3 Stellar photometry

Stellar photometry was done inside the DIGIPHOT.DAOPHOT package in IRAF. In particular we found the stars using DAOFIND on the deep, new F555W images, then aperture and psf-photometry were obtained using PHOT and ALLSTAR on the same images. For the aperture photometry we used an aperture of 2 pixels. For the psf-photometry psfs variable over the chips, constructed from the images themselves, were used and the output statistics, i.e. $\chi$and sharpness, were used to weed out non-stellar sources.

This created a list of stars that were then used for obtaining aperture photometry from the new F814W images, which together with the F555W data gave a first colour-magnitude diagram. The positions of the stars making up the colour-magnitude diagram provided our source list of objects to find proper motions for.

For the images with short exposures we performed aperture photometry on the first epoch images for each filter and then merged the F555W and F814W data into a colour-magnitude diagram. Keeping only stars detected in both F555W and F814W this provided a master list for bright objects to find the proper motions for in the short exposures.

The final colour-magnitude diagrams are all based on the aperture photometry.

   
3.1 Calibration

We obtained instrumental ($-2.5\log$(counts/s),) magnitudes from both the old, short and new, long exposure data. These magnitudes were corrected to the HST/WFPC2 magnitude system, and merged to form a final single photometry set as described below.

The final stellar magnitudes within the HST/WFPC2 system i.e. V₅₅₅ and I₈₁₄, were obtained by applying a number of corrections to the instrumental magnitudes, essentially following Holtzmann et al. (1995b).

The photometry was obtained from drizzled images. The drizzling procedure removes the geometric distortion from the images, therefore no geometric distortion correction was required for the photometry.

Aperture corrections to a 0.5 arcsec aperture were obtained from our own images. These corrections were allowed to vary with distance from the centre of the chip i.e. correction $= a + b\times$distance The values of the constants a and b are given in Table 3. For the WF4 long data there were not enough stars to obtain both a and b from our data. In this case we used the value of b from Gonzaga et al. (1999) and only fit the value of afrom our data.

 

 

Table 3: Values of constants found for aperture correction in each image, see Sect. 3.1.

	F555 W		F814 W
	a	b	a	b
Short
WF2	0.17	1.70e-4	0.21	1.78e-4
WF3	0.21	1.47e-4	0.23	1.90e-4
WF4	0.27	2.21e-4	0.23	1.59e-4
Long
WF2	0.24	5.42e-5	0.27	4.35e-5
WF3	0.24	2.16e-4	0.28	1.84e-4
WF4	0.24	1.17e-4	0.23	1.07e-4

Corrections for the charge transfer (in)efficiency (CTE) were applied. The short data were obtained before the WFPC2 cool-down on 23/4/1994. The CTE problem is worse for the pre-cool-down data, but the corrections are not well modeled. For these data the CTE problem was corrected by assuming that the star counts lost were a linear function of the y-position on the chip, with stars at the top of the chip losing the maximum 10% (Holtzmann 1995b). For our new long data we used the equations in Whitmore et al. (1999) to correct for CTE.

The zero-points used to transform to HST/WFPC2 magnitudes are from Baggett et al. (1997).

3.2 Merging long and short photometry

Figure 1: Colour-magnitude diagram for all three WF. All stars measured in both the new and old images and which satisfied the cuts imposed in $\chi$ and sharpness variables (see Sect. 3) on the individual WFs are included. Stars with fitting errors are excluded.

To make the final colour magnitude diagrams we have merged the photometry from the long and short exposures. We compared the magnitudes of the same stars from the short and long exposure data, and found no evidence for any offsets. Further, we found that the long exposure data were saturated for V₅₅₅<16.8 and I₈₁₄<15.

The two sets of long and short photometry were then merged according to the following rules

• if the star has both long and short photometry then use long if V₅₅₅>16.8 and I₈₁₄>15, otherwise use short;
• if the star just has short photometry then use it if $V_{555}\le 16.8$ or $I_{814}\le15$;
• if the star just has long photometry then use it if V₅₅₅>16.8 and I₈₁₄>15.

Our final colour-magnitude diagram is shown in Fig. 1. This includes all stars selected according to our selection criteria for $\chi$ and sharpness. We also require the stars to all have good positions according to the photometric routines in DAOPHOT. No corrections for differential reddening have been applied.