3 Photometry

To check whether profile fitting photometry could be carried out on the modified NTL images, we compared the photometries carried out on original image and on modified images and the results of a particular night are displayed in Fig. 1. The comparison shows good agreement between the magnitude derived from original and modified images. Therefore we carried out photometry on the modified images. In order to increase the signal to noise ratio, all the images of a night were stacked together resulting in one frame per filter per night. This then provides us with 133 R and 166 I data points which were used in further analysis.

Using the DAOPHOT "FIND'' routine, we identified $\sim$4400 resolved stars in the reference frame at $4\sigma$ detection level. Stellar photometry for all the images in both filters has been carried out for these resolved stars in "fixed-position mode'' using DAOPHOT photometry as described by Stetson (1987). PSF was obtained for each frame using 25-30 uncontaminated stars. The DAOPHOT/ALLSTAR (Stetson 1987) routine was used to calculate the instrumental magnitude of the detected stars for each individual frame. The internal errors estimated from the S/N ratio of the stars as output of the ALLSTAR are given in Table 3 as a function of brightness. The error become large (>0.1 mag) for stars fainter than $R \sim20.0$ mag.

Figure 1: The plot shows difference in photometries carried out on original and modified images as a function of magnitude. In the upper panel original pixel size (0.74 arcsec) was scaled down (0.5 arcsec), whereas in the lower panel the pixel size (0.37 arcsec) was scaled up (0.5 arcsec).

3.1 Photometric calibration

The absolute calibration has been done using Landolt's (1992) standard field SA98 observed on a good photometric night of 25/26 October, 2000. The airmass ranges from 1.3 to 2.1 during the observations. Atmospheric extinction coefficients determined from these observations are $0.11\pm0.01$ and $0.08\pm0.01$ mag/airmass in R and I filters and they have been used in further analysis. Thirteen standard stars having a range in brightness (11.95 < V < 15.84) and colour (0.09 < (R-I) < 1.00) were used to derive following transformation equations.

$$\Delta(R-I) = (0.96 \pm 0.01) \times \Delta(r-i)$$ (1)

$$\Delta(R-r) = (0.03 \pm 0.03) \times (R-I)$$ (2)

where R and I are the Landolt standard magnitudes while r and i are the corresponding instrumental magnitudes. The zero point errors are about 0.02 mag in R and 0.01 mag in (R-I). Equations (1) and (2) were used to generate 50 secondary standards in the target field observed on the same night by accounting the differences in exposure times and air-masses. To standardize the remaining stars, differential photometry has been performed with these secondary stars rejecting those which were showing more than $3\sigma$ deviation. This process yields an accuracy of 0.03 mag in zero points. A variation of $\pm$0.1 mag was found around the mean value in the secondary stars itself during the whole observing runs which can be treated as an accuracy in NTL photometry.

3.2 Comparison with the previous photometries

A comparison of the present I band data with those available in the literature (Magnier et al. 1992; Mochejska et al. 2001 for the DIRECT collaboration) has been shown in Fig. 2 as it is the only common filter. An offset of 0.13 mag is observed between our and Mochejska et al. (2001) Iband photometry while a difference with Magnier et al. (1992) data indicates a magnitude dependence with a slope of 0.02. A similar slope is seen between Mochejska (2001) and Magnier et al. (1992) data. These results suggest that Magnier et al. (1992) data is colour dependent while Mochejska et al. (2001) data has zero point offset.

Figure 2: Comparison of the present I band CCD photometry with those given by Magnier et al. (1992) and Mochejska et al. (2001).