![]() | ![]() |
Astron. Astrophys. 338, L91-L94 (1998) 2. Observations and reductionsSpectroscopic observations was performed using a Boller & Chivens cassegrain spectrograph at 1.52m ESO telescope of La Silla (Chile). A UV-flooded thinned Loral Lesser CCD #39 (2048 x 2048, 15µm/pixel) detector was used. Two gratings, with different dispersions, were used. One grating, with 600 l/mm which has a reciprocal dispersion of 1.9 Å pixel-1 and with a resolution of 4.6 Å, was used for spectral coverage between 3500 Å and 7500 Å. The other grating, with 1200 l/mm which has a reciprocal dispersion of 1.0 Å pixel-1, with a resolution of 2.0 Å was used for spectral coverage between 3100 Å and 5100 Å. The slit width was 4 arcsec. The slit orientation in the blue range was aligned with the parallactic angle in order to minimize the slit loss due to atmospheric refraction. Table 1 shows the log of observations and the typical S/N noise ratio in the continuum achieved in each observation. Table 1. Observation log of SPH 2. The data were reduced to the linear scale, i.e. wavelength versus
flux, using IRAF. The data were treated with the help of the
ccdproc and onedspec tasks of IRAF. We followed the
standard procedure consisting of bias subtraction, flat-field
normalization and wavelength calibration through a He-Ar lamp. Counts
were corrected from atmospheric extinction and calibrated from
instrumental chromatic response through observations of standard stars
from Oke (1974) and from Hamuy et al. (1994). In the linearized
spectra, the line fluxes were measured with the splot task and
blends were resolved using the deblend option. Figs. 1 and 2
show the reduced spectra of our sample. We estimate the errors in the
fluxes to be about 20% for weaker lines (line fluxes
© European Southern Observatory (ESO) 1998 Online publication: September 17, 1998 ![]() |