Science: Older than the stars

Scientists have long struggled to date the Universe. Now they've reached a final conclusion.

Since last year, there has been overwhelming cosmological evidence that the Universe is a couple of billion years older than the ages of stars, when calculated from astronomical evidence.

That seems logical enough, but in the mid-Nineties apparent evidence that the Universe is younger than the stars it contains made headlines, providing scope for stories poking fun at the supposed stupidity (or incompetence) of astronomers. In the last couple of years this crisis in cosmology has been resolved. Yet although the crisis itself was big news, the resolution of the crisis has passed almost unnoticed.

Astronomers measure the age of the Universe by measuring how fast the galaxies - systems like our Milky Way, each containing hundreds of billions of stars - are receding from each other. The faster the galaxies are moving apart, the less time it must have taken them to get to their present state since the Big Bang. The snag is that you have to relate the speeds (which are quite easy to measure, using the red shift in the light from galaxies) to the distances between the galaxies today (which is quite hard to measure). This relationship is measured in terms of a number called Hubble's Constant, {\i H}, named after Edwin Hubble, who discovered the expansion 70 years ago. A bigger value of {\i H} implies a smaller age for the Universe.

Hubble's own estimates of {\iH} were so big that they implied that the Universe was born less than 2 billion years ago. This was embarrassing, since even in the Thirties geologists knew that the Earth itself was roughly twice that old. So it was a major landmark in cosmology when Walter Baade found that Hubble's data had been miscalibrated: all the galaxies were twice as far apart as Hubble had thought. This meant that they must have taken twice as long to get to where they are today. As the headlines of 1952 put it, the age of the Universe was doubled overnight, happily making it a little more than the age of the Earth.

At the time, there was no conflict with the estimated ages of stars, since astrophysicists had only the vaguest idea how old the stars were. But over the next few decades, as successive improvements in telescopes and observing techniques pushed the measured distances to the galaxies out even further, and the age of the Universe up towards 10 billion years, the stars were also becoming better understood, and estimates of their ages (at least of the oldest stars in the Milky Way) were looking uncomfortably large.

By the start of the Nineties, when the Hubble Space Telescope (HST) began to play a part in improving cosmological observations, estimates of the age of the Universe hovered around 10-12 billion years, but estimates of the ages of the oldest stars were about 16 billion years. This was the cause of the headlines, some simply poking fun at astronomers, others suggesting that the whole Big Bang theory of the origin of the Universe might be wrong.

Now that gap has been closed from either side. The really important feature of the cosmological work is that several completely different ways of measuring the expansion rate of the Universe all point to the same answer. Hubble's original technique was based on measuring distances to nearby galaxies, by studying stars known as cepheids in those galaxies. Their brightnesses are compared with cepheids in our own Milky Way, and the comparative brightnesses give you the distances to the galaxies. Then, you have to compare properties (such as size and brightness) of nearby galaxies with the properties of more distant galaxies to estimate how far apart they are. This traditional approach has been extended by the HST, which can measure more cepheid distances to more galaxies than any previous instrument.

At the same time, distances were being measured by comparing the brightnesses of supernovae in different galaxies. A supernova marks the explosive death of a star, when it briefly shines as brightly as a whole galaxy of ordinary stars. So supernovae can be seen a long way away. All kinds of supernovae have the same brightness, so their distances can be found simply by measuring how faint they look in our telescopes. But this technique still depends on calibrating the brightnesses of nearby supernovae using cepheids.

Two other techniques do away with cepheids altogether. One involves gravitational lensing - the way the gravity of whole clusters of galaxies bends light from a galaxy far away across the Universe to form a distorted image in our telescopes. The geometry of the lensing can, under some circumstances, give the distance to the galaxy being lensed. And there is another technique that measures the Hubble constant from the effect that hot gas in clusters of galaxies has on the microwave background radiation, the echo of the Big Bang, passing through it.

Crucially, all four techniques give a similar value for the constant, and this implies that the minimum age of the Universe (the time since the Big Bang) is 13 billion years. It was the Bellman, in The Hunting of the Snark, who said: "What I tell you three times is true." The Universe has told us four times that it is at least 13 billion years old. It is true.

If astrophysicists were still telling us that the oldest stars are 16 billion years old, this would still be embarrassing. But at the same time that the new results were coming in, data from the Hipparcos satellite was changing our understanding of old stars. Hipparcos measured distances to stars with unprecedented precision. Clear of the blurring effects of the Earth's atmosphere, it located stars in the sky to an accuracy of one-thousandth of a second of arc, equivalent to being able to measure the size of a golf ball at a distance of 3,000 miles. A key result of this work was that it shifted the measured distances of the oldest stars out slightly. If the stars are further away from us, they must be shining more brightly than astrophysicists used to think, in order to look as bright as they do in the sky. And if they are shining more brightly, they must have used up their nuclear fuel (converting hydrogen into helium, like the process which powers a hydrogen bomb) more quickly. So they must be younger than had been thought. The best estimates of the ages of the oldest stars - the oldest known objects in the Universe - now range from 10 to 12 billion years, comfortably less than the "new" age of the Universe.

There is even icing on the cake. Recently, supernova studies have suggested that the rate at which the Universe is expanding is increasing. Given the way the tide of opinion now flows, this has been presented as another crisis for cosmology. It is nothing of the kind. The effect can be explained in terms of a parameter, called the cosmological constant, which Albert Einstein himself introduced into the discussion more than 80 years ago. And if the Universe is expanding a little faster than it used to, it was expanding more slowly before, and took even longer to get to its present state. We have to increase the age of the Universe by another 10 per cent or so, to a minimum of 14 billion years. The astrophysicists agree that the 2 billion years that allow for the oldest stars around today to have formed after the Big Bang, are ample. If anything in science is a Deep Truth, this match between astrophysics and cosmology is a prime candidate. It should be shouted from the rooftops.

Dr John Gribbin is a visiting Fellow in astronomy at the University of Sussex and author of `The Birth of Time' (Weidenfeld & Nicolson). His lecture on How We Measured the Age of the Universe is to be given the Royal Institution, 21 Albemarle Street, London W1, on Wednesday 17 March at 7.30pm