Forum Springer Astron. Astrophys.
Forum Whats New Search Orders

Astron. Astrophys. 323, 387-392 (1997)

Previous Section Next Section Title Page Table of Contents

4. The absence of mass-loss from 1978 onward

4.1. IUE spectra

IUE began its observations in 1978, and RR Tel was a favourite target among symbiotics. We concentrate on the three multiplets N V [FORMULA] 1240, C IV [FORMULA] 1550, and He II [FORMULA] 1640. P Cygni profiles with absorption in the continuum of the hot star would be the most direct proof for a stellar wind from the hot star.

The resonance doublet N V [FORMULA] 1240 is ideal for wind diagnostics, and it is prominent in the spectrum, e.g. Fig. 1. Penston et al. (1983) measured its FWHM as 57 km/s. The doublet lies practically at rest wavelengths. The dynamical range of IUE did not allow to observe P Cygni profiles of very strong emission lines on top of a weak continuum. However, wind lines, hidden by nebular emission might still betray their presence by a wide foot. For an illustration see the example of AG Peg given in Vogel & Nussbaumer (1994) and Nussbaumer et al. (1995). An investigation of a series of IUE spectra from 1978 to 1992 confirms that the N V [FORMULA] 1240 doublet in RR Tel is typical of nebular emission as generally observed in symbiotic systems, but it shows no contribution from a fast stellar wind.

[FIGURE] Fig. 1. N V [FORMULA] 1238.8, 1242.8 observed by IUE on November 21, 1978 with an exposure time of 450 seconds. The position of a reseau mark is indicated at [FORMULA]. Fluxes are in [FORMULA].

C IV [FORMULA] 1548.2, 1550.8 and He II [FORMULA] 1640 are well exposed on many IUE spectra. They present qualitatively the same picture as N V [FORMULA]. They show typical nebular profiles with no broad feet or P Cygni features. There was no detectable qualitative change in the line profiles from 1978 to 1992.

4.2. HST spectra: Search for wind lines of N V, C IV, and He II

HST has a much higher sensitivity and dynamic range than IUE. It is possible to obtain simultaneously high quality line profiles and a good signal/noise in the underlying continuum, see also Harper et al.(1995). In Figs. 2 and 3 we show the strong resonance doublets N V [FORMULA] 1240 and C IV [FORMULA] 1550. Both stand on a wide foot to be discussed below. None of them shows any sign of a P Cygni profile. The increase shortward of [FORMULA] is due to the wing of Ly [FORMULA], as is evident from a low resolution HST spectrum.

[FIGURE] Fig. 2. N V [FORMULA] 1238.8, 1242.8 taken on July 16, 1995 with HST HRS grating G160M, exposure time 762 s. Fluxes are in [FORMULA]. For the upper spectrum the scale has to be reduced by 200.

[FIGURE] Fig. 3. C IV [FORMULA] 1548.2, 1550.8 taken on July 16, 1995 with HST HRS grating G140M, exposure time 381 s. Fluxes are in [FORMULA]. For the upper spectrum the scale has to be reduced by 400.

In Table 1 we give the FWHM expressed as velocities for the important lines in Figs. 2 - 6. At their bottom the lines reach about twice the FWHM-values. As is often seen in symbiotics, the widths increase with the degree of ionization, note that the recombination O V [FORMULA] 1643.7 is emitted in the O [FORMULA] region. None of these velocities is a candidate for a wind line from a hot white dwarf.


Table 1. Emission observed by IUE (1978, 1992) and HST (1995). Fluxes are in [FORMULA] for lines and [FORMULA] for the continuum; [FORMULA] stands for [FORMULA]. The last column gives FWHM in km/s from the July 16, 1995 HST spectrum.

[FIGURE] Fig. 4. He II [FORMULA] 1640 taken on July 16, 1995 with HST GHRS grating G160M, exposure time 1088 s. Also seen are O I [FORMULA] 1641.3 and O V [FORMULA] 1643.7. The flux is given in [FORMULA]. For the upper spectrum the scale has to be reduced by 40.
[FIGURE] Fig. 5. N IV [FORMULA] 1483.3, 1486.5 taken on July 16, 1995 with HST HRS grating G160M, exposure time 979 s. Fluxes are in [FORMULA]. For the upper spectrum the scale has to be reduced by 40.
[FIGURE] Fig. 6. Si III [FORMULA] 1882.7, 1892.0 and C III [FORMULA] 1906.7, 1908.7 taken on July 16, 1995 with HST GHRS grating G200M, exposure time 1523 s. Also seen is Fe II [FORMULA] 1881.2, 1884.1. Fluxes are in [FORMULA]. For the upper spectrum the scale has to be reduced by 100.

Table 1 shows the evolution of the nebular lines since 1978. There is no significant change in the flux ratios of lowly to highly ionized lines from 1978 to 1995.

The wide foot underneath the nebular emission of N V [FORMULA] 1240, C IV [FORMULA] 1550, and He II [FORMULA] 1640 could be the signature of a fast wind. However, it is much more likely that the foot is due to electron scattering of line photons on free electrons. In a nebular gas of [FORMULA] K the Doppler width of electron scattering at [FORMULA] is [FORMULA]. The flux in the wide scattering foot of N V [FORMULA] 1240 is [FORMULA]. Calculations done for us by Dr.W. Schmutz confirm that this flux is compatible with model expectations.

4.3. HST spectra: Search for a collision zone with intercombination and forbidden lines

If the hot star emits a fast wind, we expect a shock zone of high temperature where the hot wind collides with the wind from the Mira. Nussbaumer et al. (1995) have interpreted the profile of the forbidden N IV [FORMULA] transition seen in AG Peg as evidence for a collision zone. Fig. 5 shows the N IV intercombination doublet. There are four emission features centered at [FORMULA]. For the N IV [FORMULA] flux ratio of [FORMULA] Nussbaumer & Schild (1981) give an electron density of [FORMULA]. The profile of the N IV intercombination doublet is quite different from that of AG Peg, and there is no hint of a wind-wind collision zone.

Two further intercombination multiplets, Si III [FORMULA] 1882.7, 1892.0 and C III [FORMULA] 1906.7, 1908.7 are shown in Fig. 6. From Nussbaumer (1986) we find from the Si III flux ratio densities of [FORMULA] for a one-point model. Similarly we find from Nussbaumer & Schild (1979) for the observed C III flux ratio [FORMULA].

We have verified that the additional lines in Figs. 5 and 6 do not correspond to blue or red shifted components of our doublets.

For the nebula of RR Tel Espey et al. (1996) find [FORMULA] K and [FORMULA]. This supports the traditional picture (Hayes & Nussbaumer 1986) of radiative ionization and collisional excitation of the nebular spectrum. Schild & Schmid (1996) resolve the O III [FORMULA] 5007, 4363 line profiles into a [FORMULA] component near zero radial velocity, and a [FORMULA] component shifted by -20 km/s. The volume ratio of the low to high density components they estimate as 1000. They did not see any evidence for high speed ([FORMULA] km/s) mass motion.

4.4. The flux in the continuum

In Fig. 7 we show the decline of the continuum from 1978 to 1995. We give the mean of the fluxes at [FORMULA], as well as the flux at [FORMULA]. As can be seen from Fig. 8 the fluxes at [FORMULA] correspond to the stellar continuum, whereas [FORMULA] measures the nebular flux. For an approximate idea about the relative contributions of stellar and nebular fluxes to symbiotic spectra see Fig. 1 of Nussbaumer & Vogel (1989). From 1978 to 1988 there was a decline by a factor two in the stellar continuum, and of a factor three in the nebular continuum. The evolution after 1988 is compatible with the assumption of approximately constant luminosity.

[FIGURE] Fig. 7. Evolution of the continuum flux from 1978 to 1995 from IUE (before 1995) and HST (1995) observations. [FORMULA]: mean flux of measurements at [FORMULA] and [FORMULA]. [FORMULA]: flux at [FORMULA]. Fluxes are in [FORMULA]. The error in the 1995 HST value is much lower then in the IUE measurements.
[FIGURE] Fig. 8. The continuum flux of RR Tel. [FORMULA]: HST observations of July 16, 1995. [FORMULA]: ORFEUS data taken in September 1993. (J. Krautter and H.M. Schmid personal communication). [FORMULA]: HUT observation of March 12, 1995 (Espey et al. 1995). We also show the black-body radiation for [FORMULA] K and [FORMULA], reddened with [FORMULA]. Fluxes are in [FORMULA].

We now combine the HST continuum observations at [FORMULA] of July 16, 1995 with ORFEUS (Orbiting and Retrievable Far and Extreme Ultraviolet Spectrograph) and HUT (Hopkins Ultraviolet Telescope, Espey et al. 1995) observations at [FORMULA] ; they are given in Fig. 8. Mürset & Nussbaumer (1994) as well as Jordan et al. (1994) find for 1992 an effective temperature of [FORMULA] K. We fit a black-body emission of [FORMULA] K, corrected for interstellar reddening with [FORMULA] (Jordan et al. 1994) to the observed continuum. Whitelock (1988) derived a distance of 2.6 kpc. With these parameters a best fit is obtained with [FORMULA] which implies [FORMULA]. A comparison between a black-body emitter and a line blanketed NLTE model appropriate to RR Tel is given in Jordan et al. (1994).

The comparison of the observed continuum with the model calculation shows that the strong nebular lines observed with IUE and HST, in particular N V [FORMULA] 1240 are indeed seen on top of the stellar continuum. P Cygni profiles, if present, should therefore not be much disturbed by nebular emission.

Previous Section Next Section Title Page Table of Contents

© European Southern Observatory (ESO) 1997

Online publication: June 5, 1998