Astron. Astrophys. 357, 241-254 (2000)

3. Analysis

The high dispersion spectra have been processed using the standard IRAF routines. The data reduction process includes bias and flat-field corrections, sky subtraction and wavelength calibration. After that, the continuum was normalised and the equivalent widths of the absorption and emission lines were measured using IRAF software. We have used Kurucz (1993) stellar model atmospheres and SYNSPEC (Hubeny 1988) radiative transfer code for calculating the theoretical spectrum. For the emission line analysis of the nebula, we used the NEBULAR package in STSDAS inside IRAF. The gf values and the atomic data for the forbidden lines were taken from Wiese and Martin (1980), Hibbert et al. (1991), Parthasarathy et al. (1992) and Sivarani et al. (1999 and references therein).

3.1. Description of the spectrum

The high resolution spectrum of SAO 85766 (Fig. 1) is found to show absorption lines due to He I, O II, N II, C III, C II, Fe III, Si III, Si II, etc. (Fig. 1). The presence of these lines indicate that the present spectral type of SAO 85766 is B1I. Arkhipova et al. (1999) also classified the recent spectrum of SAO 85766 and found it to be B1-1.5 II in 1996-1997.

Fig. 1a-p. High resolution spectra of SAO 85766 obtained with the 2.5M Issac Newton Telescope (INT) at La Palma

Fig. 1a-p. (continued)

Fig. 1a-p. (continued)

Fig. 1a-p. (continued)

The Balmer lines H and H appear strongly in emission but superposed on the corresponding stellar absorption lines. The Paschen line P12 shows a complex absorption and emission profile. Several permitted and forbidden lines of Fe II are found in emission. Emission lines due to Mg I, Ca I, Ni I, Ni II, Co I, Ti I, Ti II, V II, Cr II, Si II, etc. are detected. Some of the N I, O I and He I lines are found to be in emission (Fig. 1). The nebular emission lines due to [O II], [N II], [S II], and [C I] are also present (Fig. 1), indicating the presence of a low excitation nebula. The [O III] 5007Å line is not detected. The He I 4388Å and 4471Å absorption lines do not show any indication of emission. Arkhipova et al. (1999) found the He I 5015 Å line to be in emission. In our spectra, it appears to be a weak and narrow absorption line although it might be affected by emission in the red wing. This could be due to blending by the nearby Fe II emission lines or may be due to a P-Cygni type profile. The 7065Å line is in emission. Our spectra do not cover the HeI 5876Å region. Arkhipova et al. (1999) found the He I 5876Å line also to be in emission. The full list of absorption and emission lines detected in our spectra are listed in Tables 1 and 2 respectively.

Table 1. Absorption lines in the spectrum of SAO 85766

Table 2. Emission lines in the spectrum of SAO 85766
Note
Flux above the normalised local continuum

Table 2. (continued)

Table 2. (continued)

Table 2. (continued)

3.2. Atmospheric parameters and abundances

The number of absorption lines in the spectrum of SAO 85766 are not too many. Therefore, it is rather difficult to determine the atmospheric parameters with high accuracy. We have used the O II, C II, and C III absorption lines (Table 1) to derive the effective temperature T_eff, surface gravity log g and microturbulent velocity . We have used the Kurucz (1993) stellar model atmospheres and SYNSPEC (Hubeny 1988) radiative transfer code for calculating the theoretical spectrum with LTE approximation.

The mean UBV magnitudes of SAO 85766 in 1995 were V= 11.47, B = 11.52, and U = 10.79 (Arkhipova et al. 1999). The spectral energy distribution of SAO 85766 obtained by Arkhipova et al. (1999) in 1995 (Fig. 2) is compared with the spectral energy distribution in 1986 which was obtained by Downes and Keyes (1988). In Fig. 2 we also show the theoretical flux distribution using the model atmosphere with T_eff = 22000K and log g = 3.0. As SAO 85766 is a high galactic latitude star (b = + 19.8), we can estimate an upper limit for the interstellar extinction of E(B-V) = 0.14. However, from the observed B-V and spectral type (B1I as derived from our high resolution spectra), we infer an E(B-V) of 0.2. This indicates that SAO 85766 has significant amount of circumstellar reddening due to the presence of circumstellar dust with far-IR colours similar to planetary nebulae (Manchado et al. 1989). Therefore we have not used the B-V colour to estimate the temperature of the star. The T_eff obtained from the present assigned spectral type was adopted for further analysis of the spectrum.

Fig. 2. The flux calibrated spectra at different epochs by (Downes and Keyes (1988) and Arkhipova et al. (1999)) compared with the theoretical continuum for Teff=22000K, log g=3.0, Vt=15km s-1. The observed spectrum shows higher flux in the red.

We have used 10 O II absorption lines in the spectrum of SAO 85766 (Table 1) and derived the microturbulent velocity =152 km s^-1 and T_eff of 22000 500K. From the C II and C III lines we derived log g=3.00.5. The references for the atomic data and gf values used here can be found in the papers of Brown et al. (1986), McCausland et al. (1992), Conlon et al. (1993) and Hambly et al. (1996) who have analysed the spectra of other high latitude hot post-AGB stars with atmospheric parameters similar to those of SAO 85766.

We could not use the helium lines and Balmer lines of hydrogen to determine the atmospheric parameters because of the filled-in emission. The He I 4388Å and He I 4471Å absorption lines in the spectrum of SAO 85766 appear to be not affected by emission and the analysis of these lines indicates that the He abundance is normal.

The microturbulent velocity of 15 km s^-1 found from the analysis of absorption lines in the spectrum of SAO 85766 is similar to that found in other hot post-AGB stars (McCausland et al. 1992; García-Lario et al. 1997b). It is also consistent with that found in other low gravity early B stars (Underhill and Fahey 1973). However, for SAO 85766 the abundance estimates of metals are insensitive to the value chosen for the microturbulence since the corresponding line strengths are small. We have carried out spectrum synthesis calculations using the above mentioned atmospheric parameters and derived the abundances. The observed and synthetic spectra are shown in Fig. 3. The abundances are given in Table 3. The number of absorption lines used in deriving the abundances are also given in Table 3. The uncertainties in abundances due to uncertainties in T_eff of 1000K, in log g of 0.5 dex and 5 km/sec in microturbulence are estimated to be of the order of 0.2 dex. We could estimate the abundances of C, N, O, Mg, Al, Si, S and Fe, based on the absorption lines.

Table 3. Chemical composition of SAO85766 for T_eff = 22000, logg = 3.0 and V_t = 15km s^-1

Fig. 3a-f. The dotted line corresponds to the observed and the solid line is the theoritical synthetic spectrum for the model Teff=22000K log g=3.0 and Vt=15km s-1

Fig. 3a-f. (continued)

The abundance pattern in the atmosphere of SAO 85766 (Table 3) indicates that it is metal poor ([Fe/H] = -0.6). The significant underabundance of carbon ([C/Fe] = -1.0) is similar to that observed in other high galactic latitude hot post-AGB stars (McCausland et al. 1992). The C/N and O/N ratios suggest that products of CNO equilibrium reactions are visible at the photosphere. The abundances of S and Fe indicate that under abundance of Fe is intrinsic and there is no depletion of refractory elements. The underabundance of metals and high galactic latitude indicates that SAO 85766 is a low mass old disk population II star. It is not a population I B supergiant and also it is not a Luminous Blue Variable (LBV).

The underabundance of carbon and metals, high galactic latitude, detached cold circumstellar dust shell and the coincidence of the atmospheric parameters with the post-AGB tracks of Schönberner (1983, 1993) suggest that SAO 85766 is a low mass star in the post-AGB stage of evolution. The post-AGB mass loss appears to have speeded up the evolution of SAO 85766 into the early stages of a low excitation planetary nebula.

3.3. Analysis of emission lines

SAO 85766 shows numerous emission lines of Fe II and [Fe II]. The method used for the analysis of the emission lines is described in our previous paper on the post-AGB star HD 101584 (Sivarani et al. 1999). From the forbidden lines of [Fe II] we derived T_e=10000K 500K. For T_e=10000K we estmated N_e=2.50 10⁴ 100 cm^-3 from the [S II] lines and for these values of N_e and T_e the nebular abundances of C, N, O, S and Fe were derived (Table 4).

Table 4. Elemental abundance of the nebula around SAO 85766

The elemental abundance pattern of the nebula is quite similar to the photospheric abundances. Carbon and sulfur abundances in the nebula also indicate that it is metal-poor and strongly underabundant in carbon. The oxygen abundance estimated by Arkhipova et al. from the 5577Å line is very high compared to the value derived from the analysis of the 6300Å and 6363Å [O I] and other [O II] lines. There is quite a large uncertainty in the derived N_e and T_e values. The N_e, T_e contours calculated from the [S II] lines and from the [O I] lines do not overlap (see Fig. 4).

Fig. 4. Contour plot of electron temperature and electron density. Each line corresponds to a particular line ratio. The dotted line is for the observed line ratio. The vertical lines are for line ratio are for [S II] 6716/6731. The horizontal lines are for [O I] 6300+6363/5577.

3.4. Radial velocities

The absorption lines give an average radial velocity of 465 km s^-1. The emission lines also show almost the same radial velocity. The radial velocities derived from the emission lines show large scatter, however. There appears to be a weak correlation for emission lines between the optical depth and the radial velocities (Fig. 5). The radial velocity reduces as the optical depth becomes high. That is to say that the velocity increases in the outward direction. But the scatter in the radial velocities from emission lines could be due to pulsation, shock waves or mass motions. Arkhipova et al. (1999) have found SAO 85766 to show irregular rapid light variations with an amplitude up to 0.3 magnitudes in the V filter. The cause for such variations may be due to pulsation of the central star. The present high resolution spectra of SAO 85766 do not show any signature for binarity or symbiotic type. It is most likely a single star. It is not a binary star and it is not a symbiotic star.

Fig. 5. The plot shows a correlation between the heliocentric radial velocities and the optical depth for the [FeII] emission lines.

© European Southern Observatory (ESO) 2000

Online publication: May 3, 2000
helpdesk.link@springer.de