2 Observations and reduction

 

 

Table 3: Number of standard stars (N) and their rms residuals ($\sigma$) of each night of observation. V_X denotes V magnitude computed using colour index X in the colour-term-dependent part of the transformation equations

Night	VB-V			B-V			U-B			V-R			V-I			VV-R
	N	$\sigma$		N	$\sigma$		N	$\sigma$		N	$\sigma$		N	$\sigma$		N	$\sigma$
1991-11-05	22	0.021		21	0.018		22	0.027		23	0.014		23	0.023		22	0.021
1991-11-06	21	0.019		21	0.012		20	0.035		21	0.015		13	0.020		21	0.019
1991-11-07	20	0.026		20	0.018		17	0.018		20	0.015		20	0.020		20	0.026
1991-11-08	17	0.014		15	0.009		17	0.022		13	0.011		18	0.018		16	0.010
1991-11-09	21	0.014		20	0.010		19	0.030		21	0.012		20	0.023		21	0.014
1993-10-07	19	0.017		18	0.025		16	0.043		19	0.014		19	0.022		19	0.017
1993-12-10	18	0.021		-	--		-	--		17	0.017		17	0.020		18	0.021
1993-12-11	24	0.011		24	0.022		23	0.024		-	--		-	--		-	--
1993-12-12	23	0.049		21	0.030		19	0.030		18	0.021		17	0.019		16	0.050
1993-12-15	17	0.019		17	0.016		16	0.042		15	0.013		16	0.029		17	0.019
1993-12-16	23	0.039		22	0.038		21	0.043		19	0.035		18	0.050		22	0.034
1993-12-17	23	0.037		22	0.048		23	0.035		21	0.036		19	0.027		24	0.041
1993-12-18	19	0.009		20	0.018		18	0.050		20	0.012		18	0.017		19	0.009
1993-12-19	25	0.018		23	0.028		23	0.041		23	0.016		23	0.028		23	0.025
1994-12-08	25	0.013		24	0.024		22	0.041		25	0.013		25	0.022		23	0.011
1994-12-09	19	0.028		19	0.028		17	0.035		20	0.021		19	0.021		19	0.028
1994-12-10	10	0.030		10	0.040		9	0.045		11	0.016		10	0.025		11	0.040
1994-12-11	18	0.021		20	0.012		17	0.054		20	0.015		19	0.023		19	0.022
1994-12-12	35	0.017		32	0.013		32	0.037		32	0.011		30	0.021		32	0.019

The data were acquired in the course of several campaigns devoted to obtain deep UBVRI Johnson-Cousins CCD photometry of open clusters and stellar associations (Cep OB3, IC 348, NGC 1750/NGC 1758; Jordi et al. 1995; Trullols & Jordi 1997; Galadí-Enríquez et al. 1998) in November 1991, October 1993, December 1993 and December 1994 at the telescopes of Centro Astronómico Hispano-Alemán (CAHA) and Observatorio Astronómico Nacional (OAN), both in Calar Alto, Almería (Spain). In these observational runs, Landolt stars were used as reference for the transformation to the standard photometric system. Table 1 shows the telescopes and chips used in each observation period. Table 2 gives the quantum efficiency of the detectors at the central wavelengths of the standard UBVRI filters.

The reduction from raw images to standard photometry was performed following Jordi et al. (1995), and we refer to them for a fully detailed description of the procedure. In the following paragraphs we summarize the main steps of this process.

Bias level was evaluated individually for each frame by averaging the counts of the most stable pixels in the overscan areas. The 2-D structure of the bias current was determined from a number of dark frames with zero exposure time. As pointed out in previous papers (Galadí-Enríquez et al. 1994; Jordi et al. 1995), shutter timing effects can affect the photometric results, specially when dealing with bright stars (short exposure times). The shutters of every CCD-camera were analyzed following Galadí-Enríquez et al. (1994), and shutter effects were removed from flat-field and object frames.

The frames were processed using the ESO image processing software MIDAS. Aperture photometry was obtained using DAOPHOT, and aperture corrections were determined and applied with DAOGROW (Stetson 1987, 1990). Cross-identification of stars in different frames was performed using DAOMATCH and DAOMASTER programs (Stetson 1993).

In order to perform the transformation to the standard system, between 15 and 30 different Landolt (1983, 1992) standard stars, carefully selected to cover a wide range of spectral types and air masses, were observed each night.

The coefficients of the transformations were computed by a least square method using the instrumental magnitudes of the standard stars and their standard magnitudes and colours in the Johnson-Cousins system. Standard stars with residuals greater than 2$\sigma$ were not used in the transformation procedure. The rejected stars in each night were few (from 0 to 3 stars). Since these stars were different from night to night, the problem cannot be associated to their standard values: indivudal measurement problems are most probably the cause. Each rejection was checked in order not to reduce the colour range covered by the standard stars. The calculation was done in two steps, determining first the extinction coefficients (Eqs. (2) and (3) in Jordi et al. 1995) and, then fitting the remaining model parameters. Independent reductions were made for each night. The differences among instrumental coefficients from night to night within the same observing period were small and within the determined errors. The rms residuals of the standard stars are given in Table 3.

The transformation equations were applied to the instrumental data of the stars detected in the neighbourhood of Landolt standard stars. The internal errors of individual measurements were computed as described in Jordi et al. (1995), taking into account the errors in the instrumental magnitudes on the one hand, and the errors in the transformation equations on the other hand. No evidence of systematic differences between data acquired in different observational runs was found, what indicates that the transformations succesfully compensated the sensitivity differences among the instrumental systems.

Thus, final magnitudes and colours were obtained by averaging the individual measurements of each star using the internal error for weighting (Jordi et al. 1995; Rosselló et al. 1985), after rejecting obviously wrong measurements (those with deviation from the mean larger than $2\sigma$). The photometric errors were computed as the mean error of the mean of the final magnitudes and colours.