Astron. Astrophys. 322, 598-614 (1997)

3. Evolution of the UV line blanketed spectrum at Z=0.020

Our line blanketed models allow us to predict a large number of detailed observable line features, which are well suited for comparison with UV spectra. In Paper II we have presented the spectral evolution in the 850 to 2200 Å range along MS tracks with initial masses between 40 and 85 . To provide the complete set of UV spectra at solar metallicity we here present the corresponding results for the additional tracks ( = 20, 25 and 120 ).

The synthetic UV spectra of the models given in Table 1 are plotted in Figs. 2 to 4. Together with the plots from Paper II, these figures illustrate the behaviour of the strongest UV features as a function of luminosity and effective temperature. The progressive behaviour of the strongest wind lines of CNO and Si agrees with the conclusions drawn in Paper II, where we refer to for more details.

Fig. 2. Synthetic UV spectra showing the spectral evolution on the 20 track. Plotted is the logarithm of the emergent luminosity. Approximate spectral types from Table 1 are given. Starting with the second model, each spectrum has been shifted downwards by 0.7 dex with respect to the previous one, in order to allow a good comparison. The marks on the top and bottom indicate the location of the CNO and Si lines taken from the lists of Bruhweiler et al.  (1981) and Dean & Bruhweiler (1985). The strongest CNO, and Si features are labeled

Fig. 3. Same as Fig. 2 for the 25 track (models B1 to B4)

Fig. 4. Same as Fig. 2 for the 120 track (models F1 to F3)

(a) The predictions for the N V resonance line follows the observed decrease in line strength towards later spectral types. The spectra of the B-type stars show some tendency to produce less N V than observed, which reflects the problem of super-ionization.

(b) The strong luminosity dependence of the Si IV line is reproduced.

(c) While the C IV 1550 doublet is predicted too weak for ZAMS models with 40 , we obtain a strong C IV P-Cygni line for less massive ZAMS stars. Compared to observations (see Snow et al.  1994) the C IV line of the 20 and 25 ZAMS models is too strong. Both results indicate that the predicted carbon ionization balance is shifted towards too high ionization stages for a given temperature (see Paper II).

(d) The predicted O V 1371 line shows a strong P-Cygni profile for the hottest models, while it is mostly observed in absorption. The reasons have been discussed in Paper II.

In Paper II we have performed several quantitative comparisons of metal line features. In particular we have been able to reproduce the strong observed Fe 1920 Å feature in late O and early-B giants and supergiants. This feature was found to be a good temperature indicator for these stars. The new tracks confirm this result.

© European Southern Observatory (ESO) 1997

Online publication: June 5, 1998

helpdesk.link@springer.de