Direct proof of dark matter?

Scientists claim that a violent cosmic collision provides evidence of `dark matter', the hypothetical dominant stuff of the universe.

R. RAMACHANDRAN in New Delhi

ASTRONOMERS have claimed that they may have finally seen `dark matter', the hypothetical and elusive but dominant stuff of the universe, in isolation. On August 21, scientists of a major multi-institutional astronomy experiment said that a violent cosmic collision of two large clusters of galaxies had provided the first direct evidence for the existence of `dark matter'. The remnants of the collision, collectively called the `bullet cluster' 1E 0657-56, is at a distance of about 3.4 billion light years. The clusters collided at an estimated velocity of 4,700 kilometres per second and the event itself occurred around 100 million to 200 million years ago.

The collision, according to the scientists' interpretation, caused `dark matter' to be separated from normal matter, thus allowing dark matter to be observed in isolation and providing unambiguous proof of its existence. Data from about 140 hours of observation of the `bullet cluster' by the Chandra X-ray Observatory of the National Aeronautics and Space Administration (NASA) and complementing optical observations by several other telescopes, including the Very Large Telescope (VLT) of the European Southern Observatory (ESO), the Hubble Space Telescope (HST) and the Magellan Telescope, were used to arrive at the conclusions. The team of scientists was led by Douglas Clowe of the University of Arizona, Tucson, and the findings are to be published soon in The Astrophysical Journal and The Astrophysical Journal Letters.

"This is the most energetic cosmic event that we know about," said Maxim Markevitch of the Harvard-Smithsonian Centre of Astrophysics, a member of the team, while announcing the findings at a NASA press conference. "A universe that is dominated by dark matter seems preposterous. So we wanted to test whether there were any basic flaws in our thinking," said Clowe. "The results are direct proof that dark matter exists," he added. "We had predicted the existence of dark matter for decades but now we have seen it in action. This is ground-breaking," said team member Marusa Bradac of the Kavli Institute for Particle Astrophysics and Cosmology (KIPAC) at Stanford University.

So what exactly is `dark matter'?

For nearly 70 years astronomers have lived with the observation that if the familiar Newtonian dynamics were valid in the large scale of the entire cosmos, there should be much more mass in the galaxies than what is seen by telescopes. Way back in 1937, Fritz Zwicky observed a cluster of galaxies called the Coma Cluster and measured the velocities of some of the outermost galaxies. He found that these galaxies were moving too fast to be gravitationally bound to the cluster. At such high speeds, the galaxies should be flying apart much as a rocket with a velocity greater than the `escape velocity' needed to overcome the gravitational pull of the earth.

It was argued that the gravitational pull of the galaxy cluster had to be much greater than that of the combined mass of the observed galaxies, the stellar dust and the gas to hold the cluster together. Since then many more examples have been found that need a lot of additional matter just to hold galaxies together. The "missing mass" was termed `dark matter' because it does not emit light and, therefore, not visible. And also it seems to interact with ordinary matter only through gravity.

Such dark matter is also required to give rise to structures such as galaxies and galaxy clusters a billion years after the Big Bang, which Einstein's conventional General Relativity and cosmology cannot give. If Einstein's theory of gravity (and Newtonian dynamics) are the same in the large scale, this dark matter should be five times the normal matter (consisting of atoms, neutrons, protons and electrons) that makes up the visible universe.

(The observation in recent years, that the expansion of the universe may be accelerating, has led scientists to postulate the existence of an unknown form of energy as well, called `dark energy'. It is effectively a repulsive force pushing cosmic objects apart. Nearly three-fourths of the universe is believed to be `dark energy'. This means that only about 4 per cent of the universe is made up of visible matter.) But at present what the nature of dark matter (and dark energy) is entirely a subject of theoretical speculations.

Some scientists have argued that the anomalous speeds of galaxies can be explained by modifying the theories of gravity due to Newton and Einstein, generically termed as Modified Newtonian Dynamics (MOND) theories. Variants of these are claimed to be able to explain galactic cluster data and dark energy as well. However, Clowe and associates believe that it would not be easy to explain the new results with the `bullet cluster' by modifying gravity.

Most of normal matter (that is visible to telescopes) in the galactic clusters is in the form of high-temperature gas in an ionised plasma state (about 85 per cent), which can be profiled by their X-ray emissions. Using a technique known as `gravitational lensing' - which causes light from distant galaxies to be bent by the gravitation of intervening galactic structures to form multiple images much like an optical lens - astronomers are able to map the mass distribution in galactic clusters from the extent of gravitational stretching of the lensed images.

It has been found that "dark matter" and normal matter always coexist, with their centres of gravity nearly the same, but with the "dark matter", or the gravity field to be accurate, distributed much more smoothly than ordinary matter. (In fact, this is one of the motivating factors for approaches that modify gravity to mimic dark matter.) Therefore, lensing observation with a galaxy cluster cannot tell us how much of it is owing to dark matter alone.

In a head-on collision of clusters, however, something very interesting seems to happen. Not many examples of such collisions are known, but the `bullet cluster' 1E 0657-56, which was discovered in 1995, and X-ray observations had shown it to be to be among the hottest and most luminous clusters, may be typical of such cluster-merger systems.

The galaxies themselves - being sparsely distributed and interacting only via gravity - pass through each other unimpeded by the collision. But the hot gas in each cluster is pulled back in its motion because of the drag arising from the electromagnetic forces between the plasma constituents. This forces the visible matter to separate into two components, the galaxies moving unhindered and the hot gas forming a characteristic bullet-shaped clump - hence the name `bullet cluster' - because of the combined effect of the drag and the bow shock front of the collision. The detailed X-ray emissions revealing the bullet shape were first obtained in 2000 using the Chandra Observatory. While the optical image (picture 1) shows the clusters unaffected, the X-ray image shows the gas clumps decoupled from the galaxies (picture 2).

"The snapshot Chandra observation revealed a gas `bullet' and a spectacular shock front, a textbook example of shocks and the first one ever seen in a cluster," says Markevitch, who is from the Space Research Institute of the Russian Academy of Sciences in Moscow. "We then overlaid the X-ray image on an optical image, noticed the offset between the galaxies and the gas, and realised that this cluster offered a unique experimental set-up for dark matter studies - we only needed to map its dark matter distribution. This is where our optical colleagues joined in," adds Markevitch.

Enter `gravitational lensing'.

According to Einstein's General Relativity Theory, mass warps space around it; the greater the mass, the greater the warp. So, as mentioned earlier, gravity due to mass in a galaxy cluster distorts light from background galaxies and the distortion is proportional to the mass. By looking at the shapes of many different background galaxies, the mass distribution in the cluster can be mapped.

The estimated mass of the cluster from the lens map is a billion million (10 15) times the mass of the sun. When X-ray, optical and lensing observations on 1E 0657-56 were compared, the greatest lensing seemed to be because of that part of the cluster which is spatially removed from the hot gas component. If dark matter that is supposed to account for most of the cluster mass did not exist, gravity should be centred at the plasma, which is supposed to form the bulk of the visible matter. The composite picture (picture 4) shows the aftermath of the collision clearly, with the blue blobs representing regions where most of the mass is centred (as measured by lensing) and the pink blobs (revealed by X-ray measurements) showing regions where most of normal matter is present.

Dark matter (in blue), which too interacts only gravitationally and therefore does not experience drag, passes through unimpeded along with the galaxies and is segregated from the gas (in pink). The stages during the collision are described schematically in the panel of four pictures. The observation that most of the mass appears to be centred in regions separated clearly from gaseous regions is the strongest evidence as yet that most of the universe's gravity is due to dark matter, scientists have argued.

"With conventional physics, we are certain about the detection," Clowe told Frontline by e-mail. "What we observe is exactly what we expected using conventional cold dark matter and standard physics. Also, the observations of the X-ray plasmas are telling us quite a bit about some of the finer details of such [plasma] interactions which were not previously known," Clowe added.

"With high significance we can tell that there is some form of matter that does not emit X-ray or light is involved and it is five times more in the system," Bradac told Frontline. According to her, the separations seen between the clumps were well reproduced in computer simulations of gas plus `N-body' dynamics. The scientists are also able to set upper limits to (non-gravitational) self-interaction among dark matter particles and, according to Clowe, the results are consistent with no interactions occurring.

Do the results tell us something about the nature of dark matter by ruling out some of the many proposed candidates? "Unfortunately," pointed out Clowe, "our observations are based solely on detecting the gravity of the dark matter and not detecting any light, which is pretty much the minimum definition for any substance to be a dark matter candidate. As a result, our results are consistent with a broad range of dark matter candidates. It will be hard, but not impossible, to explain the observations using only neutrinos and we can place constraints on the type of dark matter existing in the extra dimensions of brane-world scenarios [in string theory], as long as the observed self-interaction limits are satisfied. But otherwise, these observations do not rule out or favour any kind."

"Overall," points out D. Narasimha of the Tata Institute of Fundamental Research (TIFR), "cosmology has a rough picture of matter in the universe and how structures grew. No single test to prove or disprove the picture exists. However. many probes, some mild and some strong, together show that the picture is generally right. The present one is one such test."

But what about theories that attempt to modify gravity and do away with the notion of the mysterious dark matter? "In the form they are right now, the alternative gravity theories cannot explain both the spatial offset and the signal strength in lensing," pointed out Bradac. "It is doubtful that this can completely rule out the various alternative gravity theories. [But] what we expect is that when trying to do the modelling they will have to use dark matter even in such theories, and probably enough dark matter that they will have to lower the amount [by which] they are altering gravity," said Clowe.

But the proponents of modified gravity theories disagree. "I do not believe that the group observing the `bullet' cluster 1E0657-56 can claim that they have detected dark matter beyond any doubt," John W. Moffat of the University of Toronto told Frontline. "I have demonstrated that my modified gravity [MOG] can fit the lensing data without dark matter. The results clearly show that the new observational data cannot exclude a modified gravity theory," he said.

In fact, even the significance of the new results is discounted by some. "I am actually surprised by the revolutionary significance being attributed to the new observations of the bullet cluster," Jacob D. Bekenstein of The Hebrew University of Jerusalem told Frontline by e-mail. Bekenstein is known for the Tensor-Vector-Scalar (TeVeS) theory of gravity, a relativistic version of MOND he formulated in 2004. "The bullet cluster system 1E 0657-56 is, after all, a pathological case - a cosmic collision, as messy an astrophysical system as they come. Is this really the case that is going to bring MOND down?" he asked.

"Gravitational lensing is at the centre of the claims being made. MOND in its original form could not address lensing. My TeVeS could treat gravitational lensing with results consistent with what was known then," he said. The gravitational field due to matter in TeVes, Bekenstein explained, has two parts: one, a linear part, just as in General Relativity, and the other, which is sometimes dominant, a highly non-linear part generated by the same visible matter.

"The inferred gravitational field [in lensing] gives a measure of mass distribution in space. If TeVeS is right, this measure will not be spatially distributed in the same way as visible matter because of the above non-linearity. Qualitatively this is what was found by Clowe and company. They, however, immediately proceed to put dark matter with the galaxies. This is a sure way to confuse the issue.... In TeVeS, in contrast to general relativity, the relation of the convergence to the mass distribution is quite indirect; it is non-linear and non-local," he points out.

Bekenstein, however, agrees that clusters have spelt trouble for MOND in the past, requiring proponents to invoke some kind of invisible (but normal) matter, but not much more massive than visible matter as is the case with dark matter, in addition to modifying gravity. Clowe too believes that some similar theory may be able to explain the new results. "I personally would prefer a pure modified gravity solution to the mass discrepancy question," he says. "In science the best scenario for progress is when sharply opposed paradigms confront each other. Yet who is to say if an inelegant mixture of the above sort has not occurred in nature?" asks Bekenstein.