8 Comparison with accretion/wind models

A number of hydrogen line profile calculations, within various frameworks, have been done over the years. Hartmann et al. (1990) and Calvet et al. (1992) compute hydrogen line profiles for spherically symmetric wind models and for "cone'' geometry inner disk winds respectively. Pedrosa (1996) computed H ${\rm\alpha}$ line profiles in the context of a radially spherically symmetric isothermal wind model. Grinin & Mitskevich (1991) and Mitskevich et al. (1993) drop the assumption of a continuous wind, assume a clumpy structure for the wind and develop the so called "stochastic wind models''. Bertout (1977), Bertout (1979) and Bastian (1982) compute line profiles resulting from material radially infalling in spherical or axisymmetric geometry. Calvet & Hartmann (1992) use a "cone'' geometry for the infalling material as a rough approximation to the magnetospheric accretion picture. Hartmann et al. (1994) take a more realistic geometry for infall in the magnetospheric accretion scenario and Muzerolle et al. (1998a) extend the calculations by considering a multilevel atom.

The above models concentrate on computing line profiles for Balmer lines. Explicit model Pa ${\rm\beta }$ line profiles are not found in the literature. Results for Br ${\rm\gamma }$ lines are found in Hartmann et al. (1990) and in Muzerolle et al. (1998a) only.

The model Br ${\rm\gamma }$ profile shown in Hartmann et al. (1990) is the result from a non-isothermal wind with a spherically symmetric steady flow (their model 12). Comparing the model line profile with the high resolution observations presented here show no resemblance between them. The model profile peaks at a redshifted velocity larger than $50\ {\rm km\ s}^{-1}$ while observed profiles tend to peak at slightly blueshifted velocities. Also, the Af of the model profile is clearly smaller than unity, again in disagreement with the observations. Other near infrared lines computed by Hartmann et al. (1990) (e.g. Pa ${\rm\alpha}$ and Br ${\rm\alpha}$) show either a flat top with the line peak being redshifted or to display a very prominent and broad blueshifted absorption (P Cygni profile), again unlike any of the observed NIR lines shown here.

Muzerolle et al. (1998a) show model Br ${\rm\gamma }$ line profiles, arising in a magnetospheric accretion scenario, for four different viewing angles: inclinations of 10, 30, 60 and 75 degrees. Full widths at half maximum range from $75\ {\rm km\ s}^{-1}$ for the lowest inclination to $110\ {\rm km\ s}^{-1}$ for the highest inclination. These are narrower than the typical observed lines, which as we have seen in Sect. 6.2 above, display FWHM mostly between 100 and $300\ {\rm km\ s}^{-1}$ with about 60% of the lines having FWHM between 150 and $250\ {\rm km\ s}^{-1}$. Line wings in the model profiles extend to significantly lower velocities than in the observed profiles, especially for the lower inclination models. The 10 and 30 degrees inclination model results show the blue wing extending up to $80\ {\rm km\ s}^{-1}$and the red wing extending to about $125\ {\rm km\ s}^{-1}$. In contrast, all of the observed blue wings and most of the observed red wings are significantly more extended than those in the model profiles (see distributions in Fig. 6). Furthermore, contrary to the model predictions, the red wing in observed profiles is not more extended than the blue wing. For the 60 and 75 degrees inclination models (the model IPC profiles) the blue wing extends up to about $170\ {\rm km\ s}^{-1}$, while the red wing extends up to $130\ {\rm km\ s}^{-1}$. The redshifted absorption feature extends from there up to about $240\ {\rm km\ s}^{-1}$. These velocities match reasonably well those observed for the maximum observed velocity of the IPC profiles, but once again, fail at explaining the velocities observed in the blue wing of the profiles. Another difference between the model predicted and the observed profiles lies in the maximum line intensity. Model profiles show normalized intensities of at least about 1.8. Such high intensities are not observed even in TTS with strong Br ${\rm\gamma }$ emission. Quantitatively, judging from the model Br ${\rm\gamma }$ profiles presented in Muzerolle et al. (1998a), models do not match the observations very well. However, a greater exploration of the model's parameter space is required to assess this matter fully.

From a qualitative standpoint, Muzerolle et al. (1998a)'s results for models with higher inclinations yield Br ${\rm\gamma }$ profiles which are IPC in shape and which resemble some of the observed line profiles (e.g. DF Tau, RW Aur), even if in quantitative terms they are different (e.g. in their widths and peak intensities). The double peaked profiles, characteristic of lower inclination models, are not observed at Br ${\rm\gamma }$ in the sample of stars discussed here. None of the model Br ${\rm\gamma }$ profiles presented by Muzerolle et al. (1998a) is a type I profile. These type of profiles constitute the vast majority of the observed ones.