3.8. A Really, Really Big Flow?

The results of these two case studies help to further define the relation between structure seen in redshift space to that which exists in physical space. From these studies we now know that voids are real structures devoid of mass and that walls are cold, quiescent 2D structures. We also know that virialized structures produce infall patterns and that the local velocity field is perturbed by the distribution of these virialized structures. However, we still don't know which structures really dominate the observed kinematics in the local velocity field. Adding to this confusion is the remarkable observation of Lauer and Postman (1994 - LP). In their startling paper, LP used distance estimates to elliptical galaxies to conclude that bulk motion is occurring over a scale of 15,000 km s^-1. This implies a mass inhomogeneity on a scale of 150 - 300 Mpc which is roughly 5% of the horizon size (c / H₀). This is also larger than the volume over which we presently measure H₀ which then seriously calls into question whether a precision measurement of H₀ can even be made. Unfortunately, the direction of this bulk motion as determined by LP is inconsistent with the direction of the CMB DA and this is very disconcerting as well as just plain odd. Hence, the LP result has serious implications with respect to structure formation models (Chapter 5) and the measured anisotropy of the CMB. As emphasized by Strauss et al. (1995), the LP result is largely incompatible with any existing structure formation models and if the observations are correct, then an unprecedented large scale flow has been discovered.

Given the fundamental problem potentially posed by the LP result, several groups are actively seeking to either verify or exclude it. Branchini et al. (1995) use the Abell cluster catalogs as a tracer of the mass from which they reconstruct the density field to predict peculiar velocities of clusters of galaxies over a scale of 18,000 km s^-1. Fitting these peculiar velocities to a dipole term recovers the amplitude and direction of the LG velocity which is consistent with the CMB DA and inconsistent with the LP result. However, the data don't completely rule out the LP amplitude and direction due to the large distant-dependent errors which are present in the LP sample. Furthermore, the Branchini et al. method assumes that the Abell catalog has a uniform sampling of clusters across the sky and this is a very dangerous assumption. In particular, the Southern compilation of the Abell cluster catalog relied on better photographic material than the older, Northern compilation.

Other studies have also failed to verify the original LP result. For example, the study of the Great-Wall by Dell'Antonio et al. (1996 - described above) are inconsistent with the LP bulk flow vector but only at the 75% level. Recent observations by Giovanelli et al. (1996) using relative TF distances to spiral galaxies have also failed to find significant bulk flow on the LP scale. Their results are shown in Figure 3-22 and indicate the large scale peculiar velocity field declines towards zero by redshift 6000 km s^-1. This decline to zero on this scale agrees well with previous studies (see Da Costa et al. 1996).

Figure 3-22: Peculiar velocity data from Giovanelli et al. (1996) whose sample shows that the average peculiar velocity declines to zero at observed redshift of 6000 km/s. This behavior is inconsistent with the Lauer and Postman (1994) result indicated by the +689 constant velocity line. The numbers below each data bin mark the number of galaxies in the sample.