2 Data reduction and line identification

Two separate observations of HD93521 were obtained during two successive orbits with a total integration time of 1740s (ORFEUS observation IDs 2276_2 and 2276_3, observation date: day 333 of 1996, GMT 04:56:05 - 05:14:05 and GMT 06:28:05 - 06:39:05). The two echelle images were coadded and then the standard extraction procedure was applied (Barnstedt et al. [1999]) without any smoothing. An additional radial velocity correction of -10 kms^-1 was applied, which corrects the wavelength scale for the fact that the star was not absolutely centered in the entrance diaphragm of the telescope. The maximum uncertainty due to the 20 $^{\prime\prime}$ diameter of the diaphragm was $\pm$36 kms^-1, so the deviation of -10 kms^-1 corresponds to a pointing offset of 3 $^{\prime\prime}$, which is an excellent value for the ASTRO-SPAS satellite. The value of -10 kms^-1 was estimated by comparing the observed radial velocity components with those already published (Spitzer & Fitzpatrick [1992]). The wavelength scale is heliocentric, a LSR scale would require an additional correction of -1.6 kms^-1, which is negligible.

As with all echelle spectra, the signal to noise ratio is best in the centre of the echelle orders, while it is reduced by a factor of about 1.4 at both ends of each order. Due to the blaze curve being not fully centered, the signal to noise ratio at the short wavelength end is significantly better than at the long wavelength end of each echelle order. There is also a slight deviation in the wavelength calibration at the short wavelength end of each order, which affects a wavelength range of about 5% in each order.

For line identifications we used the following line catalogues:

1.

H₂-lines:

-

Morton & Dinerstein ([1976]), except L10P1;

-

Abgrall et al. ([1993]), correct wavelength for L10P1: 982.834Å.

2.

Metal lines and HI:

-

Morton ([1991]);

-

Kurucz CD No. 23, web-page (Kurucz [1995]);

-

Kelly ([1968]; and web-page);

-

Feibelman & Johannson ([1995]).

Identified lines from IUE spectra of HD93521 above 1170Å are listed by Ramella et al. ([1980]), with exception of two interstellar lines: $\lambda$1260.4, which is SiII and not SiIII, and $\lambda$1304.4, which is also SiII and not OI.

The tables and plots show blended lines also, for which a non-ambiguous identification or estimate of the intensity is not possible.

Lines with a lower energy level greater than zero are marked with an asterisk, *. All stellar lines are marked with a bracketed asterisk, (*), and stellar wind lines are marked as (w) in the plots as well as in the tables.