Annu. Rev. Astron. Astrophys. 1978. 16: 103-39 Copyright © 1978 by Annual Reviews. All rights reserved

4.5. Warps in the Outer Parts of Spiral Galaxies

It has been known for some time that the HI layer in the outer parts of our Galaxy (R 12 kpc) is warped to large distances (Z 2 kpc) from the plane defined by the inner regions (see e.g. Verschuur 1975 and references therein). A tidal interaction model involving a close passage of the LMC has been proposed by Hunter & Toomre (1969), but the situation has become more complicated since the discovery of the Magellanic Stream (Mathewson et al. 1974).

From single-dish measurements Roberts & Whitehurst (1975) showed that the HI in the outer southern parts of M31 is also warped to distances of 5 kpc out of the conventionally defined plane. Radio-synthesis techniques are less sensitive to the extended galactic HI, which confuses the single-dish observations in the northern part of M31, and the recent synthesis observations by Newton & Emerson (1977) have confirmed the existence of a warp there in the opposite direction to that found on the southern side of the galaxy. They have also fitted a model to the HI distribution and velocity field involving concentric annular rings that are progressively more inclined at larger radial distances, in the manner first proposed for M83 by Rogstad et al. (1974, see below). The maximum deviation from the centrally defined plane is 6 kpc at about 30 kpc from from the center. Roberts and Whitehurst concluded that in their model for the bend noncircular motions associated with the deviations are likely to be so small that they do not seriously affect the rotation curve determined from conventional methods even in the warp region. However, changes in the rotation velocity of only 10 km/sec can result in drastic changes in the inferred M/L ratio (Baldwin 1975).

Lewis (1968) noted that the velocity field in the relatively extensive HI distribution of M83 indicates a change of the apparent major axis with radius. The HI-synthesis observations of Rogstad et al. (1974) confirmed this feature in the velocities (Figure 7), and these authors proposed a detailed model for the kinematics: The HI distribution is represented by annular rings in circular motion, progressively more inclined at larger radii. With a reasonable choice of the viewing geometry, their model could also represent bright ridges found in the HI surface-density distribution in the outer parts of the galaxy as effects of longer path lengths through the highly inclined annuli. Rogstad et al. also noted an important dynamical property of the inclined rings: owing to the torque exerted by the inner part of the galaxy, the rings are expected to slowly precess. This implies an age of no more than about 2 × 10⁹ years for the disturbance. A similar model with rather severe warping has been proposed for M33 by Rogstad et al. (1976) as a method of explaining the wing-like features to the north and south of the galaxy and the occurrence of double profiles in certain parts of the disk.

Figure 7. The velocity field of M83 derived from HI aperture synthesis observations shows a strong variation of the position angle of the kinematical major axis with distance from the center. The underlying optical picture from the Palomar Sky Survey shows that the horizontal axis of the figure corresponds approximately to the major axis of the optical image of the galaxy. The distortions in the velocity field are interpreted as a result of a severe warping of the outer parts of the plane of the galaxy (Rogstad et al. 1974).

The feature of a changing kinematical major axis with increasing radius has been noted for several other galaxies by Bosma (1978a, b). He suggests that the effects are distinguishable from oval distortions in that the major axis agrees with that of the optical image in the inner parts of the galaxy (compare Figures 4 and 7).

A more direct way to look for warps is to study edge-on galaxies. Sancisi (1976) has found warps in the HI distributions of four out of five systems observed (NGC 5907, 4565, 4244, and 4631); the only exception is NGC 891 (Sancisi et al. 1974).

What evidence is there that the warps found in the HI data have counterparts in the stellar disks? The photometry of NGC 4565 (see H. Spinrad's results in Frankston & Schild 1976, Bertola & di Tullio 1976) shows some indication of a warp in the far outer parts of the optical image in a way consistent with the even more distant HI warp, but the evidence is not very clear. Features found in the outer parts away from the centrally defined planes of NGC 5907 and NGC 4244 could simply be isolated bright clusters of stars and HII regions. We must conclude that the optical counterparts to the HI warps have not yet been convincingly identified. In this respect it is perhaps useful to note that the HI warps in the edge-on galaxies mentioned above do not develop gradually at ever-increasing radial distances, but instead appear abruptly at the edge of the optical disks as seen on standard photographs (e.g. Palomar Sky Survey).

The origin of these warps is not yet clear. The tidal interaction picture of Hunter and Toomre is surely an obvious one to try; however, several difficulties remain in the models of M83 and M33, perhaps the main one being the relatively short time scale (of the order of 10⁹ yrs) for disruption of the warp due to differential precession of the rings. A more devastating counter-argument is the simple observation that more than half of the galaxies with evidence for HI warps in the presently available sample do not have any obvious nearby neighbours with which to interact in the time scale available. An explanation of the warps seen in the edge-on galaxies in terms of very distant spiral arms coupled with a slight tilt of the plane with respect to the line of sight (Byrd 1977) is probably inconsistent with the observed HI velocities. Interaction with an intergalactic wind (Kahn & Woltjer 1959) or gas cloud (Sancisi 1976) remains a possibility.

Finally, with respect to the suggestion that the inner parts of NGC 5383 are warped (Peterson et al. 1978) we make the following remark: the HI evidence for an extensive underlying disk in this galaxy (Allen et al. 1973a) and its optical structure (van der Kruit & Bosma 1978a) indicate that the optical velocity field of Peterson et al. in the region of the bar should probably be interpreted instead as streaming of the gas along the bar.