10 Conclusion

From the Pa ${\rm\beta }$ and Br ${\rm\gamma }$ observations of a sample of 50 T Tauri stars we conclude that blueshifted absorptions in the profiles of those lines are very rare. Of the 42 stars observed to have Pa ${\rm\beta }$ in emission only one shows a blueshifted absorption, while of the 30 displaying Br ${\rm\gamma }$ in emission none has a blueshifted absorption. The most common profiles are of type I (54% at Pa ${\rm\beta }$ and 73% at Br ${\rm\gamma }$). IPC line profiles constitute 34% of the Pa ${\rm\beta }$ profiles and 20% of the Br ${\rm\gamma }$ profiles. These figures are in complete contrast to those of H ${\rm\alpha}$, the most studied hydrogen emission line in TTS. Comparison with higher members of the Balmer series reveal some similarities between the two sets of lines, especially in that they are more prone to display IPC structure than H ${\rm\alpha}$.

 

 

Table 8: Stars for which the NIR lines were observed more than once. The day when an observation was made, is marked by either "Pa ${\rm\beta }$'' or "Br ${\rm\gamma }$''

Star	UT9410				UT9512
	02	03	04	05	15	16	17
DG Tau	-	Pa ${\rm\beta }$	-	-	Pa ${\rm\beta }$	Pa ${\rm\beta }$	-
DL Tau	-	Pa ${\rm\beta }$	-	-	Pa ${\rm\beta }$	Pa ${\rm\beta }$	-
DR Tau	-	Pa ${\rm\beta }$	-	-	Pa ${\rm\beta }$	Pa ${\rm\beta }$	-
	-	Br ${\rm\gamma }$	-	-	-	-	Br ${\rm\gamma }$
GG Tau	-	-	-	Br ${\rm\gamma }$	-	-	Br ${\rm\gamma }$
GI Tau	Pa ${\rm\beta }$	-	-	-	Pa ${\rm\beta }$	-	-
	-	Br ${\rm\gamma }$	-	-	-	-	Br ${\rm\gamma }$
GK Tau	Pa ${\rm\beta }$	-	-	-	Pa ${\rm\beta }$	-	-
	-	Br ${\rm\gamma }$	-	-	-	-	Br ${\rm\gamma }$
RW Aur	-	-	-	-	Pa ${\rm\beta }$	Pa ${\rm\beta }$	-
	-	-	-	Br ${\rm\gamma }$	-	-	Br ${\rm\gamma }$
RY Tau	-	-	-	-	Pa ${\rm\beta }$	Pa ${\rm\beta }$	-
SU Aur	-	-	-	-	Pa ${\rm\beta }$	Pa ${\rm\beta }$	-

The Pa ${\rm\beta }$ and Br ${\rm\gamma }$ lines are very wide (FWHM $\approx 200\ {\rm km\ s}^{-1}$), slightly blueshifted and with line wings extending to about $300\ {\rm km\ s}^{-1}$ in the blue and to about $200\ {\rm km\ s}^{-1}$ in the red. The Asymmetry Factor (Af) is slightly larger than one for most cases, with the Br ${\rm\gamma }$ distribution for this parameter showing a larger spread than that of the Pa ${\rm\beta }$distribution. The former also displays larger Afs than the latter. A significant number of IPC Br ${\rm\gamma }$ lines are displaced to the blue by about $50\ {\rm km\ s}^{-1}$ relative to the IPC Pa ${\rm\beta }$ lines.

If the line emitting region for Pa ${\rm\beta }$ sits in front of the whole stellar disk, the line is optically thin. Alternatively, the filling factor of the line emitting region is small and the stellar photosphere is also directly observed.

Comparing the data presented here and the results available in the literature from models for the formation of the hydrogen lines reveal that both wind and accretion models fail to explain the observed line profiles in most cases. The accretion models, computed in the context of the magnetospheric accretion scenario, do provide a qualitative insight on how these lines might be produced in the T Tauri stars' environment but fail under a quantitative comparison. The models produce lines too narrow (by $\sim 100\ {\rm km\ s}^{-1}$ FWHM), with wings extending to velocities too small (by at least $\sim 100\ {\rm km\ s}^{-1}$) and with far too high maximum normalized intensities (by factors of a few). If nothing else, the discrepancies found between observations and model line profiles hint that the axi-symmetric models considered thus far are just a rough approximation to the real accretion flows in T Tauri stars.

The redshifted absorption feature in the Pa ${\rm\beta }$ and Br ${\rm\gamma }$ IPC profiles must be formed in infalling material. It is located at velocities of the order of the free-fall velocity from a few radii out for a typical T Tauri star. There seems to be a trend associating the amount of emission seen in the IPC profiles and accretion rates, suggesting that lines with this type of profile originate mostly, if not completely, from infalling material. A similar trend does not seem to be present for type I profiles. However, with the exception of the lack of redshifted absorption feature, the latter display similar characteristics to the IPC profiles. In particular, they are centrally peaked and slightly blueshifted, characteristics that are very difficult to obtain in wind models (Calvet & Hartmann 1992) but arise naturally in inflow models due to absorption of infalling redshifted material.

Continuous wind models tend to produce lines with normal P Cygni profiles but Pa ${\rm\beta }$ or Br ${\rm\gamma }$ calculations are seldom found in the literature. Stochastic wind models might be able to produce profiles similar to type I but there are no specific predictions for the Pa ${\rm\beta }$ and Br ${\rm\gamma }$ lines. The way in which winds affect the shape of these NIR lines should be investigated further.

The data set presented here demonstrate that current knowledge about the formation of hydrogen lines in T Tauri stars is far from providing a detailed explanation for their characteristics and origin. Also, it provides a solid database with which model results can be compared. Hydrogen lines constitute one of the most important diagnostics available for the study of T Tauri stars and understanding their origin is of crucial importance. From a theoretical point of view, models have to simultaneously explain the near infrared lines and the Balmer lines, which as we have seen above convey different information. An observational effort to try to understand how and why the lines vary is also very important. Otherwise, we will only be trying to understand T Tauri stars and what gives rise to hydrogen emission lines from a single snapshot of what is, in reality, an ever changing system.

Acknowledgements

We thank the referee, Dr. Suzan Edwards, for insightful comments that significantly improved this paper. D.F.M. Folha acknowledges financial support from the "Subprograma Ciência e Tecnologia do $2^{\rm o}$Quadro Comunitário de Apoio''. This research has made use of the Simbad database, operated at CDS, Strasbourg, France. The United Kingdom Infrared Telescope, is operated by Joint Astronomy Centre on behalf of the U.K. Particle Physics and Astronomy Research Council.