SUPERCLUSTERS, DYNAMICS AND MODELS

GUIDO CHINCARINI, ROBERTO SCARAMELLA, AND PAOLO VETTOLANI

On very large scales, gravity seems to be the dominating force so that
the  distribution of  matter is  intimately  related to the motions on
large scales if matter is clumped as we observe in the distribution of
galaxies (see Fig. 1).

  We know,  on the  other hand,  that  the observations of  the Cosmic
Microwave Background Radiation (CMBR)  show homogeneity at all angular
scales  to a  very  high degree of    accuracy. This implies that  the
universe is homogeneous on  very large  scales, as  is assumed  by the
Freidmann-Lemaitre-Robertson-Walker cosmological model.

  These very general considerations  define to a  large extent some of
the questions that the current observational cosmology addresses. What
is the local topology of the matter distribution in the universe, that
is,   how and on what  scale  is matter clumped?  What   do we mean by
matter, baryons, exotic particles,  and invisible matter? What are the
effects that the estimated distribution of matter  has on the reported
large scale motions with  respect to the Hubble  flow  and how is  the
determination of the Hubble constant affected  by local streamings? Is
it possible to devise new methods to  estimate the mass content of the
universe? How did the largest structures, usually called superclusters
(SC), form and evolve?

A PANORAMIC VIEW

Recent redshift surveys covering very  large volumes of space indicate
that in spite of our efforts we  are still limited by  the size of the
sample, and that we do not yet observe a fair  sample of the universe;
we must   observe in  a deeper   and  more complete  way.  The spatial
distribution of  galaxies  has in recent  times  been described   as a
hierarchy of  clusters, a network   of filaments, and as  an irregular
lattice of cells and/or bubbles.

  It is important to note, though, that only clusters of galaxies, and
to a lesser extent, groups of galaxies, uncontroversially appear to be
completely  meaningful entities, that   are gravitationally  bound and
that  delimit regions of space   in which localized physical processes
are at work (e.g., x-ray diffuse luminosity). In this respect, SCs are
a much less sharply defined class of  astronomical objects, in that up
to now   they  have in  general  been  defined  by  subjective, albeit
sensible, criteria of overdensity thresholds in the number of galaxies
and/or clusters  of galaxies.  Obviously these  procedures  carry with
them a large factor of uncertainty that must be well-borne in mind. As
an  example, SC  has  been used  to  denote  fairly different systems,
although in general it refers to galaxy systems with scales of tens of
megaparsecs [1 Mpc*3x 10*  ly; we will  use  the value H*=100h  km s**
Mpc** for the Hubble constant].

  Clearly,  the actual distribution of SCs  (in a broad sense) is very
much related to the initial  conditions of the matter distribution  in
our universe: On the  very large scales the  distances, the sizes, and
the masses involved are so large  that there has  been not enough time
since  the    Big Bang  for   them   to  reach a   nonlinear  stage of
evolution. This very same fact gives us hope that we can make a direct
connection between the  initial conditions and  the present observable
large-scale structure.  Moreover, a  clearly established structure  of
SCs would argue in favor of gravity as  the main driving force for the
formation of today's structure and pose insurmountable difficulties to
other  proposed mechanisms, like the  suggestions of cosmic explosions
or   radiation-driven instabilities as  main   sources for galaxy  and
cluster formation processes.

  It is  then important  to    note  that quantitative   measures   of
clustering, especially the  two-point correlation  function (a measure
of  deviation from   a  random  distribution), seem   to indicate  the
presence over a finite distance  range of a scale-free distribution of
the  same  form but with  different  amplitudes  for both galaxies and
clusters. Clusters are more spatially correlated among themselves than
galaxies are, and it is still not clear if the same applies to SC with
respect to clusters of galaxies.

  While the   detailed  picture must  await  deeper  and more extended
observations, the fundamental fact persists  that the distribution  of
observed objects is   characterized by inhomogeneities even  on  large
scales and that, assuming that gravity is at work, such irregularities
of  the distribution  of the  mass must  perturb the  otherwise smooth
expansion of the universe. We discuss this point in the next section.

DYNAMICS

The most dense large-scale structures,  SCs, consist of large, perhaps
unbound, agglomerates of galaxies and  clusters of galaxies. A crucial
problem is to determine the mass of these overdensities. This point is
directly related to  dynamical and theoretical  considerations, and is
concerned with the problem of the behavior of the mass-to-light ratio,
M/L, when the length scale increases.

  Indeed, while we  are  able to  directly measure luminosities,  mass
determinations are more complex and  usually require a few very strong
assumptions on the dynamical state of the object under study.

  To  have a  theoretically  appealing,  spatially flat universe,  one
would need  to have much more  mass than that  obtained by summing the
mass directly seen in  galaxies through their luminous component (LB).
This then  translates   into a  very  large  value  for  the   average
mass-to-luminosity ratio:  (M/LB>=1300 (this would   mean that we  are
able to see only the fraction 1/1300 of the mass  of the universe). An
experimental value smaller  than  this  ratio  would imply  that   the
universe is open, whereas  a larger value   would indicate that  it is
closed.

  This theoretical number is   much larger than those  usually derived
from dynamical studies: For galaxies, (M/LB)=10, while for clusters of
galaxies,  (M/LB>=300, a fact which suggests  that indeed the dominant
part of the matter in the universe consists of yet unseen dark matter.
Therefore, although by increasing the  size  considered the M/L  ratio
increases, these astronomical observations suggest an open universe.

  It then becomes  crucial to determine what  happens on scales larger
than those of clusters of galaxies and whether the rising trend of M/L
continues up  to the  point  of  reaching the theoretically  appealing
value.

  The major problem is that virial theorem mass estimates are reliable
only   when they are   determined   for well-relaxed, dynamically  old
systems, while this is not the case for SCs: These have crossing times
which are  comparable or even  greater than  the age  of the universe,
resulting in  dynamically  young, unrelaxed systems.   Also, while the
x-ray luminosity  of  rich clusters gives  useful  information  on the
total  mass contained  within the cluster  radius,  this method is not
applicable to SCs, for which only upper limits  to their diffuse x-ray
luminosity are available.

  Another interesting line of approach  is that of taking advantage of
the fact that SCs  are young systems, and  therefore still in a linear
stage of evolution.

  Indeed, in linear  perturbation theory, by the continuity  equation,
there     is    a  relationship   among     a  mass   overdensity, the
gravitationally-induced  peculiar velocity with  respect to the Hubble
flow,  and the circumstance that  the universe is spatially flat, open
or closed. In  other words, when a galaxy  is in the neighborhood of a
very  large mass,   this  galaxy   is  subject to  the   gravitational
attraction of the large mass, and this  has the effect of slowing down
the cosmic expansion velocity (Hubble flow) of that particular galaxy,
which  then   has a  nonzero  peculiar  velocity.   By comparing   the
amplitudes of the expansion velocities  of several galaxies close to a
SC  with those of galaxies  at the same  distance from us which are in
more homogeneous  regions and expand  with the Hubble flow (i.e., they
have zero peculiar velocity),    in principle one  should be   able to
measure the extent  of the SC gravitational  pull, and hence  its mass
and its M/L.

  Much effort  has been devoted to apply  this technique to the infall
of our   galaxy  towards the  Virgo cluster,  the   nearest cluster of
galaxies  (see Fig. 2). There  are many practical difficulties, one of
which is trying to determine the exact amount of our peculiar velocity
towards Virgo: We  can estimate  the  total peculiar velocity  of  our
galaxy  through   the measured dipole  anisotropy  in   the CMBR. This
anisotropy is not primordial, but is a sort of  Doppler effect, due to
our  specific  motion: We see  the CMBR  sky   hotter on the direction
toward  which we are  moving,  and  colder on the  opposite  side  (an
observer   on a galaxy with  zero  peculiar velocity would not measure
such a dipole effect).

  The total peculiar velocity of the few galaxies that constitute with
our own,  the Local Group of  galaxies (LG), is  estimated through the
dipole anisotropy to  be *600 km s**, but  it is not directed  towards
Virgo. Because from linear theory one finds that the peculiar velocity
is parallel to the direction of the  net acceleration that is felt, it
follows that there are other  large mass contributors besides Virgo to
the  peculiar motion of the LG.  Because gravity falls with the square
of   distance, masses farther away  than  Virgo must  have much larger
masses than the  Virgo cluster  to  have appreciable influence on  our
peculiar velocity.

  It is also important to note that, in an expanding universe, besides
the more intuitive  pulls given by  mass concentrations (infalls), one
also has  "pushes" from underdense regions:  If a galaxy happens to be
on the  edge of a void,  it will be blown away  because voids  tend to
expand  faster than average, contrary to  dense regions, which tend to
contract.

  From various studies, the   region of the Hydra-Centaurus (H-C)   SC
emerged  recently as the  most  interesting one  for  this problem. At
first it was thought that the H-C SC  was itself at  rest, and that it
was the main one responsible  for the peculiar acceleration needed  in
addition  to  that  due     to   Virgo  to  explain   our     peculiar
velocity. However,   recent studies of  peculiar  motions  of galaxies
within 50 MPC h**(where h=H*/100  km s** Mpc**)  from the LG confirmed
earlier claims of the  presence of large peculiar velocities  coherent
on  large regions  of space-the Rubin-Ford   effect. These studies not
only pointed out distortions of the Hubble flow larger than previously
thought, and hence  the need for  even larger mass concentrations, but
also showed  that the H-C  SC was  itself  moving with respect  to the
Hubble flow, and not at rest as was once assumed (see Fig. 3).

  These  motions on very large scales  have the main  component of the
flow directed towards the  Centaurus   direction and, because  the  SC
there itself appears to be moving, must be directed at some very large
mass   behind  it.  This   hypothetical  mass  concentration  has been
nicknamed the Great Attractor (GA).  Indeed,  the Centaurus cluster is
part of an overdensity located at *35 Mpc, but is itself falling so it
cannot be the  GA.  In the same direction,  behind it there is another
concentration  of galaxies at  *45  Mpc which is   likely to produce a
strong attraction on Centaurus,  our galaxy,  and  the LG, and  on the
galaxies which show the general  motion in that direction. Therefore a
very large mass is required to explain the  whole effect. On the other
hand, we  discovered that  in  the  very same direction,  three  times
farther away  at *140 Mpc,  lies  the largest nearby  concentration of
rich  clusters  known so far (see   Fig.  4).  Is this  a coincidence?
Probably not.  A very important consequence of this alignment is that,
once determined,  the fraction of the pull  due to this extremely rich
SC as a residual from the subtraction of those due  to the more nearby
structures (including the GA) that  can be  studied more easily,  will
enable us to make the first direct measure of  the mass of a very rich
SC. This in turn  will allow measurement of the  value M/L on the very
large scales and therefore directly test  if the universe is spatially
open (more likely) flat (perhaps), or closed (less likely).

PRESENT VIEW

From what has been described earlier, it is  clear that our picture of
the universe  has changed  drastically during  the last  decade. After
previously considering highly homogeneous model with a few clusters of
galaxies  scattered around  in an otherwise   uniform  sea of galaxies
expanding uniformly and smoothly,   we have now recognized   a locally
high-structured    distribution   organized  by   gravity.  Primordial
fluctuations of the order of a  few hundredths of  the size (3000 Mpc)
of the current horizon  evolved such as  to almost reach the nonlinear
stage, with gravity playing the most important role. Irregularities in
the distribution of mass  on such scales  affect the Hubble expansion,
perturbing  the  Hubble flow. On    much larger scales  there must  be
uniformity and  the whole is embedded  in a bath of uniform radiation,
the   CMBR, rather than  in a  smooth sea of  galaxies  which has been
denied by  the   observations: Excesses  of  densities alternate  with
regions  devoid   of galaxies. What  the  transition  scale is between
inhomogeneity and uniformity has yet to be  determined. Whether or not
there is  a cellular structure  of the  universe will be  a matter for
future  researchers and  surely  enough, nature will   show us further
unexpected features.

                     Additional Reading

Bahcall, N.A.(1988). Large-scale structure in the universe indicated
   by galaxy clusters. Ann. Rev. Astron. Ap. 26 631.

Broadhurst, T.J., Ellis, R.S., Koo, D.C., and Szalay, A.S.(1990).
   Large-scale distribution of galaxies at the galactic pole.
   Nature 343 726.

Chincarini, G.(1978). Clumpy structure of the universe and the
   general field. Nature 282 515.

Chincarini, G., and Rood, H.J.(1980). The cosmic tapestry.
   Sky and Telescope 59364.

Dressler, A.(1991). The Great Attractor: Do galaxies trace the
   large-scale mass distribution?. Nature 350 391.

Dressler, A., et al.(1987). Spectroscopy and photometry of elliptical
   galaxies: A large-scale streaming motion in the local universe.
   Astrophys. J. (Lett.) 313 L37.

Gott III, J.R., et al.(1989). The topology of large-scale structure
   III: Analysis of observations. Astrophys. J. 340 625.

Oort, J.H.(1983). Superclusters. Ann. Rev. Astron. Ap. 21 373.

Postman, M., Geller, M.J., Huchra, J.P.(1988). The dynamics of
   the Corona Borealis supercluster. Astron. J. 95 267.

Rood, H.J.(1988). Voids. Ann. Rev. Astron. Ap. 26 254.

Scaramella. R., Baiesi-Pillastrini, G., Chincarini, G., Vettolani, G.,
   and Zamorani, G.(1989). A marked concentration of galaxy clusters:
   Is this the origin of large-scale motions? Nature 338 562.