Forum Springer Astron. Astrophys.
Forum Whats New Search Orders

Astron. Astrophys. 355, 885-890 (2000)

Previous Section Next Section Title Page Table of Contents

3. Results and discussion

3.1. General FIR properties

The far-infrared continuum at each position is modelled with a single-temperature blackbody spectrum of the form [FORMULA] [FORMULA] [FORMULA], where the solid angle, [FORMULA], is constrained to equal the LWS beam, [FORMULA](T) is the Planck function at temperature T and [FORMULA] [FORMULA] [FORMULA]. The result for the central position is shown as the dashed curve in Fig. 2. Although the observed continuum is not a simple function of wavelength and the single temperature blackbody is not an especially good fit, particularly at wavelengths [FORMULA]m, a better calibration of straylight and the beam profile is required for anything more sophisticated. The best FIR temperature at each position is [FORMULA] 30 K.

The luminosities quoted here are derived from the line fluxes listed in Table 2. At the central position, the total luminosity in all of the far-infrared lines is 2.6 [FORMULA] [FORMULA], which is [FORMULA] 1% of the total FIR luminosity ([FORMULA] [FORMULA]). The [CII] luminosity is 1.1 [FORMULA] [FORMULA] (0.4% FIR) and the [OI] luminosity is 7.5 [FORMULA] [FORMULA] (0.2% FIR). Because full spectra are not available at the NW and SE positions we estimate the FIR continuum luminosity by integrating under the single-temperature blackbody fit to the data. At the NW position the total FIR luminosity, [FORMULA] [FORMULA]. The [CII] luminosity is 9.3 [FORMULA] [FORMULA] (0.4% FIR) and the [OI] luminosity is 3.4 [FORMULA] [FORMULA] (0.2% FIR). At the SE position the total FIR luminosity, [FORMULA] [FORMULA]. The [CII] luminosity is 3.5 [FORMULA] [FORMULA] (0.4% FIR) and the [OI] luminosity is 2.0 [FORMULA] [FORMULA] (0.2% FIR). These are typical values for starburst galaxies (c.f. Lord et al. 1996).


Table 2. Line fluxes.
Flux [FORMULA] W cm-2
- wavelength range not covered
upper limits are 3 x rms residuals from a fit to the continuum
[CII] 157 µm flux at #4 (off) is 0.04

3.2. Ionized gas lines

Photons of energy 35.12, 29.60 and 14.53 eV are required to form O++, N++ and N+, respectively, so that the observed [OIII], [NIII] and [NII ] emission must originate in or around HII regions. The [OIII] line ratio is a sensitive function of density in the range [FORMULA] cm-3. For the central position this ratio is [FORMULA] 0.9, corresponding to an electron density, [FORMULA] [FORMULA] 100 cm-3 (Rubin et al. 1994). The [OIII] lines indicate a higher electron density, [FORMULA] [FORMULA] 250 cm-3, for the starburst nuclei of M82 (Colbert et al. 1999) and M83 (Stacey et al. 1999). In contrast, the [NII] 205 µm / 122 µm line intensity ratio for the Galaxy gives an average electron density, of only [FORMULA] 3 cm-3 (Petuchowski & Bennett 1993). The thermal pressure of the ionized material in the Cen A dust lane is therefore closer to that of starburst galaxies than to that of the Milky Way.

Since the [NIII] 57 µm and the [NII] 122 µm lines arise from different ionization states of the same element, the line intensity ratio is sensitive to the hardness of the interstellar UV field and therefore to the spectral type of the hottest main sequence star. For the central position [NIII] / [NII] [FORMULA] 1.6. This is larger than the value of [FORMULA] 0.9 for M83 (Stacey et al. 1999) but smaller than the value of [FORMULA] 2.1 for M82 (Colbert et al. 1999). Assuming that the region is ionization bounded, with an electron density, ne [FORMULA] 100 cm-3 the [NIII] / [NII] line intensity ratio for Cen A corresponds to an abundance ratio N++/N+ of [FORMULA] 0.3; this corresponds to an effective temperature, [FORMULA] K (Rubin et al. 1994). Applying the same corrections to M82 and M83 with ne [FORMULA] 250 cm-3 implies an effective temperature, [FORMULA] [FORMULA] 34 500 K for M83 and [FORMULA] [FORMULA] 35 500 K for M82. If the effective temperature in Cen A corresponds to the tip of the main sequence formed in a single starburst, we are observing O8.5 stars and the burst is [FORMULA] 6 [FORMULA] years old. If the burst was triggered by the spiral-elliptical galaxy merger then its occurance was very recent. Alternatively, the merger triggered a series of bursts of star formation, of which we are witnessing the most recent.

The N++ and O++ coexist in roughly the same ionization zones, and the [OIII] 52 µm and [NIII ] 57 µm lines have roughly the same critical density. As a result the ratio of these lines is an indicator, to within [FORMULA]%, of the N++/O++ abundance ratio, which itself, is nearly equal to the N/O ratio in the hard UV field environments we are seeing here (Rubin et al. 1988). The line ratio we observe at the centre of the dust lane is [FORMULA] 0.3 - the same as found in the nucleus of M82 (Colbert et al. 1999), but much smaller than that found for the nucleus of M83 ([FORMULA] 0.67 Stacey et al. 1999).

A more precise determination of the abundance ratio requires the observed line ratio to be divided by the volume emissivity ratio. The latter ratio is dependent on the electron density because the two lines have slightly different critical densities. Using our value for the electron density [FORMULA] 100 cm-3 and Fig. 3 of Lester et al. (1987) we estimate that the N/O abundance ratio to be [FORMULA] 0.2 in Cen A. This value is consistent with the range of [FORMULA] 0.2 - 0.3 found for Galactic HII regions (Rubin et al. 1988). The nitrogen to oxygen abundance ratio is a measure of the chemical evolution and we expect it to increase with time (cf. the solar value of [FORMULA] 0.12).

3.3. Neutral gas lines

Carbon has a low ionization potential (11.4 eV), which is less than that of hydrogen. [CII] 157 µm line emission is therefore observed from both neutral and ionized hydrogen clouds. We model the [CII] line emission with three components: Photodissociation regions (PDRs) on the surfaces of UV exposed molecular clouds; cold (T [FORMULA] 100 K) HI clouds (i.e. the cold neutral medium (CNM) Kulkarni & Heiles 1987, Wolfire et al. 1995); and diffuse HII regions (i.e. the warm ionized medium (WIM) Heiles 1994).

3.3.1. HI clouds

It can be shown that the intensity in the [CII] line emitted from gas clouds with density, n(H) and temperature (T) is given by (c.f. Madden et al. 1993)


where the critical density for collisions with H, [FORMULA] cm-3 (Launay & Roueff 1977) and the fractional C+ relative to hydrogen is [FORMULA] (Sofia et al. 1997).

N(HI) is estimated from the HI 21cm map of van Gorkom et al. (1990) to be 18.8 [FORMULA] atom cm-2 at the SE position. The central and NW positions are difficult to estimate due to HI absorption against the nuclear continuum. There may be a central hole in the HI and the column density is certainly not higher than the peak observed in the SE region of the dust lane (Van Gorkom et al. 1990)

Assuming typical Galactic values for the temperature, T [FORMULA] 80 K, and hydrogen density, n [FORMULA] 30 cm-3, results in an estimated [CII] flux of [FORMULA] W cm-2 in a 70 " LWS beam at the SE position. This corresponds to 4%, 1% and 1% of the observed [CII] flux at positions SE, Centre and NW respectively. The peak HI emission line flux corresponds to [FORMULA] W cm-2 which is only 6% and 8% of the [CII] flux at the centre and NW positions respectively. We conclude that there is very little [CII] emission in our beams from HI clouds.

3.3.2. Diffuse HII regions

Ionized carbon can be found in both neutral gas and ionized gas clouds, and is an important coolant for each. We detected [OIII] 88 µm in all 3 beam positions so there is an ionized gas component in each beam. Using the constant density HII region model of Rubin (1985) with the Kurucz abundances, 1049 ionizing photons per second and our derived density, ne [FORMULA] 100 cm-3 and effective temperature, [FORMULA] 35 500 K we can estimate the [CII] emission from the HII regions. Applying the model [OIII] 88 µm / [CII ] 158 µm line ratio of 0.35 to the observed [OIII] 88 µm line flux at each position results in [FORMULA] 10% contribution to the observed [CII] line flux in each beam. Scaling the model fluxes to the distance of Cen A gives [FORMULA] 3000 HII regions in the central and NW regions and [FORMULA] 1000 HII regions in the SE region.

The estimate above assumes that the observed lines have the same filling factor in the large LWS beam. If, alternatively, we were to assume that the ionized component was instead dominated by a contribution from an extended low density warm ionized medium (ELDWIM) with ne [FORMULA] 3 cm-3, then the [CII] flux can be estimated from the ratio of the [CII]/[NII] lines to be [FORMULA] 18% at the central position. The observations of the [NII] 121.9 µm line at the NW and SE positions (with lower signal to noise) indicate a similar fractional component (21% and [FORMULA]%, respectively).

We have estimated the density in the HII regions in the centre of Cen A to be [FORMULA] 100 cm-3 with an effective temperature, [FORMULA] [FORMULA] 35 500 K. Based on the HII region models of Rubin (1985) we estimate that [FORMULA] 10% of the observed [CII] arises in the WIM.

3.3.3. PDRs

Far-UV photons (6 eV [FORMULA] h[FORMULA] 13.6 eV) from either O/B stars or an AGN will photo-dissociate H2 and CO molecules and photo-ionize elements with ionization potentials less than the Lyman limit (e.g. C+ ionization potential = 11.26 eV). The gas heating in these photodissociation regions (PDRs) is dominated by electrons ejected from grains due to the photoelectric effect. Gas cooling is dominated by the emission of [OI ] 63 µm and [CII ] 158 µm emission. Observations of these lines, the [OI] 146 µm and CO (J=1-0) 2.6 mm lines and the FIR continuum can be used to model the average physical properties of the neutral interstellar medium (Wolfire et al. 1990). Kaufman et al. (1999) have computed PDR models over a wide range of physical conditions. The new code accounts for gas heating by small grains/PAHs and large molecules, and uses a lower, gas phase carbon abundance ([FORMULA] = 1.4x10-4, Sofia et al. 1997) and oxygen abundance ([FORMULA] = 3.0x10-4, Meyer et al. 1998). The [OI] 63 µm / [CII] 158 µm line ratio and either the [OI] 146 µm / [OI] 63 µm line ratio or the ([OI] 63 µm + [CII] 158 µm) / FIR continuum can be used as PDR diagnostics to determine the average gas density (i.e the proton density, n cm-3), the average incident far-UV flux (in units of the Milky Way flux, Go = 1.6 [FORMULA] erg cm-2 s-1) and the gas temperature.

We assume that the measured [CII] flux at each position should have [FORMULA] 10% subtracted, due to the HI and WIM components, before it is used to model the PDRs (if, alternatively, a 20% ELDWIM contribution is subtracted it would not significantly affect the PDR parameters derived below). The PDR lines are plotted in Fig. 3 and the line intensity ratios are given in Table 3.

[FIGURE] Fig. 3. PDR Lines (Flux Density [FORMULA] W cm-2 µm-1). Note the [CII] lines are the total observed flux density i.e. PDR + CNM + WIM


Table 3. PDR diagnostic line intensity ratios

The results for the three regions are consistent with each other, having a gas density, n [FORMULA] 103 cm-3, and an incident far-UV field, G [FORMULA] 102.

At the NW position, only the combination of the [OI ] 63 µm / [CII] 158 µm ratio and the ([OI] 63 µm + [CII ] 158 µm) /FIR continuum ratio gives a meaningful solution for G and n. The [OI] 146 µm line is clearly detected but with a very rippled baseline due to channel fringes. The observed [OI] 146 µm line flux would need to be reduced by [FORMULA] 60% in order to obtain a consistent result with the [OI ] 146 µm / [OI] 63 µm line ratio predicted by the PDR model.

The LWS results for the nucleus confirm those previously derived from IR, submm and CO observations. The consistent set of derived PDR conditions for all three positions suggest that the observed FIR emission in a 70 " beam centred on the nucleus is dominated by star formation and not AGN activity. Joy et al. (1988) mapped Cen A at 50 and 100 µm on the KAO. They concluded that the extended FIR emission was from dust grains heated by massive young stars distributed throughout the dust lane, not the compact nucleus. Hawarden et al. (1993) mapped Cen A at 800 µm and 450 µm with a resolution of [FORMULA]10 ". They attribute the large scale 800 µm emission to thermal emission from regions of star formation embedded in the dust lane. They note that the H2 emission within a few arcseconds of the nucleus, observed by Israel et al. (1990), indicates that significant UV radiation from the nucleus does not reach large radii in the plane of the dust lane i.e. the nuclear contribution to exciting the extended gas and dust disk is small.

Eckart et al. (1990) and Wild et al. (1997) mapped Cen A in 12CO J=1-0, 12CO J=2-1 and 13CO J=1-0. All three maps have two peaks separated by [FORMULA] 90 " centred on the nucleus. It is interesting to note that our SE position only clips the lowest contours of the CO (1-0) and CO (2-1) maps of Wild et al. (1997). In spite of this the derived PDR parameters are consistent with those encompassing the bulk of the molecular emission. There must be extended low level CO (1-0) emission beyond the sensitivity limits of the Wild et al. (1997) maps. The lowest contour is 17.5 K kms-1, corresponding to [FORMULA] [FORMULA] 108 [FORMULA] if the material filled the LWS beam.

Previous Section Next Section Title Page Table of Contents

© European Southern Observatory (ESO) 2000

Online publication: March 21, 2000