 5371 to be a good candidate for a post-starburst galaxy (Koorneef 1993). The rotation curve was measured by, e.g., Zasov & Sil'chenko (1987). Gonzalez Delgado et al. (1997) studied the HII region population.

The X-ray lightcurve does not show short-timescale variability (the countrate in individual bins falls slightly outside the 1[FORMULA] error, occasionally, but this is most likely due to the closeness of the source to the PSPC support grid structure).

The source appears to be extended or double. Thus, photons from the total emission region were first extracted for analysis. A spectrum was fit to this `double' source (since their contributions cannot be disentangled from each other safely; it is these results that are listed in Table 2). In that case, a powerlaw of index [FORMULA] [FORMULA] yields a successful X-ray spectral fit. The cold absorption, if treated as free parameter, underpredicts the Galactic value, and even more so if a RS model is applied. The latter type of model only provides an acceptable fit if the metal abundances are depleted below 0.01 [FORMULA] solar.


[TABLE]

Table 2. Results of spectral fits; first columns: powerlaw with [FORMULA] (for results with [FORMULA] as free parameter see text), last columns Raymond-Smith model. Fluxes (absorption corrected) and luminosities refer to the (0.1-2.4 keV) energy band. For the sources that did not allow a PSPC spectral analysis, we assumed a powerlaw of photon index [FORMULA] =-1.9 plus an amount of cold absorption corresponding to the Galactic value in direction of the individual galaxies (Dickey & Lockman 1990) to calculate fluxes and luminosities (if RASS and pointed observations were available, we used the observation with the longest exposure time to derive f and L). For some of the sources only observed during the RASS, only upper limits on count rates are available. The errors in [FORMULA] and kT, given for successful spectral fits only, are quoted at the 68.3% confidence level. Metal abundances reported in the last but one column were fixed.
Notes:
1) fixed


Secondly, since the optical position of NGC 5371 falls on the northernmost of the two sources, source photons centered on the optical position of the galaxy were extracted within a circular region of diameter 250". In this case, the X-ray spectrum is dominated by the northernmost source, but the second one contributes to some extent. The spectral analysis then yields a best fit in terms of a powerlaw with [FORMULA] [FORMULA] ([FORMULA]=0.5), and [FORMULA] recovers the Galactic value if treated as free parameter. Again, RS emission can only successfully describe the spectrum for heavily depleted abundances.

The source appears double, or extended. No HRI observation is available for a more detailed study of the spatial extent.

4.8. NGC 6500

There are several lines of evidence for the presence of a nuclear outflow or wind as judged from radio continuum emission measurements (Unger et al. 1989) and optical emission lines (Gonzalez Delgado & Perez 1996). H97 did not detect broad H[FORMULA]. Barth et al. (1997) using HST data concluded that the resolved UV emission of NGC 6500 is likely dominated by massive stars. They also derived a ROSAT HRI flux for this galaxy, assuming a spectrum with [FORMULA] =-2.

Although NGC 6500 is detected in the HRI observation, the low number of source photons prevents a more detailed analysis in terms of source extent or variability.

4.9. Origin of nearby sources

In PSPC observations of several galaxies (NGC 2768, NGC 3642, NGC 4450) there is a second source detected near the target source with a countrate always roughly 1/10 of the central source. The same was found for NGC 4736 by Cui et al. (1997) who considered the second source to be real and of transient nature due to its presence in the PSPC and absence in the HRI observation. Given the similar locations relative to the central source, and same (factor [FORMULA] 1/10) relative countrates, we suspected these second sources to be an instrumental artifact, namely ghost imaging (Briel et al. 1994) to be at work.

To more closely examine this problem we selected the second source near NGC 4450 since it is the brightest, thus allowing the most detailed analysis, and since no potential optical counterpart shows up near the X-ray position. We extracted the photons around the X-ray center of the source and made several tests. However, we find no indications of ghost imaging: Source photons do not exclusively cover the very soft channels (ghost imaging only operates below [FORMULA] 0.2 keV; Nousek & Lesser 1993, Briel et al. 1994), and the source is not fixed in detector coordinates but follows the wobble.

Using the X-ray [FORMULA] distribution of Hasinger et al. (1994), we expect only 0.17 sources of X-ray flux greater or equal to that of the source near NGC 4450 in a region of size 10´ [FORMULA] 10´. For a discussion of an excess of bright X-ray sources around nearby galaxies see Arp (1997, and references therein).

Unfortunately, no HRI observation of NGC 4450 is available for further scrutiny. It will certainly be interesting to check for the presence of the second source once further high spatial resolution observations of NGC 4450 become available.

Previous Section Next Section Title Page Table of Contents

© European Southern Observatory (ESO) 1999

Online publication: August 25, 1999
helpdesk.link@springer.de