1 Introduction

This second paper continues the evaluation study of the expected spectroscopic performances for the GAIA astrometric mission planned by ESA and the establishment of an extensive databank of input data for simulations of GAIA observations. As for the whole series, the results are also of general interest to ground-based spectroscopists working at moderately high resolving powers ( $\lambda / \bigtriangleup \lambda \ \sim \ 10^4$) in the near-IR region of the spectrum.

The astrophysical outlines of the GAIA mission are discussed by Gilmore et al. (1998, and references therein) and an outlook of the GAIA payload and spacecraft is presented by Mérat et al. (1999). The goals of GAIA spectroscopy, the merits of the 8500-8750 Å region and a summary of the topics to be addressed in the present series of papers are given in Table 1 of Munari (1999, hereafter M 99).

In Paper I (Munari & Tomasella 1999) we have built up a homogeneous observational databank composed by the spectra of 131 standard stars mapping the MKK classification system from types O4 to M 8 and luminosity classes from I to V.

In Paper II we present a library of 254 synthetic spectra mapping the part of the [$Z/Z_\odot$], $\log~g$ and $T\rm _{eff}$  space where the majority of GAIA targets will be located. The targets are mainly F-G-K-M stars with metallicities ranging from those of the galactic globular clusters to that of the Pop. I objects, thus we computed a grid of synthetic spectra for $-2.5\leq[Z/Z_\odot]\leq+0.5, 4.5\leq\log g\leq 1.0$ and $T\rm _{eff}$ $\leq$ 7500 K. Extension to O-B-A stars will be given later on in this series by Castelli & Munari (1999, in preparation). The synthetic spectra match in resolving power the observed ones ( $\lambda/\bigtriangleup\lambda$ = 20000), thus forming an ideal companion set to the data discussed in Paper I. M 99 outlined the superior merits for implementation on GAIA of the 8500-8750Å region compared to other near-IR wavelength intervals of similar $\bigtriangleup \lambda$ = 250 Å extension (in the current GAIA baseline configuration the range available for spectroscopy amount to $\bigtriangleup \lambda \sim 250$). The near-IR triplet of Ca II and the head of the hydrogen Paschen series lie here. Strong He I and N I lines are found here in early type spectra while in cooler stars lines of Fe I, Si I, Mg I and Ti I are abundant over the 8500-8750 Å.

 

Table 1: The Metallicity-Temperature-Gravity grid mapped by our synthetic spectra. The numbers-in-the-boxes give the corresponding electronic figure where the given triplet of spectra is plotted. The entries in slanted characters are the six spectra mentioned in Sect. 2, i.e. those with a too large percentage error in the flux derivative dH/d $\tau _{\rm Ross}$

Figure 1: All spectra have been computed over a range wider that the 8500-8750 Å considered in this series of papers. This plot show the full explored range (7650-8750 Å). The filled circles point to the K I, Na I, Fe I and Ca II lines discussed in Sect. 1

Figure 2: A sample of [$Z/Z_\odot$] = -0.5, $\log~g=4.5$ synthetic spectra arranged in a sequence showing the effect of varying the temperature

Figure 3: A sample of $T\rm_{eff}$= 6500 K, $\log~g$ = 3.0 synthetic spectra arranged in a sequence showing the effect of varying the metallicity

Figure 4: A sample of $T\rm_{eff}$ = 6500 K, [$Z/Z_\odot$] = -0.5 synthetic spectra arranged in a sequence showing the effect of varying the gravity. The thick dashes mark the 1.00 level of the continuum. This is an example of the Figs. 5-89 only available electronically

The argument raised by M 99 in support of the 8500-8750 Å interval originated both from data from the literature as well as from an extensive and high resolution mapping of the MKK classification scheme from the near-UV to near-IR wavelength domains (Tomasella & Munari, in preparation). However, telluric absorptions dominate over most of the near-IR (cf. Fig. 3 of Paper I), where GAIA is currently baselined to perform the spectroscopic observations. Therefore, a firmer assessment of the capabilities of GAIA spectroscopy from space must be supported by synthetic spectra which can map a finer and more complete grid of parameters than possible with observed spectra, without the dramatic contamination by the blocking telluric absorptions.

To this aim the computation of the synthetic spectra was not limited to the 250 Å interval currently baselined for implementation on GAIA (8500-8750Å), but instead expanded to cover the whole 7650-8750 Å range, thus including the the K I doublet at 7664, 7699 Å, the Na I doublet at 8183, 8194 Å and the lines of Fe I multiplet 60 at 8327 and 8388 Å. Such K I, Fe I and Na I lines are identified by filled circles in Fig. 1, together with the Ca II triplet at 8498, 8542 and 8662 Å.

We will comment in the last section of this paper how the synthetic spectra presented here support the M 99 conclusions about the superior performance of the 8500 - 8750 Å interval compared to equivalent $\bigtriangleup \lambda$ = 250 Å centered on the K I, the Fe I or the Na I doublets.