The Grid

The NextGen model grid was calculated from $3000\,{\rm K}$to $10000\,{\rm K}$ (in steps of 100-$200\,{\rm K}$) for $3.5 \le \log(g)\le 5.5$(in steps of 0.5) and metallicities of $-4.0 \le [{\rm M/H}]\le 0.0$ (in steps of 0.5). We use the recent solar abundances given in Table 5 of [Jaschek & Jaschek(1995)Jaschek and Jaschek]. The changes in the abundances compared to the previous [Anders & Grevesse(1989)Anders and Grevesse] data are in general small, with the exception of the iron abundance which changed from 7.67 to 7.53. The grid contains a total of 2142 models in this range of parameters (the complete NextGen grid includes 3026 models). In all models, convection is treated in the mixing length approximation with the mixing length set to unity. We will discuss the model grid for lower effective temperatures, the VLMS to Brown Dwarf range of spectral class M, in a separate paper. For effective temperatures higher than 7000-$10000\,{\rm K}$ NLTE effects become important, so LTE models at higher effective temperature will become increasingly unrealistic. In cooler models, NLTE effects are important for individual atomic lines of, e.g., Ti I [Hauschildt et al.(1997b)Hauschildt, Allard, Alexander, and Baron], for effective temperatures below about $4500\,{\rm K}$ but NLTE does not have significant effects on the low resolution spectra or the structure of these models. Therefore, we include a few representative NLTE models (for $7000\le\hbox{$\,T_{\rm eff}$}\le 10000\,{\rm K}$)in the model grid presented here. A full NLTE grid (with the complete set of NLTE species available in PHOENIX) is currently being constructed and will be discussed in a later paper (Hauschildt et al, in preparation).

[Aufdenberg et al.(1997a)Aufdenberg, Hauschildt, Shore, and Baron,Aufdenberg et al.(1997b)Aufdenberg, Hauschildt, Sankrit, and Baron] have shown that for effective temperatures higher than about $18000\,{\rm K}$ the combined effects of line blanketing and spherical geometry are of crucial importance for main sequence stars with $\log(g)\le 4.5$. In this parameter range, plane parallel models do not deliver enough EUV flux compared to spherical models or observed EUV spectra. However, for effective temperatures below about $10000\,{\rm K}$the use of the plane parallel approximation in the model construction for $\log(g)\ge 3.5$ does not result in significant differences to spherical models. Therefore, we have used the plane parallel geometry to calculate the present grid. A grid of NLTE spherical model atmospheres for OB main sequence stars is currently being constructed and will be discussed in a separate paper (Aufdenberg et al, in preparation). For effective temperatures lower than about $3000\,{\rm K}$ the effects of dust formation and/or dust opacity become important. This significantly changes the physics of the model atmospheres and the formation of the spectrum, therefore, we will discuss these models in a separate paper (Allard et al, in preparation).

The synthetic spectra and the model structures are available via anonymous FTP from ftp://calvin.physast.uga.edu/pub/NextGen or via the WWW URL http://dilbert.physast.uga.edu/yeti and constitute about 560 MB of data. For each model, the model structure (in the form of the PHOENIX output file of the last model iteration) and the synthetic low resolution spectra (directly the result of the model iterations) are available. The model fluxes are given as tables of $F_\lambda$ in erg/s/cm²/cm versus wavelength in Å to make comparisons with observed spectra and other model calculations easier. The spectra are given on an semi-regular wavelength grid with about 23,000 wavelength points from $1\hbox{\AA}$ to $1\,$cm, the actual wavelength grid depends on the model.