Astron. Astrophys. 331, 1051-1058 (1998)

4. Calculations of damping for a sample of lines

To illustrate the effects of adopting the different broadening computations discussed in Sect. 3, damping widths and abundances were calculated for a set of red Fe I lines. The line list is given in Table 1, with the wavelengths and lower excitation potentials in columns (1) - (2). The angular momentum quantum numbers of the initial and final states are listed in column (5).

Table 1. Sample list of red Fe I and Fe II lines

The mean square radii of the final and initial states are given in Bohr radii², firstly for the hydrogenic approximation (columns 6-7) and secondly for the Unsöld approximation (columns 8-9). Recall that depends on the difference between the final and initial values (Eq. (1)). In comparison, WIDTH6 adopts a single value for all Fe I transitions - 19  - and one value for all Fe II transitions - 10  .

Columns (10) - (14) give the damping values log  computed for the five formalisms described in Sect. 3, for a temperature of 5000 K, which is typical of the atmospheres of G-K stars. Ratios of the values for various pairs of formalisms are given in columns (15) - (19). Three of these ratios are plotted in Fig. 1.

Fig. 1. Ratios of damping values computed by various formalisms: (Anstee & O'Mara)/ (Unsöld) (solid circles), and (WIDTH6)/ (Unsöld) (open circles). The asterisks and solid line show the empirical enhancement factors of Simmons & Blackwell and the linear approximation adopted here.

Fig. 1 shows that the WIDTH6 approach resembles the Unsöld approximation for lines around 2 eV. Recalling that Kurucz's motivation was to fit strong, low-excitation-potential lines, we see that WIDTH6 provides values for them comparable with mild enhancements () of Unsöld's approximation. However, considerably larger differences exist for the higher excitation potential lines, where WIDTH6 values are a factor of two lower than Unsöld's. Presumably this difference arises because Kurucz's treatment is based on a 4p excited state, which is not appropriate for all Fe I lines.

Fig. 1 also compares the Anstee & O'Mara damping values with those of Unsöld. The Anstee & O'Mara calculations correspond to mild enhancements of the Unsöld values for lower excitation potential lines, consistent with empirical results, but gives higher enhancements for lines around 4 eV. The Anstee & O'Mara values differ considerably from the WIDTH6 ones for lines above 2 eV, differing by a factor of 5 at 4 eV. Also shown in the figure are the empirical enhancement factors of Simmons & Blackwell (crosses) and the linear approximation (solid line) adopted for the calculations in Table 1.

© European Southern Observatory (ESO) 1998

Online publication: March 3, 1998
helpdesk.link@springer.de