Astron. Astrophys. 347, 55-62 (1999)

4. Emission lines and diagnostic diagrams

After reddening and radial velocity corrections, the corresponding stellar population template (Table 1) was subtracted from each spectra of the galaxy sample. This procedure is illustrated in Fig. 4 for a red (E2) and a blue (S5) stellar population cases. This figure shows the observed spectra, the corresponding stellar population templates and the resulting pure emission spectra of the interacting galaxies AM 2238-575A and AM 2322-821B.

Emission line fluxes were measured on the pure emission spectra and were reddening corrected using the expression (12) of McCall, Rybski & Shields (1985), which should take into account any residual underlying stellar absorption, and which is based on the extinction law of Schild (1977). On the other hand, we caution that extragalactic reddening laws for extended objects seem to work differently than interstellar ones deduced from stars (Calzetti, Kinney & Storchi-Bergmann, 1994; Calzetti et al., 1995). However, the Calzetti extragalactic obscuration law follows very closely the law in the visual spectral range, and significant differences arise only in the ultra-violet (Bonatto et al. 1999). The internal extinction for each galaxy was also taken into account assuming an intrinsic ratio of 2.85 which corresponds to the case B Balmer recombination in an optically thin plasma at T = 10⁴ (Osterbrock 1989). We estimate an rms of 0.06 magnitudes for residuals of the system sensitivity function fitting. Nevertheless, errors in the fluxes are about 15% due to inaccuracies in the continuum subtraction and deblending technique. The corresponding emission line ratios for both components of the galaxy pairs are listed in Table 2.

Table 2. Reddening-corrected emission line ratios
Table Notes:
CS - composite spectrum (AGN + H II region);
RH II - pure H II region spectrum;
NE - no emission spectrum.

Usually, diagnostic-diagrams with emission line ratios of easily observed lines are employed to classify the spectra of extragalactic objects according to the main excitation mechanisms. Unfortunately, due to border effects we could not use , which restricted the number of diagnostic-diagrams available for the analysis.

The classical (Veilleux & Osterbrock 1987) diagnostic-diagrams and , are shown in the upper and lower panels of Fig. 5, respectively. These diagrams are divided in three zones corresponding to the different degrees of activity in the nucleus. Nuclear H II regions and starburst galaxies lie to the left of the solid line, while Seyferts and LINERS are located to the right of this line, with the Seyferts presenting larger values of . The solid line in Figs. 5 and 6, which separates H II regions from AGN, has been adapted from Veilleux & Osterbrock (1987). The consistency of this curve as a true boundary between these two types of objects has been checked with the photoionisation models of Rola et al. (1997) for the diagnostic-diagrams used in the present paper. We noticed that their hottest model (corresponding to the extreme H II regions) matches very well the separating line adopted in the present work.

Fig. 5. Top panel: . The long-dashed lines indicate the effect on the line ratios due to the "contamination" of the Seyfert 1 spectrum (J 15.22) with the H II region spectra R1a, R10a and R5a. The H II region contributions (at Å) are indicated. Corresponding models for the LINER AM 2054-433S are shown as empty diamonds. The heavy solid line represents the "classical" boundary between H II regions (left), LINERS (right) and Seyferts (top right). Filled squares are the line ratios measured in our sample galaxies. The heavy-dashed line marks off the locus of metallic H II regions. The location of NGC 1068 is indicated. Bottom panel: same as above for . Typical error bars are shown.

Fig. 6. Same as Fig. 5 for the line ratios . The dot-dashed line indicates the H II region models calculated for effective temperature T=50 000 K, electron density , ionisation parameter in the range , and solar metallicity.

Fig. 5 shows that the line ratios of most of the interacting galaxies fall in the zone of H II region spectra. However, some of these are very near the transition zone between H II regions and LINERs. In fact, in the diagnostic-diagram, which is a good indicator of galaxies with LINER emission line spectrum (Fig. 6), 4 galaxies have line ratios typical of LINERs and 1 is probably a Seyfert 2, which represents 9.2% of the total sample.

Although some line-ratios indicate the presence of an AGN, broad components in emission lines have not been detected in the present spectra, suggesting that the AGN might be a very low-luminosity source, heavily contaminated by the H II region emission.

© European Southern Observatory (ESO) 1999

Online publication: June 18, 1999
helpdesk.link@springer.de