2 Observations and data reduction

The multi-star CCD photometric monitoring was carried out for 21 nights during February/March 1994 using two 1-metre class telescopes located at Naini Tal and Kavalur in India, at a separation of $\sim$2500 km. Due to the high declination of the blazar the airmass was generally larger than 1.5, but the object could be monitored typically for about 5 hours in a night. The positional offsets, the calibrated BV Johnson and RI Cousins magnitudes and instrumental colours of the three comparison stars used for differential photometry of the blazar are listed in Table 1. The photometric measurements for stars 1 and/or 2 have been carried out by Takalo et al. (1994); Ghisellini et al. (1997) and Villata et al. (1998) in the different passbands. Within the errors, the magnitudes derived by them for a star agree fairly well. For stars 1 and 2, the BVR data are taken from Villata et al. (1998) as they are the most accurate measurements; while the I magnitudes are adopted from Ghisellini et al. (1997). For star 3, these values are determined from stars 1 and 2, using the differential instrumental magnitudes measured in our campaign. A log of observations for the entire campaign is given in Table 2. Further details of the observations are summarized below.

 

Table 1:  Positional offsets in arcmin and the differential ($\Delta$) instrumental (b - v) and (v - i) colours in mag of the comparison stars are relative to the blazar S5 0716+71. Stars observed by Takalo et al. (1994), Ghisellini et al. (1997) and Villata et al. (1998) have been identified with prefix T-, G- and V- respectively. BVRI are standard magnitudes

 

Table 2:   Log of the CCD observations

2.1 The UPSO observations

The BV Johnson and RI Cousins photometric observations were carried out using the Photometrics CCD system at the f/13 Cassegrain focus of the 104-cm Sampurnanand reflector of the Uttar Pradesh State Observatory (UPSO), Naini Tal on 19 nights in February/March 1994 (Table 2). Typical seeing was around 1.5$^{\prime\prime}$. Details of the CCD system are reported in Mohan et al. (1991). Typical exposure times were 20-60 s in R and I, 1-2 min in V and 3-5 min in B. The observed comparisons are stars 1 and 2 (Table 1).

2.2 The VBO observations

CCD photometric observations in R and I Cousins passbands were made using the Thomson-CSF TH7882 CCD chip mounted at the f/13 Cassegrain focus of the 102-cm telescope of the Vainu Bappu Observatory (VBO) at Kavalur. Typical seeing was around 2$^{\prime\prime}$. Details of the instrumental set up can be found in Sagar & Pati (1989). Typical exposure times were 5-10 min in both R and I passbands (owing to large zenith distance of the object and frequent cloudy conditions). The observed comparisons are stars 1 and 3 (Table 1).

2.3 The differential light curves (DLCs)

At both sites, bias and dark frames were taken intermittently throughout the observations. The pixel size of 23 $\times$ 23 ($\mu$m)² of both CCD systems corresponds to 0.36 (arcsec)² on the sky. The chip size of 384 $\times$ 576 pixels covers an area of 2.3 $\times$ 3.4 (arcmin)². In order to improve the signal/noise ratio the data were acquired in the binning mode of 2 $\times$ 2 (pixels)². In order to detect small intensity fluctuations, we performed differential photometry, generally ensuring that the locations of the blazar and the comparison stars did not change by more than a few pixels from exposure to exposure during a night. This largely obviated the need to flat-field the frames (which introduces additional noise). The background was subtracted using the measured counts within the annuli (between radii of 25 and 35 pixels) around the circular aperture of 5 pixel radius, centred at the desired object.

  

Figure 1: DLCs in the B, V, R, I passbands deduced from the UPSO observations on 19 nights (excepting for February 28 on which I-band data could not be obtained, whereas for the last 3 nights B-band observations could not be made). The symbols crosses, squares, filled circles and triangles refer to B, V, R and I passbands respectively. In order to plot the DLCs with clarity in the B, V, R, I sequence from bottom to top, we have applied fixed offsets of 0.25 mag in B, 0.0 mag in V and -0.1 mag in R and I for the DLCs involving the blazar and 0.1 mag in B, 0.0 mag in V and -0.05 mag in R and -0.1 mag in I for the DLCs involving the two comparison stars 1 and 2

  

Figure 2: DLCs in the R and I passbands obtained at VBO for 5 consecutive nights. The temporal overlap with the UPSO data on the first 3 nights (Fig. 1) shows good consistency between the two datasets. Note that the comparison star 1 is common to the two datasets. For clarity in plot, the offsets applied are -0.1 mag in both R and I for the DLCs involving the blazar, and -0.1 mag in I and 0.0 mag in R for the DLCs involving the two comparison stars 1 and 3

The data were reduced at the VBO VAX 11/780 system using DAOPHOT software package, as described in Gopal-Krishna et al. (1993; 1995). In order to maximise the signal/noise ratio, we performed the aperture-growth analysis described by Howell (1989). Accordingly, a circular aperture of 10 pixels (=7$^{\prime\prime}$) diameter was adopted for the photometry of the blazar as well as the comparison stars. The DAOPHOT algorithm also provided estimates of the statistical errors of the computed relative magnitudes, which were usually < 1% for all the passbands. Using the computed instrumental magnitudes, differential light curves (DLCs) were generated. These represent the relative instrumental magnitudes of the blazar relative to the two comparison stars, and the same for the pair of comparison stars (Figs. 1 and 2). Note that the comparisons used are stars 1 and 2 for the UPSO DLCs and stars 1 and 3 for the VBO DLCs. It is seen that the average instrumental colours of the blazar are not much different from those of the three comparison stars (Table 1).