Astron. Astrophys. 337, 253-260 (1998)

2. Magnitudes of the central stars

2.1. Selection of the objects

The PN were pre-selected from their appearance similar to a compact bright stellar object, or to a bright nucleus surrounded by a very faint nebula. Therefore, most of the observed PN may belong to two extreme evolutionary phases: either very young and compact PN, or very old and extended PN.

Table 1 lists the 19 PN from the programs 106 and 143 submitted and accepted by the Hipparcos project. For 7 objects (He 2-138, He 2-438, Hu 2-1, M 2-54, SaSt 2-12, SwSt 1, He 3-1333), the central star with the nebula appears like a very bright nucleus and the observed data concern both the star and the nebula. Large nebulæ of very faint surface brightness surround a bright central star for NGC 1360, A 35, A 36, PHL 932, LoTr5.

Table 1. Sample of planetary nebulae observed by HIPPARCOS.

Table 1 lists the denominations: HIC number, galactic denomination following Acker et al. (1992), usual names of the nebula and of the central star (col. 1-4), astrometric data, with errors, for the epoch 1991.25: parallaxes (col. 5), proper motions cos and (col. 6-7). Most of the objects - which are some of the brightest central stars known - are well studied.

Note that two objects from the programs are not listed here as they do not belong to the PN-class: HIC 114552 is a misclassified PN, and HIC 76881 corresponds to the B8/B9 giant star HD 139636 = CPD- 6854, nearby the faint PN He 2-133 (see Acker et al., 1992).

2.2. HIPPARCOS magnitudes of the nuclei

The HIPPARCOS H magnitudes were measured along a large spectral range (the FWHM of the H response curve extends from 400 to 620 nm), and through an aperture with a diameter of 30 arcsec. For the 19 PN identified by their usual names (col. 1), Table 2 presents the H magnitudes with their uncertainties (col. 4 and 5) and the different photometric quantities used for our determination of the magnitude of the nuclei (col. 7).

Table 2. Determination of the magnitude of the central stars of the planetary nebulae (* see the text). Original fluxes taken from (a) Pottasch and Acker, 1998 and (b) Acker et al., 1991

In a first step, we estimate the total nebular contribution. We determine the total nebular flux by calculating the convolution of the H response curve and of the total nebular spectrum. The H fluxes and the relative lines intensities are taken from the spectrophotometric survey done by Acker and Stenholm, and reported in Acker et al. (1992), excepting some cases indicated in Table 2. The nebular continuum was taken from Tylenda et al. (1991). The uncertainty in the estimation increases with the value of the relative contribution of the nebula.

Then, we apply a factor depending on the ratio of the surface of the nebula and the surface of the 30 arcsec HIPPARCOS aperture. The angular diameter given in col. 2 is taken from Acker et al 1992, from Cahn et al. 1992, and from Kaler (1983) for some large PN. The scaling of the nebular fluxes to the HIPPARCOS aperture produces errors depending (1) on the inhomogeneity of the nebular density, (2) from neglecting external structures and haloes. This geometrical effect is already taken into account in the calibration of the H fluxes (see the discussion in Acker et al., 1991). The resulting total nebular contribution is given in (col. 3).

The second step leads to the determination of the stellar magnitude (col. 6), after correction from the nebular contribution.

Finally, we use the values (col.9) extracted from Acker et al. (1992), and with the relations given in Perrymann (1997) between , and , we determine (col.10) in order to estimate the magnitude for the central stars (col. 7). Given the uncertainties on the various corrections, we cannot define better than 0.03 mag.

2.3. Comparison of the HIPPARCOS magnitudes with values from the literature

We compare the magnitude with the magnitude (col. 8) deduced from the continuum (Tylenda et al, 1991). A good correlation is observed. However, significant differences appear for 3 objects (identified by *): NGC 246; He 3-1333 and He 2-438, which are also detected as variables by HIPPARCOS (9 stars are noted v in col. 5 on Table 2). For all these objects, we can propose an explanation, in terms of binary systems, or unstable stars (pulsating or WR).

• BINARIES: the nucleus of NGC 246 is known as being a visual binary. The O star has a G8-K0 companion with a magnitude of 14.3, the separation of the two stars being 3.8 arcsec. The HIPPARCOS magnitude V_H = 11.84 corresponds to the total magnitude of the system. We calculate a value of 11.96 for the magnitude of the O star, in perfect agreement with the literature magnitude.

  Other nuclei are known spectroscopic binaries, a situation which could explain the variability detected by HIPPARCOS: case of NGC 2346 (central star V 651 Mon), and of A35 (see Sect. 5). A35 and LoTr 5 are very old extended PN, whose central stars are binaries or more complicated systems. The primary is (for these two objects) a cool giant; the IUE satellite has revealed an extremely hot companion, very faint for A35, and with a magnitude of 14.9 for LoTr 5 (Kaler and Feibelman, 1985).

• PULSATING STAR: The central star of He 1-5 is the pulsating star FG Sge. The magnitude V is variable, increasing from 8.9 in 1970 to 9.3 in early 1992, and fading by about 4 mag. in August 1992 over a period of 40 days, due probably to an ejection of obscuring matter; then the luminosity increases in 1993, and fluctuates around V = 12 (see Iijima, 1996). Over the 3 years of observations by HIPPARCOS, the mean value of the ground-magnitude is in agreement with the value V_H = 9.18.

• WC STARS: The nuclei of He 3-1333 and He 2-438 are late [WC]-stars, with strong mass loss. Their photometry over 3 years is provided in Fig. 1, and shows new results.

  The light curve of the [WC11] nucleus of He 3-1333 shows long-term, possibly periodic, variations, with strong (about 0.5 mag) dips lasting several days, possibly related to dust formation episode (comparable to the variation found for the Pop. I star WR 121, by Marchenko et al, 1997). Variations were already reported by Pollaco et al. (1992) from ground-based observations. The star appears variable on time-scales ranging from hours to months, with an amplitude nearly V =1, with associated colour variations. During the period outside the strong variations, the V-magnitude is stabilized around 11.15.

  BD+ 3639 (V1966 Cyg) is the variable [WC9] nucleus of He 2-438 , with inhomegeneous and variable mass loss (Acker et al , 1997). As shown in Fig. 1, the luminosity increases over the 3 years by about 0.1 mag. Such a long-term variation is compatible with the variability already detected, and with the different values of V found in the literature. The mean magnitude V_H=9.92 corresponds to (i) the V-continuum of the star plus (ii) the bright stellar emission lines (CIII and CIV lines around 565-580 nm) which represent about 0.7 mag. The magnitudes given in the literature are fainter, and show a large spread, going from 10.9 in Perek and Kohoutek (1967) and 9.95 in Pottasch (1983), (both magnitudes are not corrected for the nebular continuum nor for the stellar emission lines) to 12.5 in Tylenda et al. (1991), (the magnitude corresponds to the V-continuum of the star, corrected for the nebular continuum and for the stellar emission lines).

Fig. 1. 3-years photometry of 2 variable nuclei of planetary nebulae: He 3-1333 and He 2-438. The error bar is equal to 2m.p.e.

The behaviour of these late-[WC] stars need further photometric observations. Note that two other emission-line stars appear as variables: the [WC8] nucleus of NGC 40 with variables CIII/CIV emission lines, and the wels nucleus of Hu 2-1 (wels means weak emission lines star , classified by Tylenda et al., 1994), considered by Miranda (1995) as a possible mass exchanging binary.

© European Southern Observatory (ESO) 1998

Online publication: August 6, 1998
helpdesk.link@springer.de