4.3 Non-photometric Corrections and Standard Calibration

A few of the galaxy images were obtained in non-photometric conditions. For these, we determined the offsets needed to bring them to the system of the photometric images. This is described in Sect.  (p. ). The derived offsets are listed in Table  (p. ). In summary, two Gunn r and two Johnson B images were found to be non-photometric, with needed offsets in the range $-0\hbox{$.\!\!^{\rm m}$ }01$ to $-0\hbox{$.\!\!^{\rm m}$ }11$. The five Johnson U images were brought to an internally consistent system with offsets in the range $+0\hbox{$.\!\!^{\rm m}$ }03$ to $-0\hbox{$.\!\!^{\rm m}$ }06$. The offsets were added to ${m_{\rm ell}}(r)$, ${m_{\rm circ}}(r)$, $\mu(r)$, and ${< \hspace{-3pt} \mu \hspace{-3pt}>}(r)$.

The standard calibration is described in Appendix  (p. ). A summary is given here. Three standard star fields were observed: PG1633+099 (5 stars), SA110 (8 stars), and M67 (23 stars). Since the M67 stars are very bright, they were observed with the telescope defocused in order to avoid very short exposure times. Standard star observations were made in Gunn r and Johnson V, B, and U on all the nights where the HydraI Gunn r and Johnson B galaxy images were taken, i.e. night 1, 2, 3, 4, and 6. In addition, standard star observations were made on night 7. No standard star observations were made on night 9 where the HydraI Johnson U galaxy images were taken. Aperture photometry was performed on the standard stars, and the magnitudes were corrected for the finite size of the aperture. The aperture corrected stellar magnitudes are denoted $m_{\rm raw}$.

The field SA110 was observed twice each night, at low and high airmass (typically X = 1.15 and 1.65, respectively), to determine the extinction coefficients for the four passbands. No significant color dependence was found for for the extinction coefficients. The derived extinction coefficients for night 4 and 6 were significantly larger than those for the other nights. It was decided to use the mean value of night 1, 2, 3, and 7 as extinction coefficient for all nights. See Table  (p. ).

After the observed stellar magnitudes had been corrected for extinction, the magnitudes still showed night-to-night variations at the level of $0\hbox{$.\!\!^{\rm m}$ }01$- $0\hbox{$.\!\!^{\rm m}$ }04$ (depending on passband). These relative offsets (so-called night coefficients) were used to bring the different nights to an internally consistent system. The night coefficients are given in Table  (p. ).

Instrumental magnitudes were then calculated from the raw magnitudes as

$$m_{\rm inst} = m_{\rm raw} -kX + n \enspace ,$$ (4.6)

where -kX is the extinction correction and +n is the night shift.

Standard star magnitudes were taken from Landolt (1992), Jørgensen (1994), and Montgomery, Marschall, & Janes (1993). Ten of the M67 stars did not have magnitudes in Johnson U, giving a total number of 26 stars in this passband instead of the 36 stars available in the other passbands.

We determined the following transformations

$$\arraycolsep=2pt % \begin{array}{llllrlrll} {r_{\rm std}}&=& ... ...=0.039 \\ & & & & \pm 0.017 & & \pm 0.006 & & \\ \end{array}$$ (4.7)

${\sigma_{\rm fit}}$ is the rms scatter. 3-7 stars were excluded from the fits, either because they had extreme colors outside the range needed, or because they deviated systematically; cf. Sect.  (p. ).

The raw galaxy magnitudes (as determined in Sect. ) were then standard calibrated using Eq. () and (). The Gunn r and Johnson B observations were made in pairs (cf. Table , p. ), and were standard calibrated in this way. The Johnson U observations were paired with the best seeing Johnson B observations. A night coefficient of zero was assumed for night 9.