next up previous
Next: Stellar kinematics Up: Spectroscopic results Previous: The absorption calcium triplet

Gas kinematics

We have measured the systemic velocity in H1#1 by integrating all the spatial increments along the slit corresponding to the disk emission (i.e. excluding the nucleus and the circumnuclear ring of Hii regions), and fitting the two peaks of the resulting integrated profile. The mean velocity of these two peaks is 141717#174 km s-1; this corresponds to a scale of 92 pc/arcsec, for H0=75 km s-1 Mpc-1. The velocity curves obtained from the pixel to pixel measurements of the H1#1 emission line are shown in Fig. [*]; the H1#1 line flux is also plotted for reference. Allowing for the different spatial and spectral resolutions, these velocity curves agree very well with those presented by MM93. In the following we analyze the detailed and the global kinematic structures found along the three position angles.

First we confirm that PA=138 presents the largest velocity amplitude, followed by PA=84 and PA=48, indicating that 138 is closest to the major axis. At first glance, there seems to be much velocity structure at PA=48 that does not correspond to a kinematic minor axis, where we would not expect a systematic residual velocity curve. On a more detailed analysis, this velocity structure present along PA=48 is related to local morphological structure, and we can see that the outer parts of the disk velocity correspond to the systemic velocity, as indicated by a horizontal dotted line, 1417 km s-1. Thus we adopt 48 and 138 as the kinematic minor and major axes respectively.

For this value of the kinematic major axis and an inclination angle of 42 (MM93), the total dynamical mass enclosed within the circumnuclear region (a radius of 6 arcsec, equivalent to 644 pc), is 31#31 M32#32. Kohno et al. (1999) report a molecular gas mass of 33#33 M32#32 within a 6 arcsec radius; this amounts to 25% of the total dynamical mass.

Much detailed structure is present in the velocity curves at the different PA. For example, at PA=48 we detect the 50 km s-1 streaming into the main NE arm, between -56.5 arcsec and -53 arcsec (labelled A in Fig. [*]). Also at this PA we detect the 50 km s-1 streaming into the bar region at -36 to -20 arcsec in the NE (labelled B), and at 32 to 23 arcsec to the SW (labelled B). At PA=84, the 50 km s-1 streaming into the bar is seen at the eastern edge, between 40 and 30 arcsec (labelled C), and at the western edge of the bar between 39 and 32 arcsec (labelled C).

When deprojected for an inclination of 42 and a kinematic major axis of 138, the velocity curves at PA 84 and at PA 138 generally match in the outer disk regions, except for local distorsions produced by the spiral arms. However, the circumnuclear region shows a different picture. One important feature of the velocity curve at PA 48 is the existence of an apparent counter-rotation of the gas within the circumnuclear region with respect to the gas in the disk at the same PA. Indeed, when we look into the 17#178 arcsec circumnuclear velocity gradients at the three position angles more closely (insets in Fig. [*]), we see three features. First, if we deproject the circumnuclear velocity curves using the same two parameters as for the main disk, i=42 and 14#14, the deprojected amplitude at PA=84 becomes significantly larger than at PA=138; this would imply that the circumnuclear region requires a different set of deprojection angles. Second, there is a continuous change of slope between the three PA, flattening from PA=138 to PA=84 to PA=48. If the circumnuclear velocity curve along PA=48 is also produced by disk rotation, then the kinematic axis of this circumnuclear rotation must be different from the kinematic axis of the main disk. Under this assumption of disk rotation, we can compute a kinematic major axis for this circumnuclear region. We obtain that in this case the major axis of the circumnuclear region would correspond to 118; with this circumnuclear major axis and for the same[*] inclination of 42, the three deprojected circumnuclear velocity curves agree quite well. Third, at PA=84, the circumnuclear velocity amplitude, 200 km s-1, is significantly larger than the galaxy main disk velocity amplitude at the same PA, 130 km s-1. These three facts, namely, the apparent nuclear counter-rotation at PA=48 and the greater velocity amplitude at PA=84 (both, with respect to the disk amplitude at PA=84, and with respect to the deprojected circumnuclear amplitud at PA=138), argue for a gas dynamics in the circumnuclear region of NGC 6951 that is decoupled from the dynamics of the main body of the galaxy. We shall return to this point in the discussion section.

The FWHM of H1#1 and [Nii] are plotted in Fig. [*]. The values shown have been corrected for an instrumental resolution of 43.5 km s-1. The influence of the nuclear velocity gradient on the widths is negligible; the gradient is 44 km s-1 arcsec-1, that amounts to a quadratic correction on the nuclear widths of only 2 km s-1. The H1#1 line flux is shown (dotted) for reference. The velocity dispersion in the circumnuclear region is very high, in the H1#1 line there is a central plateau of 180 km s-1 in the nuclear 17#171.2 arcsec, it then decreases down to a minimum of 70 km s-1 that is reached at the maximum of H1#1 emission in the Hii regions. After this local minimum, the velocity dispersion increases again to 120 km s-1 and above in the two quadrants NE-NW and SE-SW across the ring of Hii regions. This trend is consistent with the general picture of the dynamics deduced from our data and from the molecular content (Kohno et al. 1999), that will be developed in the discussion section. The behaviour of the velocity dispersion for the [Nii] 34#346583 line is qualitatively similar to but systematically higher than that of H1#1 by a factor 1.2 in the inner 17#173 arcsec, reaching values of up to 220 km s-1 in the nucleus. These emission line widths indicate dynamically hot nuclear and circumnuclear regions. The kinematic data and the results from the emission line ratios and density values presented in section 4.1 seem to imply the existence of shocked material.

The width of H1#1 is not resolved further out in the disk of the galaxy, except across the spiral arms where it reaches values of 25 to 45 km s-1.


next up previous
Next: Stellar kinematics Up: Spectroscopic results Previous: The absorption calcium triplet
Enrique Perez
1999-09-29