We have obtained the gas kinematics of NGC 2992 by means of long slit data along nine position angles. The results are in general agreement with previous determinations (Heckman et al. 1981, who presented data along PA= with higher spectral resolution but lower spatial resolution; Colina et al. 1987, who presented data of lower spectral and spatial resolution along 3 PAs; Keel 1996, for data with a resolution similar to ours along PA ).
We have modeled the kinematics of NGC 2992 by circular rotation in a gaseous disk to which is added constant radial outflow in the disk plane, as suggested by line splitting. Disk rotation can be accounted for with the following parameters: major axis along , inclination , velocity amplitude =250 km s-1, and parameters p =1.1 and =5 arcsec (see Sect. 3). This value of , which allows a fairly good representation of the observed velocity field in the central regions, results to be different from that of the large scale disk as derived by continuum images. This discrepancy could imply that the disk of NGC 2992 is warped, probably due to interaction with NGC 2993. Outflow was modeled along a triangular region of axis PA= , with an opening angle of and a constant velocity of 150 km s-1 along the outflow region (measured 5 arcsec from the nucleus). Opening the angle of the outflow to in the east could give a better fit only for PA= in the northeast, in the region where H emission is detected by Wehrle & Morris (1988).
It can be noted that low excitation gas follows better the pure rotation model in some regions, whereas high excitation gas is better represented when outflow is also included in the model. In our scenario, the outflow takes place radially close to the plane of the gaseous disk, with a spatial extension which is much larger on the east than on the west side.
This simple model accounts rather well for the kinematics of NGC 2992, if one excepts the regions located more than 6 arcsec northwest of the nucleus. The kinematics of the gas in that region confirm results by Heckman et al. (1981) and Colina et al. (1987), who reported large velocities and velocity dispersions. As noticed before, these regions with the largest discrepancies may be understood in terms of more complex dynamics and structure, and/or by taking into account the possible line asymmetries due to the dust distribution. Since our spectral resolution is not sufficient to observe line splitting in high excitation lines, and we don't have line ratios, we cannot constrain a more sophisticated model.
© European Southern Observatory (ESO) 1998
Online publication: April 20, 1998