The largest change from the input physics of the NG-giant models is due to the different and larger line lists for TiO and water vapor. In Figures 1 and 2 we compare NG-giant models (dotted lines) to the PMS grid. The main difference between these models is the selection of the input line lists, we used the same version of PHOENIX and the same thermodynamical and opacity data (other than TiO and water vapor) for the calculations. Both sets of models were iterated to convergence with their respective setups, so the differences in the spectra are the results of both direct opacity changes and changes in the structure of the model atmospheres due to the different opacities. In the optical spectrum, the result is generally weaker TiO bands for the PMS models compared to the NG-giant models at but slightly stronger TiO bands at . The situation is slightly different for the water bands shown in Figure 2. The NextGen-type models show stronger H2O bands with less inter-band opacity than the PMS models. More importantly, the shape of some water bands are noticeably different between the two setups. The reason for this behavior is the changes in the structure of the atmosphere caused by the different degree of completeness of the water vapor line lists used (see Allard, Hauschildt & Schwenke, submitted). For low , the strengthening of the overall water opacity going from the NG-giant models to the PMS models produces weaker TiO bands due to changes in the structure of the atmospheres. However, in hotter models the water bands are not as important as the TiO bands and the net effect of the overall slightly stronger TiO lines produces stronger TiO bands in the optical. These effects are more pronounced for higher gravities (water opacity is relatively more important for larger gravities at the same effective temperature).
In general, the PMS setup applied to M dwarfs produces somewhat better fits to field stars , although the water bands are still not perfectly reproduced by the models . One reason for this are problems with the water line lists, but other opacity sources (such as dust formation in very cool models) as well as the treatment of convection in optically thin layers are additional sources of uncertainty.
. One reason for this are problems with the water line lists, but other opacity sources (such as dust formation in very cool models) as well as the treatment of convection in optically thin layers are additional sources of uncertainty.