2 Observations and photometric reduction

Three first epoch plates were taken with the 40 cm refractor (f=6895 mm) at Sheshan station of Shanghai Astronomical Observatory in 1958 (see Table 1). All of these were taken without filter but are roughly in the B-passband. Perhaps due to weather changes during the observations, the number of detectable stars on the plates varies significantly, although the exposure times of these plates are not very different. Altogether, there are 115 different stars measured on the three plates.

  

Table 1: The three first-epoch plates taken with 40 cm refractor at sheshan station of Shanghai Observatory

   

Table 2: CCD Observational material taken with the Vainu Bappu Telescope

Frame	Observation date	Passband	Exposure (s)
24m4147i27f	1996 - 03-24	I	60
24m4147i28f	1996 - 03-24	I	180
24m4147i29f	1996 - 03-24	I	900
24m4147r30f	1996 - 03-24	R	60
24m4147r31f	1996 - 03-24	R	180
24m4147r32f	1996 - 03-24	R	900
24m4147b33f	1996 - 03- 24	B	180
24m4147b34f	1996-03-24	B	300
24m4147b35f	1996-03-24	B	1800
24m4147v36f	1996-03-24	V	60
24m4147v37f	1996-03-24	V	180
24m4147v38f	1996-03-24	V	1200
24m4147b30f	1996-03-25	B	1800
24m4147v31f	1996-03-25	V	1200
24m4147r32f	1996-03-25	R	900
24m4147i33f	1996-03-25	I	900

All plates were scanned using the PDS-1010MS instrument of Purple Mountain Observatory, Chinese Academy of Sciences. Window scanning was used, with the initial coordinates converted from the CCD data (see below). The second epoch data in this paper are CCD frames taken in 1996 with 2.34 m Vainu Bappu telescope at Kavalur, India (focal length 7558 mm, geographic position: E $121^{\circ}11^{\prime}11.3''$, N $31^{\circ}05^{\prime}47.6''$, altitude 713 m). Table 2 gives the detailed observing information. The main purpose of taking these frames is to do BVRI photometry, but they are also of interest for astrometry. Of these, four B-passband frames were used for the proper motion determination. The CCD was made by Photometrics Inc., type TK1024AB2, $1024\times 1024$ pixels, each $24 \mu \times 24 \mu$, the field is $11^\prime \times 11^\prime$. This field size defines the investigated region in our work. Image processing of the CCD data frames was done in the usual manner using bias subtraction and flat-fielding. The coincidence of the flat field frames (summed for each colour band) is better than a few percent in all the filters. The magnitudes were determined using DAOPHOT2 and ALLSTAR2 profile fitting softwares (Stetson 1987, 1992). The stellar point spread function (PSF) was evaluated using the Penny model of DAOPHOT2 from several uncontaminated stars present in each frame. The image parameters and errors provided by DAOPHOT2 and ALLSTAR2 were used to reject poor measurements. About $10\%$ of the stars were rejected in this process. After all the frames were reduced, Stetson's (1992) DAOMATCH and DAOMASTER routines were used to cross identify the stars measured on different frames of the cluster region.

   

Table 3: Coefficients of linear transformation of 3 B-passband CCD frames to the standard CCD frame (24m4147b33f)

Frame	X coordinate			Y coordinate
	a	b	c	a'	b'	c'
24m4147b34f	1.00	0.00	0.38	0.00	1.00	0.82
24m4147b35f	1.00	0.00	-0.31	0.00	1.00	1.49
24m4147b30f	1.00	0.00	16.42	0.00	1.00	-21.80
Note: x'=ax+by+c,y'=a'x+b'y+c', in units of pixel.

In order to estimate the astrometric accuracy of the four B-passband CCD frames, we took the first frame (24m4147b33f) as a standard frame. Rectangular stellar positions (in units of pixel) in the other three B-passband frames were linearly transformed to the same system as the standard frame (coefficients of the transformation are listed in Table 3). The few stellar positions in those three frames that have significant coordinate deviations (exceeding 2 pixels) in comparison with the standard frame (probably due to erroneous identification or severe image blending) were rejected before further reductions. Then we obtained an average frame in which each stellar position is the mean of the (x, y) coordinate in the four frames (in the same system). Coordinate deviations for stars in all four frames with respect to the average frame were calculated and stars with x or y deviations larger than 0.7 pixel were rejected. The resulting (x, y) coordinate dispersion for the remaining stars is about 0.2 pixel, corresponding to 0 $^{\prime\prime}_{\raisebox{.6ex}{\hspace{.12em}.}}$13 (see Table 4). We only used the remaining stars for further astrometric reductions (Table 4 also lists the number of stars used in each frame).

   

Table 4: Average positional accuracy of the position of one star on the B-passband CCD frames (unit: pixel)

Frame	$\sigma_{x}$	$\sigma_{y}$	star number
24m4147b33f	0.22	0.20	106
24m4147b34f	0.18	0.17	103
24m4147b35f	0.20	0.19	104
24m4147b30f	0.21	0.16	101

   

Table 5: Photometric errors (output of DAOPHOT) from the fit to the star profile

Parameter	Median error	Maximum error
V	0.025	0.296
B-V	0.043	0.404
V-R	0.045	0.301
V-I	0.043	0.303

The CCD instrumental magnitudes have been calibrated using local standards in the cluster field by Christian et al. (1985). The equations relating the instrumental magnitudes to standard magnitudes are:

$$\begin{eqnarray*}B & = & b-(0.3213\pm 0.0165)(b-v),\\ V & = & v-(0.1528\pm 0.0... ...(0.0738\pm 0.0287)(v-r),\\ I & = & i+(0.0706\pm 0.0156)(v-i), \end{eqnarray*}$$

where b,v,r,i represent instrumental magnitudes while B,V,R,Irepresent standard magnitudes.

The limiting magnitude of the photometry is close to V=22 mag, in this paper however, only stars with proper motion data available were considered, whose limiting magnitude is about B=17.6. Table 5 shows the photometric error distribution for the 115 stars. The complete photometric work will be given in a separate paper by A.C. Gupta.