Now you see it, now you don't

The 'discovery' of dark energy - invisible stuff that makes up the bulk of the Universe - was heralded as a breakthrough in 1998. New research suggests that it's just not there. Marcus Chown investigates

It's invisible, it permeates every pore of space and it is remorselessly driving the galaxies apart. The discovery of "dark energy" nearly six years ago was a bombshell dropped on the world of cosmology - and voted scientific breakthrough of 2003 by the journal Science. Now, however, an international team of astronomers is controversially claiming that it could all be a big mistake. "The dark energy could be a huge cosmic mirage," says Subir Sarkar of the University of Oxford. "It may not exist at all."

Dark energy burst on to the world stage in 1998 when a team of scientists led by physicist Saul Perlmutter at the University of California at Berkeley announced that they had discovered something odd about Type Ia supernova, a kind of exploding star thought to detonate with a standard luminosity. It seemed that supernovae very far away were fainter than they should be, given their distance from the Earth. Everything made sense if, while the light from the supernovae had been travelling across space to us, the expansion of the Universe had actually speeded up, pushing the supernova farther away than expected and making them appear fainter in telescopes.

The problem was that the galaxies - the building blocks of the Universe - were known to be flying apart from each other like pieces of cosmic shrapnel in the aftermath of a titanic explosion known as the Big Bang. The only force acting on them should be their mutual gravity, which by pulling them together should be "braking" the expansion of the Universe, not speeding it up. Remarkably, the theory went, empty space must be filled with some kind of weird anti-gravity stuff - dark energy - which was counteracting gravity and driving the galaxies apart.

It's this claim that is now in dispute. A team led by Alain Blanchard of Laboratoire d'Astrophysique de l'Observatoire Midi-Pyrenees in Toulouse has used the European Space Agency's XMM satellite to measure the X-rays emitted by ultra-hot gas in clusters of galaxies. The galaxy clusters are at very large distances from the Earth and consequently we see them as they were very far back in time.

The key thing that the observations have allowed is to be able to count the number of clusters with a particular X-ray luminosity, so that a cluster seen far back in time can be compared with a more recent cluster with the same luminosity. "This is crucial," says Sarkar, who is Blanchard's collaborator. "It allows a like-for-like comparison of the number of clusters of a given luminosity back then and in today's Universe."

What Blanchard's group has found is that nowadays there are far more galaxy clusters in a given volume of space than there were in the past. "This is at odds with what is expected in a universe with dark energy," says Sarkar.

Because dark energy is a property of space, as space expands, the amount of dark energy grows and, with it, its anti-gravity effect. Consequently, the gravity that is sucking in matter to make new galaxy clusters is eventually overwhelmed by dark energy, preventing any more galaxy clusters from being born. "If the Universe contains dark energy, therefore, the number of galaxy clusters in a given volume should have levelled off at some time, not continued increasing until the present time," says Sarkar.

Astronomers such as Cambridge's Sir Martin Rees think the behaviour of X-ray emitting gas in galaxy clusters is highly complex and that our lack of understanding of the complexities may be leading Blanchard and his colleagues to an erroneous conclusion. Taken at face value, however, the XMM results imply that there is no dark energy.

The results are compatible with us living in a high-density universe in which the total matter - visible stars and gal- axies plus invisible, or "dark", matter - adds up to close to the so-called critical density. This is a value strongly favoured by theorists because it is required by "inflation", the current best theory of what happened in the Universe's first split-second of existence.

If Blanchard's people are right, the question is: how did astronomers manage to miss all the extra matter in their surveys of the Universe? Sarkar points out that astronomers cannot see dark matter directly but must infer its existence from the motion of visible material such as stars and galaxies. "They could easily have underestimated the ratio of dark matter to ordinary matter," he says. "And this could have caused them to underestimate the total amount of dark matter."

Clearly, if there is no dark energy, then the original supernova observations which indicated the existence of dark energy were wrong. Sarkar thinks so. The basic assumption made by Perlmutter and his colleagues is that supernovae in the ancient Universe were exactly the same as supernovae today. This may be wrong. Certainly, two Indian physicists have recently drawn attention to a peculiarity of the supernova analysis. Perlmutter and his colleagues use a "correction" to the brightnesses of supernovae that are extremely far away and ones which are relatively nearby. They then deduce that the Universe's expansion has speeded up as a result of dark energy. According to Roy Choudhury of the International School for Advanced Studies in Trieste and Thanu Padmanabhan of the University of Pune, if the two categories of supernovae are analysed separately, the two sets of data can individually be explained without any need for dark energy.

Not surprisingly, the scientists who made the original supernova observations that led to the announcement of dark energy stand by their result. According to Alexei Filippenko of the University of California at Berkeley, leader of a second team at Berkeley beside Perlmutter's, "the supernova observations are really pretty convincing now".

Bizarrely, even if the supernova observations are wrong, it may not matter. According to Rees, this is because the dark energy result is now not dependent on a single observation but on a large, interlocked set of observations. One of the most important is an observation made last year by Nasa's Wilkinson Microwave Anisotropy Probe (WMAP).

The probe observed the cosmic background radiation, the relic heat of the Big Bang fireball. Tiny variations in the temperature of this "afterglow" across the sky turned out to be compatible with the Universe having the critical density, with 30 per cent of its mass in the form of matter - both visible and dark - and 70 per cent in the form of dark energy. In other words, they confirmed the existence of dark energy.

Sarkar and his colleagues, however, point out that the analysis of the probe's data was based on a number of cosmic assumptions. If any were slightly wrong, the Big Bang radiation might not be compatible with dark energy after all.

Sarkar and his colleagues, who include Michael Rowan-Robinson of London's Imperial College and Marian Douspis of the University of Oxford, need the "lumpiness" of the matter in the Big Bang, which led to the temperature variations seen by the probe, to be slightly different from the standard assumption.

They also require that 12 per cent of the mass of the Universe be in the form of ghostly particles called neutrinos. "We know from experiments that neutrinos have mass," says Sarkar. "The only thing not known is whether they have sufficient mass for our purposes."

Critics such as David Spergel of Princeton, who has worked on the WMAP data, point out that a lot of things need to be wrong for observations of the Universe to make sense without dark energy. "It is better to be slightly wrong about a number of things than incorporate a parameter which is 10 followed by 123 zeroes bigger than theory predicts!" counters Sarkar.

This is a reference to the fact that our best theory of physics - quantum theory - predicts a value for the energy density of the dark energy which is 10-followed-by-123-zeroes times bigger than observed. This spectacular discrepancy has been described by Nobel laureate Steven Weinberg of the University of Texas as "the worst failure of an order-of-magnitude estimate in the history of science".

Certainly, if the dark energy goes away, physicists will sleep easier in their beds. Nature, however, is under no obligation to respect the sensibilities of physicists or anyone else. Ultimately, it will be further observations of the Universe that determine whether dark energy is real or one of the biggest mirages in the history of science.

Marcus Chown is the author of 'The Universe Next Door', published by Headline (£7.99)