2 Observations and data reduction

2.1 Photometry

S 1082 was monitored in the U, B, V, I and Gunn i bands on three occasions (see Table 1). In run 1 we observed the star during twelve nights with the 0.91m ESO-Dutch telescope at La Silla. The observing schedule was divided in four blocks of three nights; every first night the star was observed for an average of 5 consecutive hours in a U B V Gunn i-exposure sequence while every second and third night typically one to three exposures were taken in each filter. During both nights of run 2, S 1082 was observed for several hours with the 1m Jacobus Kapteyn Telescope on La Palma. In run 3, also on the Jacobus Kapteyn Telescope, we aimed to complete the phase coverage of the light curve between phases 0-0.3.

 

 

Table 1: Log of the photometric observations. For each run we give the date of the observations, the telescope and filters used and the typical exposure time for each filter.

Run	Dates	Telescope	Filters	$t_{\exp}$ (s)
1	Feb. 8-19 1999	0.91m ESO	U B	300 120
		Dutch	V Gunn i	120 120
2	Dec. 25, 26 1999	1m ING JKT	B V	75 30
3	Feb. 13-16, 20 2000	1m ING JKT	U B V I	350 30
				15 8

Standard reduction steps of bias subtraction and flatfielding were performed with IRAF routines. Aperture photometry for all the stars in the field was done with the DAOPHOT.PHOT task. Differential light curves for each individual dataset were computed with ensemble photometry (Honeycutt 1992). The variability properties of the other stars in the fields are discussed in van den Berg et al. (2001) and Stassun et al. (2001, in preparation). We refer the readers to these papers for a full description of the observations and the photometry reduction.

2.2 Spectroscopy

   
2.2.1 High-resolution spectra

High-resolution ( $R\approx49~000$) echelle spectra were taken with the Utrecht Echelle Spectrograph on the 4.2m William Herschel Telescope on La Palma. S 1082 was observed on four nights in 1996 and 2000 (see Table 2 for a log of all spectroscopic observations).

The 1996-spectra were centred on a blue (4250 Å) and red (5930 Å) wavelength. The 31 lines mm^-1 grating was used in combination with the $1024\times1024$ pixels² TEK-CCD. For a full description of the spectra of run 1 we refer to van den Berg et al. (1999).

The spectra of 2000 were all taken with the same instrumental setup: the 79 lines mm^-1 grating was used with the $2148\times2148$ pixels²SITe1-CCD while the spectra were centred on 5584 Å. In run 2 the seeing was about 2 $^{\prime\prime}$, while light clouds were present during the start of the run. The slit width was set to 1 $^{\prime\prime}$. To secure stability, no changes were made to the instrumental setup during the night. In the first two observations of run 5 the seeing was about 2 $^{\prime\prime}$; this deteriorated to 3 $^{\prime\prime}$  with cloudiness during the last two observations. As the slit width was kept fixed at 1 $^{\prime\prime}$, these spectra are of bad quality. In run 6 the seeing was 2 $^{\prime\prime}$ during the first two observations but later improved to 1 $.\!\!^{\prime\prime}$6. The slit width was accordingly changed from 2 to 1 $.\!\!^{\prime\prime}$2. Due to the wider slit these spectra have a lower resolution than the spectra of run 1 and 5. All frames were exposed for 1200 s, except for those of run 5 that were exposed for 1800 s to account for the bad seeing conditions. During each run we observed radial-velocity standards. Flatfield images were made with exposures of a tungsten lamp. Thorium-argon lamp emission-line spectra were taken for wavelength calibration.

 

 

Table 2: Log of the spectroscopic observations. For each run we give the date of the observations, the wavelength coverage in Å, the number of observations, the exposure time in seconds, the spectrograph used and (if applicable) the radial-velocity standard.

Run	Dates	$\Delta \lambda$	#	$t_{\exp}$	Inst	RV
1	Feb. 28 1996	3920-4920	2	600	ues	HD 136202a
				300
		4890-7940	3	360	ues	HD 136202
2	Feb. 16 2000	4380-8650	17	1200	ues	HD 89449b
3	Feb. 20 2000	3535-5035	1	900	ids	-
4	Feb. 22 2000	3535-5035	1	900	ids	-
5	Mar. 13 2000	4380-8650	4	1800	ues	HD 89449
6	Mar. 20 2000	4380-8650	5	1200	ues	HD 89449

^aF8III-IV, ^bF6IV (Simbad).

Data-reduction was done in IRAF with CCDRED and ECHELLE routines. After correcting the frames for the electronic bias and after flat fielding, spectra for each echelle-order were extracted with optimal extraction. Towards the red, gaps occur in the wavelength coverage. For the wavelength calibration, fifth-order polynomials were fitted in both directions to the positions of the arclines on the CCD; the maximum residuals to the fit were $\sim$0.0125 Å  corresponding to 0.75 km s^-1. Low-order polynomials were fitted to the spectra for continuum normalisation.

2.2.2 Intermediate-resolution spectra

Two intermediate-resolution ( $R\approx3600$) spectra were obtained with the Intermediate Dispersion Spectrograph (IDS) mounted on the 2.5m Isaac Newton Telescope on La Palma, on February 20.90214, 2000 (UT) and February 22.85523 (UT). The R1200B grating and the EEV10 CCD combination gave a spectral resolution of 1.19 Å (FWHM) and a useful wavelength range of 3533-4825 Å. The exposure times were 900 s each, and the signal-to-noise level in the extracted spectra ranged from about 65 per pixel near the Balmer jump to about 150 near 4750 Å.