NEWS & VIEWS
ASTROPHYSICS
Dark matter is real
The combined data from four systems of telescopes offer the strongest
evidence yet that a modification of gravity cannot do away with the need for
dark matter.
SEAN CARROLL
is in the Department of Physics, California Institute of
Technology, Pasadena, California, USA.
e-mail: [email protected]
t’s hard to decide whether cosmologists’ belief in
the existence of dark matter represents hubris or
humility. Is it overly bold to believe in a substance
we’ve never directly seen? Or is it humble to recognize
that most of the matter in the Universe is something
different from the stuff with which we are familiar?
Either way, the notion of dark matter has served
cosmologists well. In the 1930s, Fritz Zwicky argued1
that the visible matter in the Coma cluster of galaxies
fell far short of the dynamical mass implied by the
velocities of the cluster’s constituent galaxies. This
conclusion was greatly strengthened in the 1970s by
Vera Rubin2, who measured the velocity with which
gas orbited around individual galaxies. These days,
we have multiple lines of evidence that all point in
the same direction, including observations of the
cosmic microwave background, the distribution of
large-scale structure, galactic dynamics, gravitational
lensing and hot X-ray-emitting gas in clusters, to
name a few.
The need for dark matter would seem to be firmly
established, save for a persistent loophole. All of
our evidence for dark matter so far is indirect — we
don’t see the dark matter itself, we only measure its
gravitational field. To make the leap from an observed
gravitational field to an amount of matter, we need
to assume that we understand how gravity works; in
particular, that Einstein’s general relativity is valid on
the scales of galaxies and clusters. But it is certainly
conceivable that gravity is modified on scales much
larger than the Solar System. If that were the case
— and models along those lines have certainly been
proposed, most notably in Moti Milgrom’s modified
newtonian dynamics (MOND)3,4 — it may be that we
have been tricked into believing in dark matter, when
gravity is actually to blame.
The difficulty in distinguishing between the
dark-matter and modified-gravity hypotheses is that
ordinary matter and dark matter tend to accumulate
in the same places in the Universe, at least over
sufficiently large scales. Thus, wherever we found a
sufficiently large concentration of ordinary matter,
dark matter would be there along with it, giving rise to
I
Figure 1 Dark matter in the Bullet Cluster. The Bullet Cluster formed through the collision of two
clusters of galaxies; the galaxies (orange and white) are shown in an optical image from Magellan and
the Hubble Space Telescope. ‘Normal’, or baryonic, matter makes up the hot gas (in pink) detected
by Chandra. But the ‘lensing map’ drawn up using data on gravitational lensing (from Magellan and
European Space Observatory telescopes at Paranal) shows that the concentration of mass in the
cluster (blue) is separate from the normal matter. This concentration of mass outside the location of the
ordinary matter is strong evidence of dark matter.
a stronger gravitational field than we would otherwise
expect — exactly as if gravity itself were modified.
What we would like is a cluster of galaxies in which
the ordinary and dark matter have somehow been
separated from each other, so that the gravitational
field (due primarily to dark matter) would emanate
from a different direction from the location of the
ordinary matter.
The Universe has kindly provided us with
precisely the kind of situation we are looking for,
in the form of the Bullet Cluster, 1E 0657−56. This
system is really two clusters, which have collided
recently (cosmologically speaking). When the two
clusters met, the gas in each interacted with that in the
other, producing a shock front that shows up vividly
in images from the Chandra X-ray satellite (Fig. 1).
653
nature physics | VOL 2 | OCTOBER 2006 | www.nature.com/naturephysics
©2006 Nature Publishing Group
NEWS & VIEWS
The two concentrations of dark matter, meanwhile,
should pass right through each other.
Fortunately, we have a way of determining where
the dark matter is, using the lensing of galaxies
behind the cluster. Clowe et al.5, in work to appear
in the Astrophysical Journal, have done exactly this.
Using data from the Hubble Space Telescope, the
Magellan telescopes and European Space Observatory
telescopes in Chile, and from the Chandra X-ray
observatory, they have constructed detailed maps of
the gravitational field, which can be compared directly
with the location of the ordinary matter as revealed by
the X-ray observations.
Lo and behold, the gravitational field is not
pointing towards the ordinary matter; instead, it
points towards two separate concentrations, exactly
where we would expect the dark matter to be. This
is the most vivid example yet of a gravitational field
produced by dark matter without a corresponding
amount of ordinary matter; it establishes beyond
reasonable doubt that there really is dark matter, not
simply a modified gravitational-force law.
What are the implications of such a finding? We
can be confident that the dark matter is not simply
ordinary matter that is for some reason invisible. In
ordinary matter, almost all of the mass is in the form
of baryons — protons and neutrons. But we have strict
upper limits on the number of baryons in the Universe
from the abundances of light elements formed during
primordial nucleosynthesis, and independently from
the patterns of temperature fluctuations in the cosmic
microwave background. Given these data, there is no
way for ordinary baryons to account for all the mass
implied by the dynamics of galaxies and clusters.
The dark matter, then, must be some new kind of
particle. The search is on to produce such particles in
654
Earth-based experiments, as well as to detect directly
the existing cosmological dark-matter particles. There
is also an active theoretical programme devoted to
the construction of models for dark matter, many
of which involve supersymmetry or other marked
extensions of known physics. Whatever it is, dark
matter is something outside the heretofore successful
Standard Model of particle physics.
However, the fact that dark matter exists does not
imply that gravity is not modified. Strictly speaking,
it is difficult to cleanly distinguish between ‘modified
gravity’ and ‘dark matter’. Modifications of general
relativity generically involve the introduction of
new fields, with their own energy densities, just
as with dark matter particles. To the extent that
there is a difference, it is that dark matter is freely
propagating, whereas the fields in a modified theory
of gravity are presumably tied to their ordinary-matter
sources, analogous to the distinction between freely
propagating photons and the electric field produced
by a distribution of charge.
The lesson of the Bullet Cluster is that there are
appreciable gravitational fields in the Universe that
point in a direction other than where the ordinary
matter is; regardless of the details, the source for those
fields is worthy of the name ‘dark matter’. Whether
it represents hubris or humility, we can state with
confidence that we are made of completely different
stuff from most of the matter in the Universe.
REFERENCES
1.
2.
3.
4.
5.
Zwicky, F. Helv. Phys. Acta 6, 110–127 (1933).
Sofue, Y. & Rubin, V. Annu. Rev. Astron. Astrophys. 39, 137–174 (2001).
Milgrom, M. Astrophys. J. 270, 365–370 (1983).
Bekenstein, J. Phys. Rev. D 70, 083509 (2004).
Clowe, D. et al. Astrophys. J. (in the press); http://arxiv.org/abs/astroph/0608407 (2006).
nature physics | VOL 2 | OCTOBER 2006 | www.nature.com/naturephysics
©2006 Nature Publishing Group