2 Choosing indices

The first task at hand is to choose the indices to be calibrated. The indices that have been considered for this problem are listed in Table 1. At the head of the list is the DDO CN index $\delta$CN, for which a number of calibrations have been published. Calibrations of interest here have been derived by [34, Taylor (1991]) and [45, Twarog & Anthony-Twarog (1996]). The first of the previous calibrations is based on an earlier version of the [40, Taylor (1999a]) catalog (see [34, Taylor 1991]). The second previous calibration is based on an expanded version of the 1991 catalog.

  

Table 1: Kinds of data considered for calibration

• [$^{\mathrm{a}}$]Relations given by [18, Janes (1975)] are used to obtain $\delta$CN from DDO photometry.
• [$^{\mathrm{b}}$]G is a weighted mean of feature-strength residuals for the strong features named (see Sect. V and Table 5 of [, Taylor & Johnson 1987]). "G'' is short for "giant,'' and does not refer to the G band. The data calibrated here include the original measurements of [32, Spinrad & Taylor (1969)].
• [$^{\mathrm{c}}$]No calibration is derived. See notes "d,'' "e,'' and "f.''
• [$^{\mathrm{d}}$]Metallicities derived from these data by [16, Hansen & Kjaergaard (1971)] are not used. Instead, the data are transformed to $\delta$CN (see Table 2, note "a''). This is done only for stars that lack direct measurements of the DDO indices.
• [$^{\mathrm{e}}$]See Fig. 1 of [27, McClure (1976)].
• [$^{\mathrm{f}}$]See Eq. (12) of [10, Eggen (1989b)].
• [$^{\mathrm{g}}$]D is a weighted mean of feature-strength residuals for the strong features named (see Sect. V and Table 5 of [, Taylor & Johnson 1987]). "D'' is short for "dwarf,'' and does not refer to the D lines. The data calibrated here are from [33, Taylor (1970)].
• [$^{\mathrm{h}}$]Measured values of U-B and $[(R-I)_{\rm C} - 0.007$ mag] are compared with a grid derived from model atmospheres. The $(R-I)_{\rm C}$ correction allows for the difference between the Buser-Kurucz solar value of $(R-I)_{\rm C}$ and the solar value deduced by [38, Taylor (1997)].

The DDO CN index, $\langle$Fe$\rangle$, and G were all considered in a previous review of supermetallicity [37, (Taylor 1996]). M₁ and $(R-I)_{\rm E}$ are included specifically to retrace the steps of [11, Eggen (1989c]), who used these parameters to derive a photometric metallicity for the SMR candidate star $\mu$ Leo. As noted in Table 1, Copenhagen photometry [8, (Dickow et al. 1970]) is not used in its published form, but some of it is accepted after being converted to $\delta$CN. The (38-42) and m₂ indices are noted in Table 1 because, like M₁, they have been calibrated by [10, Eggen (1989b, 1989c)]. They are not calibrated here because their spectral information duplicates that in $\delta$CN, G, and M₁.

For dwarfs, the D parameter discussed by Taylor & Johnson (Table 1) is calibrated. In addition, formal metallicities are obtained by using a theoretical grid [3, (Buser & Kurucz (1992]), with U-B and Cousins R-I chosen as arguments. The values of [M/H] from the grid are then calibrated.

For $\langle$Fe$\rangle$, G, M₁, $(R-I)_{\rm E}$, and D, source papers for the adopted calibration data are listed in Table 1. The sources used for DDO data are listed in Table 2.

  

Table 2: Sources of DDO photometry

• [$^{\mathrm{a}}$]This source contains Copenhagen photometry, and is used only if no directly-measured DDO data are available. The data are converted to DDO color indices by using the transformations of [18, Janes 1975] (see his Table 5).
• [$^{\mathrm{b}}$]Data from this source are preferred, as McClure & Forrester recommend.
• [$^{\mathrm{c}}$]This valuable secondary source contains data from almost all the primary sources in a readily accessible form.