2 Observations

Figure 1: R-filtered images of the observed comets (C/1999 F2: June 30, C/1999 J2: July 9, C/1999 N4: July 4, C/1999 T2: July 2, C/2000 H1: June 30, C/2000 K1: July 4). North is up, west is left. The images were scaled nonlinearly to enhance the visibility of diffuse outer regions.

Johnson V and $R_{\rm C}$ filtered CCD observations were carried out at Calar Alto Observatory (Spain) on three nights in July, 2000 (4th, 6th and 9th). The instrument used was the 1.23-m telescope equipped with the SITe#2b CCD camera (2048x2048 pixels giving an angular resolution of 0 $.\!\!^{\prime\prime}$49/pixel). The projected sky area is 16 $.\mkern-4mu^\prime$0$\times$16 $.\mkern-4mu^\prime$0, 10 $.\mkern-4mu^\prime$0$\times$10 $.\mkern-4mu^\prime$0 unvignetted.

The targets were selected following a few practical restrictions. Since the main aim was to observe distant comets with small expected coma contributions, we chose every visible comet at solar distances larger than 5 AU. The central brightness was restricted to be brighter than 19.0 mag, since time-series observations were planned to reveal rotation. Four comets remained as possible targets, and two further comets at a solar distance between 3 and 4 AU were added as auxiliary candidates. Finaly, six comets were examined.

We captured all of them a few days before starting the comet observations (on June 30th and July 2nd) in order to check which comets would be well-suited for our purposes. Three of them turned out to be unsuitable for the detailed analysis, thus we only characterize their comae and tails. Three objects remained as final targets for the morphological and photometric studies.

The exposure time was limited by two factors: firstly, the comets were not allowed to move more than two times the FWHM of the stellar profiles (varying from night to night) and secondly, the signal-to-noise (S/N) ratio had to be at least 20. This latter parameter was estimated comparing the peak pixel values with the sky background during the observations. The adapted exposure time was 240 s. The observing log is summarized in Table 1, listing also the main geometric parameters and sky conditions.

Figure 2: R-filtered surface brightness distributions (in mag/arcsec2) compared with the coma models. Note the logarithmic scale of the horizontal axes.

The CCD frames were reduced with standard tasks in IRAF. We obtained flat-fields taken during the evening twillight, and a master flat-field was formed with the task flatcombine. The photometric reductions were done with the corresponding tasks in IRAF/APPHOT. The trailed images and the presence of faint comae did not permit the use of psf-photometry. Therefore, a simple aperture photometry was performed. The applied differential photometry consisted of using two stars nearby as comparison and check stars. The cometary magnitudes are relative to the ensemble mean of the comparison, thus improving the precision of the differential data. We have carefully examined the aperture choice and a 2 $.\!\!^{\prime\prime}$6 diameter was accepted with respect to the mean FWHM and its doubled value. The time-series accuracy was estimated by selecting two nearby stars at similar brightnesses to those of the nuclei for the comparison and check stars. The usual scatter of comp minus check measurements was about $\pm$0 $.\!\!^{\rm m}$02.

Since all three nights were photometric, we could make an absolute "all-sky'' photometry using photometric standard stars taken from Landolt (1992). The standard field of PG1633+099 used was fairly close to the celestial positions of the observed comets. This field contains five standard stars covering V magnitudes between 12 $.\!\!^{\rm m}$969 and 15 $.\!\!^{\rm m}$256 and V-R colour between -0 $.\!\!^{\rm m}$093 and 0 $.\!\!^{\rm m}$618. The extinction was monitored by observing an A-type standard at different air-masses, and the nightly zeropoints of the standard transformations were determined with the other standards. The standard deviation of the linear fits is $\pm$0.02 mag, implying similar precision for the absolute values.

Another important correction specific for comet photometry was also applied. Licandro et al. (2000a) discussed the effects of the varying seeing on the photometry of blurred diffuse surfaces, such as those observed in comets. These authors outlined the following method: the actual seeing is determined in every CCD frame by examining stellar profiles. Then a seeing-magnitude relation is found with help of artificially blurred images of a non-variable comet. In this way the "seeing-effect'' on the magnitude determination carried out in the inner coma can be corrected for each observed frame with varying seeing. This "seeing-subtraction'' removes the some part of the atmospheric effect and the remaining variation can be attributed to the comet itself. The typical corrections did not exceed 0 $.\!\!^{\rm m}$1. The procedure adds a further $\sim$0 $.\!\!^{\rm m}$01 noise to the data and including all of the mentioned uncertainties, the photometric accuracy is estimated to be $\pm$0 $.\!\!^{\rm m}$05. The data reduction ends with the correction for the light time.