Astron. Astrophys. 343, 536-544 (1999)

2. The data

2.1. Observations and observation strategy

We observed BE 381 and S 61 at the 3.5m ESO/NTT telescope (La Silla, Chile), on the nights of January 26-28, 1996 and July 31, 1998, using the STScI Coronograph mounted at the SUSI focal plane, and the EMMI spectrograph in Red Medium Dispersion (REMD) and Blue Medium Dispersion (BLMD). The journal of both coronographic and spectroscopic observations is given in Table 1. The NTT/STScI coronograph is described in detail in Clampin et al. (1994). The coronographic images were taken using an occulting bar of , in the light of H ( = 6562.7 Å, FWHM = 28 Å) and in a narrow V continuum filter ( = 5475.3 Å, FWHM = 77 Å), in excellent seeing conditions (as good as during the observations). The detector employed was a CCD TK1024AF with a format of 1024 1024 pixels, which was rebinned by 2 in order to achieve a plate scale of pixel^-1. The usual set of bias, dark and flat-field images was acquired together with images of the standard star HD60753 in order to properly calibrate the scientific exposures. Since both nebulae around BE 381 and S 61 were quite faint, images were taken with the central star located both under and outside the occulting mask, to achieve optimal contrast and full spatial coverage of the nebular region.

Table 1. Journal of coronographic and spectroscopic observations

The EMMI spectra were obtained with gratings 4, 6 and 7 which cover the wavelength ranges 3380-5280 Å, 6280-6870 Å and 4520-5820 Å with a spectral resolution (FWHM) of 6.3 Å, 6.3 Å and 2.6 Å, respectively. Two Tektronix CCDs were used: TK2048EB for the EMMI red arm (with a plate scale of pixel^-1) and TK1024AB for the blue arm with a plate scale of pixel^-1. We employed a longslit of dimensions by , oriented East - West. We obtained comparison spectra of Ar and HeAr for the wavelength calibration of the gratings, together with detectors bias and flat-field maps. We also observed the standard stars GD108 and Feige110 for the photometric calibration of the spectra. For the observations of BE 381, we positioned the slit on the star, and at two positions on the nebula, North and South respectively. In the case of S 61, we took spectra with the slit on the star and on the nebula, at North, North (July 1998) and South, respectively. We estimate our overall pointing accuracy to be of the order of .

2.2. The data reduction

All data were corrected for bias, flat-fielded, and cleaned to remove cosmic rays and bad pixels using IRAF routines. On the coronographic images, we masked the occulting bar by setting to zero the value of the pixels in the bar region. The spectra were reduced by following the reduction recipe outlined in the IRAF LONGSLIT package. First, the sky background of each frame was modelled with a Chebyshev function of low order into a surface, which was then subtracted from the frame itself. Second, we derived the wavelength calibration in three steps: we identified the emission lines of each comparison spectrum at its central rows and determined the dispersion correction using Chebyshev functions of order between 3 and 4. Then, we re-identified the same emission lines over the entire frame adopting a step of 2 rows and readjusting the dispersion correction at each step. The dispersion corrections obtained in this way were finally fitted into a surface with Chebyshev polynomials of order 6 which was applied to each frame for the final wavelength calibration. Third, we corrected the spectra for atmospheric extinction and calibrated them in flux using the sensitivity functions derived from the standard stars spectrum. We estimate an uncertainty on the resulting fluxes of about 10.

© European Southern Observatory (ESO) 1999

Online publication: March 1, 1999
helpdesk.link@springer.de