B00B7H7M2E EBOK (39 page)

Wendy Freedman is a native of Toronto, Canada, and one of her most vivid childhood memories is of a trip with her father
to
northern Canada, where they watched the stars and he explained to her how long it takes their light to reach us on the Earth. When Freedman entered the University of Toronto in 1975 she intended to study biophysics, but she soon switched to astronomy. She got her doctorate in astronomy and astrophysics from Toronto in 1984, then received a Carnegie Fellowship at the Carnegie Observatories, and in 1987 was the first woman to join the permanent faculty there, where she remains to this day.

At the heart of Freedman’s Extragalactic Distance Scale Key Project lies the effort to measure Cepheid distances to 20 galaxies with the Hubble telescope. It is these distances that are then expected to provide an absolute scale for other methods which give only relative distances (Type Ia supernovae, Type II supernovae, the Tully-Fisher relation, and surface-brightness fluctuations –
see here
).

In 1994, Freedman and her team were attempting to measure more precisely the distance to the centre of the Virgo supercluster. They found twenty Cepheids in the spiral galaxy M100 in the Virgo cluster, at the core of the supercluster, the first sure identification of Cepheids that far away. The Hubble data indicated that these Cepheids are approximately 56 million light years from earth. That was nearer than earlier estimates put the centre of the Virgo supercluster.

From this new distance measurement and M100’s recession velocity (learned from its red shift), Freedman and her colleagues calculated a new value for the Hubble constant, about 80 kilometres per second per megaparsec. Experts led by Allan Sandage had previously calculated that its value was about 50, nowhere near 80. Thus began one of the most heated debates in modern astronomy – either an extremely significant controversy or much media hype about nothing, depending on which side of the issue you stand.

Freedman’s announcement came as a shock. The Hubble findings were an embarrassment. Deciding between values of
50
and 80 was not mere nit-picking: if the universe is expanding so much more rapidly than previously thought, it follows that less time has elapsed since the Big Bang than the 10 to 20 billion years most experts had settled on. Depending on the density of matter in the universe, a Hubble constant of 80 means the universe must be only eight to twelve billion years old, probably nearer eight.

It isn’t uncommon in science for new findings to challenge earlier thinking, sometimes eventually undermining what nearly everyone has been assuming was virtually unassailable scientific knowledge. But this challenge was one of the most disquieting so far in the 20th century, for astronomers were fairly certain, based on what they considered sound understanding of nuclear physics and the rate at which hydrogen converts to helium in stars, that some of the oldest stars in the Milky Way are 14 billion years old, probably even older. The universe can’t be younger than the stars in it.

The glitch made newspaper front pages all over the world. Scientists ground their teeth. Since the future of astrophysics and astronomy depends on massive public spending, these branches of science have an enormous stake in maintaining their credibility. Researchers wonder what will motivate the allocation of funds for science now that the old Cold War rivalry is history. If public opinion is to favour continuing support for this wondrous but expensive adventure, it really doesn’t do to have announcements indicating that tax money is buying nonsense! Like the Church in Galileo’s day, the modern scientific establishment has much to lose if simple faith – in science this time round – is undermined.

Freedman and her team were young astronomers. Those whose numbers they were questioning were some of the most highly – and deservedly – respected older members of the astronomy community. Sandage had spent the best part of a lifetime developing new measuring techniques and making careful observations to arrive at the Hubble constant value of 50. But
iconoclastic
as the team’s announcement was, it didn’t come entirely out of the blue, nor was such a conundrum unprecedented. In 1929, Hubble himself calculated the Hubble constant to be 500 kilometres per second per megaparsec, making the universe younger than geologists knew the Earth was. Baade refigured H
_o
at 250. Sandage reduced it still further to 180, then to 75, then (in the mid-1970s) to 55 plus or minus 10 per cent. These corrections succeeded in making the universe old enough to allow for the formation of even the most ancient stars and globular clusters, but not before discussions had taken place that resembled those that now followed Freedman’s announcement. Nor had Sandage and Tammann’s value for the Hubble constant previously gone unchallenged.

In the late 1970s and 1980s, when Sandage had settled on a Hubble constant close to 50 and an age of the universe of 15 to 20 billion years, Gerard de Vaucouleurs of the University of Texas, for one, took serious issue with those numbers. Shortly before Freedman made her findings public in October 1994 there were other studies whose results implied that current estimates of the universe’s expansion rate and age might be headed for another revision. A team led by Robert Kirschner of the Harvard-Smithsonian Center for Astrophysics, using the Cerro Tololo Inter-American Observatory in Chile, measured the expanding debris from five supernovae and judged the universe might be from 9 to 14 billion years old. But Freedman’s team’s calculations, based on data from the Hubble telescope, were more convincing than any of these other challenges to the older numbers.

Astronomers leave a wide margin for error in calculations like these, and the immediate temptation is to wonder whether the numbers are sufficiently fuzzy to allow the universe to be just barely old enough and the oldest stars just barely young enough. However, stars didn’t pop into existence the instant the universe began. Estimating that stars are the
same
age as the universe would be unsatisfactory. There must be a cushion of at
least
a billion years after the beginning to leave comfortable time for them to form. The leeway in Freedman’s numbers and in current estimates of the age of the oldest stars is not enough. For reference: a Hubble constant of around 50 indicates an age for the universe of around 15 billion years; a Hubble constant of around 70 or 80, a much younger universe – about 10 billion years or less.

A negative reaction to Freedman’s team’s announcement came almost immediately, and not unexpectedly, from Sandage, whose office was right down the hall from Freedman’s at the Carnegie Observatories. Sandage had served under Hubble himself in this same observatory when it was known as Mount Wilson. According to Sandage, the glitch was being grossly overpublicized and its importance exaggerated – mostly media hype. There were plenty of possibilities of error in the Hubble team’s results. In their measurements of the apparent magnitudes of the Cepheids, for instance, and in their assumption that the galaxy where these Cepheids are is actually in the centre of the Virgo supercluster. Perhaps instead it is in the foreground, nearer our own Galaxy. Arguing for that is the fact that M100 is a spiral galaxy, and it is elliptical galaxies, not spirals, that are more commonly found in the centres of clusters like Virgo. Freedman countered that her team had already taken that possibility into account in assigning a wide margin of error to their estimate. What’s more, as they had also reported in their original paper, the relationship of Virgo to a more distant cluster, the Coma cluster, had made it possible to step out to there and calculate the Hubble constant at that distance – a calculation that gave the same result.

The question also arose whether the rate at which Virgo is moving away is a dependable indicator of the recession rate of the universe as a whole. Tammann reported that his studies indicated that Virgo is actually moving away more rapidly than the rest of the universe. Here, again, was the perennial difficulty of sorting out the actual ‘Hubble flow’ from all the other
movement
that’s going on among and within clusters and superclusters. How to extract from this complicated picture the part of all that motion that is directly attributable to the expansion of the universe? Any sample of the universe is likely to give a faulty reading unless it is an extremely large sample indeed. No one knows for certain how large a sample that would have to be.

Freedman and her team hadn’t claimed that their result settled once and for all the value of the Hubble constant, but neither were they convinced by the opposition. Sandage’s own measurements had recently been challenged on another front. He had been using Type Ia supernovae for making his distance measurements. Some of the measurements that gave Sandage and his collaborators Hubble constants of around 50 were based on the assumption that these all reach the same maximum brightness and are good standard candles. However, Mark M. Phillips, an astronomer at the Cerro Tololo Inter-American Observatory in Chile, had recently found that not all Type Ia supernovae do have the same brightness characteristics. Brighter ones appeared to occur in spiral galaxies or galaxies with many bright stars. Phillips had developed a technique for analysing the light curves (how the supernova brightens and dims) to recognize these differences and make allowances for them, but, as of 1994, Sandage had not corrected his data. Ominously for Sandage, Robert Kirschner and his colleagues at the Harvard-Smithsonian Center for Astrophysics
had
corrected theirs, and their estimation for the Hubble constant had risen from 55 to around 67.

At the American Astronomical Society meeting in January 1995, just two months after Freedman’s announcement, the new measurements and the controversy they’d stirred up about the value of the Hubble constant and the age of the universe were centre stage. Mark Phillips and Mario Humay, Phillips’s colleague in Chile, were there reporting their measurement of 25 supernovae, some as far away as one billion light years.
Compensating
for the differences in maximum brightness that Phillips had discovered, they’d arrived at a Hubble constant of 60 to 70, in the middle range between Sandage’s measurements and those of Freedman’s team. Freedman announced that the Hubble telescope had now measured distances to 40 more Cepheids in M100 and distances to two other galaxies in the Virgo cluster, M101 and NGC925. The new data were consistent with her team’s earlier results.

Eight months passed, and in September 1995, Nial Tanvir at Cambridge University, with colleagues at Durham University in England and the Space Telescope Science Institute in Baltimore, estimated – based on fresh Hubble observations of Cepheids – a distance of 38 million light years to the M96 galaxy, in the direction of the constellation Leo. From this they inferred a distance to the much more remote Coma cluster. The team’s calculation gave the universe an age of 9.5 billion years, give or take a billion.

In March of 1996, a year and a half after Freedman’s initial announcement, Sandage and colleagues had rallied and were ready to report the results of their ongoing supernova study. Sandage, whistling into the wind, it seems, told an interviewer, ‘We believe that this marks the end of the “Hubble wars”.’ In 1990 there had been a Type Ia supernova in the galaxy NGC4639. Sandage’s team had been observing the light curve of this supernova since 1992. What was particularly significant about this investigation was that the Hubble telescope had been able to see individual stars in this same galaxy and 20 of them were Cepheids. From their brightness, Sandage’s group had been able to calculate the distance of the NGC4639 galaxy as 82 million light years, and from that they knew the absolute magnitude of the Type Ia supernova in a way that didn’t depend on comparing that supernova’s brightness with the brightness of other Type Ia’s. Applying this fresh knowledge to previous measurements of the apparent peak brightnesses of six other Type Ia supernovae, Sandage recalculated their distances and
came
up with a Hubble constant of 57, in the range he had been insisting on all along.

Sandage considered the case closed. Freedman and her colleagues were not won over. Sandage’s new results, compelling as they seemed, didn’t make the team’s own Hubble findings go away or point up any flaw in their calculations. The Freedman team’s numbers did come down a little. At the Princeton ‘250th Birthday Conference’ in June 1996, she reported a value of 73, based on a combination of the distance measurements to Cepheids, the study of Type Ia and Type II supernovae, the Tully-Fisher relation and surface-brightness fluctuations. Freedman said that with so much data accumulating so rapidly, the debate would be settled in the next three years. Sandage was sure that it will finally be settled close to his own number, but perhaps not until well into the next century, too late for him personally to enjoy his victory. He has said of de Vaucouleurs’s death in 1995 (the de Vaucouleurs who was his critic), ‘Very unfortunate. Anybody in the middle of a crisis should live to see the resolution of the crisis.’

What about the age of stars? There has been less controversy about that than there has been about the age of the universe. In the winter of 1996, study of some of the most distant galaxies ever observed showed them to be as much as 14 billion years old, shoring up faith in earlier calculations. However, in the late summer and fall of 1997, physicists at Case Western University in Ohio, led by Lawrence M. Krauss, re-examined the age of some of the oldest and most distant stars, using measurements from the Hipparcos satellite. They recalculated the age of globular clusters previously thought to be as much as 15 billion years old or even older. Hipparcos’s measurements, made with unprecedented precision, revealed that these globular clusters are further away than earlier estimates had put them, and so, in order for them to appear as bright as they do, they must also be brighter than previously thought. And if they are brighter, that means they are burning faster and have evolved more quickly,
making
them younger – perhaps 11 billion rather than 15 billion years old.