The most astonishing thing about COBE’s discovery of hot spots in the microwave background was how readily most scientists came to believe in the result. ‘It was a terribly difficult measurement and the COBE team made only a marginal detection,’ says Jim Peebles.
So were the hot spots real? In the past, experimenters who had tried to measure the coldest thing in the Universe had mistakenly measured stray radiation coming from the Earth or the Galaxy, from their own equipment, the exhausts of their rockets and countless other spurious sources. So were COBE’s hot spots really imprinted on the radiation from the beginning of time, or had the COBE team been hoodwinked by something altogether more mundane and closer to home?
‘I still wake up in the middle of the night thinking, “Have we accounted for everything?”,’ says Dave Wilkinson. ‘That’s my biggest worry about COBE – that we didn’t measure everything we should.’
One thing Wilkinson worries about is that stray radiation from the Earth may have got into the sensitive instruments by bending round the satellite’s ground shield. The COBE scientists were unable to measure this effect and instead had to estimate it from theory. Their calculations assumed, for the sake of simplicity, that the metal shield had a knife-sharp edge, but, as Wilkinson points out, the edge if looked at closely was bound to be ragged. The unanswered question is whether by oversimplifying the calculation the team underestimated the amount of Earth radiation the COBE instruments were seeing.
Charles Bennett had another worry. He was concerned about spurious radiation coming from our Galaxy, and whether the team had subtracted it from their signal correctly. ‘I dedicated myself to satisfying myself that that wasn’t what we were seeing,’ says Bennett.
The radiation given out by the Milky Way is complicated. It comes from glowing dust and also from electrons broadcasting radio waves as they spiral around the Galaxy’s magnetic field lines.1 The team had to have a theoretical ‘model’ of how the different types of radiation should vary with wavelength, and then it had to make sure its own measurements agreed with it. These were made at just three wavelengths, hence Bennett’s worry. But in the end he satisfied himself that all was okay. ‘Whatever model you pick for the Milky Way, it accounts for virtually none of our signal,’ says Bennett. ‘That’s what really convinced me we weren’t seeing the glow of the Galaxy.’
Of course, the COBE team could still have been hoodwinked. ‘If nature was malicious, it could have filled the halo of the Galaxy with three-degree dust,’ says Bennett. ‘That would look exactly like the cosmic background radiation!’
But many scientists who were not on the COBE team were ready to accept the result as correct. ‘I’m betting the COBE result is real and correctly measured,’ says Peebles. ‘First, I observe the people on the project – like Dave Wilkinson – and I have a deep faith in their ability to track down every last thing, which they were doing for six months or more.’
Peebles also believes that the evidence presented by the team is good. ‘Of course, it’s statistical only – they don’t have a picture of the face of God – but there are statistical tests of what they have found, and the ones I’ve examined seem pretty good.
‘I’m betting it’s good,’ he continues. ‘I would give you odds – not a million to one but, oh, three to one, something like that.’
By an odd coincidence, the hot spots detected by COBE were lurking just below the level ground-based experiments could pick out. ‘We certainly lucked out finding them,’ says Dave Wilkinson.
‘If COBE had been launched a year or two later, it would have been scooped,’ says Peebles.
It was clear that ground-based experiments, equipped with the latest detectors, would soon be able to see whether the satellite’s hot spots really existed. ‘There was a minor gold rush to be the first person to check the COBE result from the ground,’ says Peebles.
‘The unfortunate thing will be if people don’t find anything,’ says Bennett. ‘Everyone will wonder which experiment was right, which was wrong. It could drag on for ever. I’m pretty confident about what we did, but if we made a mistake, I’d rather it was caught sooner than later.’
As it happened, Bennett did not have to wait long.
In December 1992, a team of astronomers from Princeton University, MIT and Goddard found hot spots in the cosmic background radiation. They were similar in all ways to those the COBE team had announced finding eight months earlier.
Ironically, the scientists involved – Lyman Page, Stephan Meyer and Ed Cheng – found their hot spots before COBE did. But the effect was so small that it took the team nearly three years to confirm that what they were seeing was really in the Big Bang radiation and not something else – for instance, a spurious signal in their instrument.
COBE had the huge advantage that it orbited high above the atmosphere, which strongly absorbs the cosmic background radiation. The research budget of Page and his colleagues, on the other hand, did not quite stretch to COBE’s $60 million. They had to get their peek at the Universe by hoisting their instrument package to high altitude beneath a balloon.
The balloon experiment used bolometer detectors that were 25 times as sensitive as those on COBE, the technology of which was frozen in the early 1980s. This meant that it could make the same measurement as the satellite 625 times faster. ‘We could do in six hours what COBE took a year to do,’ says Page.
Wilkinson admits that the detectors used on COBE were ‘medieval’. But COBE succeeded, despite its insensitivity, because it was relentless, sitting up in orbit observing the background radiation day in, day out for more than a year. Instruments flown on balloons rarely get to observe for more than ten hours before high-altitude winds blow them out of range over the sea or over mountains.
Page and his colleagues had begun building their experiment in 1984. But when they first flew it, in 1988, disaster struck. ‘Our balloon burst,’ says Page. However, in October 1989, the team took its instrument to the launch site at Fort Sumner in New Mexico. ‘This time everything worked perfectly,’ says Page.
The balloon reached an altitude of 40 kilometres, where it stayed for ten hours. During this time, the sensitive instruments on board observed the Big Bang radiation for a total of six hours, scanning a quarter of the entire sky for tiny temperature differences.
‘When the data came down, it was clear there was a temperature variation in it,’ says Page. ‘The trouble was, we couldn’t tell whether the variation was in the background radiation or whether it had a more local source.’ It could have been caused by our Galaxy, the atmosphere or even by the instrument itself.
‘One by one we eliminated all the possibilities,’ says Page. The instrument observed at four wavelengths – one more than COBE – and this helped Ken Ganga at Princeton determine the Galactic emission, mainly from dust, and so subtract it from the signal.
Finally, the team had eliminated everything they could think of. They presented their result at a workshop held at Berkeley in December 1992.
The hot spots they had found were only 14.5 millionths of a degree hotter than the average temperature of the sky, slightly less than the 17 millionths of a degree found by COBE. But, as in the COBE map of the sky, the hot spots existed on all sizes, from seven times the apparent diameter of the Moon up to a quarter of the sky.
Page and his colleagues compared their map of the sky with the one obtained by COBE to see if the bumps and the wiggles were the same. They were. ‘It’s a pretty neat result,’ says Page. ‘We’re really happy with it.’
Meyer and Cheng had both worked on COBE as well. But several others working on the satellite made the same comparison independently, and they had confirmed that the agreement was very good. ‘The COBE team loves it,’ says Page.
What gave everyone so much confidence in the new result was that it had been obtained with an instrument that was very different from the one on board COBE. The balloon experiment used a single horn which pointed at 45 degrees from the vertical and spun round a vertical axis once every minute. In this way, it was able to compare the temperature around a ring of sky. In the six hours the balloon was performing experiments the Earth was turning beneath the sky, so the horn was able to sweep out overlapping rings covering a quarter of the sky.
Apart from using a very different instrument, Page’s experiment operated at slightly shorter wavelengths than COBE – between 0.44 and 1.8 millimetres. The difference was important since the signal from the cosmic background radiation is the residue left when the emission from our Galaxy is subtracted. At the shorter wavelengths of the balloon experiment the main emission from the Galaxy was from warm dust, whereas the Galactic emission seen by COBE was from electrons spiralling around magnetic field lines.
The Galactic emission is the major uncertainty in any experiment, so it was a great relief when, after using two entirely different models – one for emission from dust and one for emission from electrons – the balloon and space experiments were left with precisely the same hot spots in the cosmic background radiation.
So the COBE result was vindicated.
‘Back in 1974,’ says John Mather, ‘we set out to do a job so terrific that no matter what the theorists came up with we’d have all the data that anyone could get. We would reach the limits set by our location in the Universe.
‘You can’t send a space probe out of our Galaxy. You can’t even send one out of the Solar System. But we said we would do the best we can living here. We’ve just about done that.’
With both of COBE’s major discoveries now confirmed, attention turned to studying how the Big Bang radiation varied across the sky on an even finer scale. Even the smallest lumps of matter COBE had seen in the early Universe were bigger than the largest collections of galaxies astronomers have so far seen in today’s Universe. But lumps of matter in the early Universe should have arisen in all sizes. So if people zoomed in on small portions of the sky, they ought to be able to see lumps small enough to have been the seeds of individual galaxies like the Milky Way. The aim was to discover hot spots in the sky as small as half a degree across, which is the apparent diameter of the Moon and 14 times smaller than the smallest spotted by COBE.
Lumps of matter of this size were potentially much more important than the ones found by COBE. Those lumps were larger from end to end than any light signal could have traversed 380,000 years after the Big Bang, so there was no way they could have been affected by any processes occurring in the Universe at that time. If they told scientists about any epoch, it was a much earlier one, perhaps the first split second of the Universe. However, lumps of less than about two degrees across were small enough to have been affected by processes occurring 380,000 years after the Big Bang. Potentially, they would provide a panoramic window onto the Universe at the instant galaxy formation got under way.
If anything, the hot spots on the small scale should be hotter than those found by COBE. The reason is that COBE’s hot spots were not caused directly by matter but indirectly, through its gravitational effect on the fireball photons. But, on the small scale, theorists fully expected to see the direct effect of electrons on the photons of the fireball radiation. Before they combined with protons to form atoms, these electrons could have collided with photons, boosting their energy and making them appear hotter.
In the wake of COBE, there is now intense interest in the afterglow of creation. We now know that written across the sky is the story of the early Universe, and we are only just beginning to read that story.
With the help of COBE, we have seen the Universe through the most sensitive microwave glasses ever made. What at first appeared to be the unbroken whiteness of the fireball has resolved itself into a complex patchwork of light and shadow, telling us of the birth of the giant clusters of galaxies at the beginning of time.
Heartened by COBE’s discovery, an army of men and women with microwave glasses is now peering ever more closely at the Big Bang radiation. They are zooming in on smaller and smaller patches of sky in the hope of finding the seeds of individual galaxies like the Milky Way.
Until now, we have managed to glean only a few scraps of fundamental information about the nature of the Universe. But the cosmic background radiation promises to increase that knowledge greatly. The temperature of the fireball radiation is already the most precise thing we know about the Universe, and we are only just beginning to decode the secrets of this oldest fossil in creation. ‘There’s an awful lot more life left in this beast,’ says John Mather.
Wringing the precious secrets from the cosmic background radiation has been a long slog, but we have been lucky. Though the afterglow of creation is terribly faint, it is still possible to pick it out from the bright microwave glow of our Galaxy. Had humans evolved much later in the history of the Universe, it might have been a different story …
In another 13.7 billion years, the remorseless expansion of the Universe will have driven the galaxies twice as far apart as they are today.2 The photons of the background radiation will be stretched to longer wavelengths and diluted even more. The Universe, instead of being pervaded by a background glow at a temperature of three degrees, will be filled with radiation at only 1.5 degrees above absolute zero. From inside a galaxy like the Milky Way it will be hard to pick out ripples like those found by COBE. Hard but not impossible. It will simply take a lot more patient observing.
But when the Universe is three times the age it is now, the temperature of the cosmic background radiation will be only a third of what it is today; when the Universe is four times as old, just a quarter. By the time 137 billion years have elapsed since the Big Bang, the relic of the fireball will have all but died out. It will be a pathetic 0.3 degrees above absolute zero. If there are any intelligent species around 137 billion years from now, they will not be nearly as fortunate as we have been. In their Universe, the afterglow of creation will be essentially undetectable, its secrets for ever beyond reach.
The fate of the fireball radiation in the very distant future depends on whether the expansion of the Universe one day runs out of steam and goes into reverse. If this never happens and the Universe expands for ever, its dying galaxies becoming ever more isolated islands in an ever-growing ocean of space, then the radiation will simply be diluted out of existence.
On the other hand, if the Universe does stop expanding and embarks on a runaway collapse, the relic radiation will be rescued from such an ignominious end. As the Universe shrinks inexorably down to a big crunch – a sort of mirror image of the Big Bang, in which all of creation is squeezed again into an impossibly small volume – the background radiation will get hotter and hotter as it is squeezed to shorter and shorter wavelengths. No longer a few degrees at radio wavelengths, it will be a few tens of degrees in the infrared. Then, as the burnt-out hulks of galaxies are crushed together, the Universe will blaze again with visible light, corresponding to a temperature of thousands of degrees.
This will be the mirror image of the epoch probed by COBE. Atoms, instead of forming for the first time, will be broken apart. The Universe, instead of becoming transparent to radiation, will become utterly opaque. The billions of years of domination by matter will be over and radiation will at last be king again.
In the last minutes before the Big Crunch, all of creation will be a raging inferno. The ferocious light of the fireball radiation will begin to blast apart the nuclei of atoms into their constituent protons and neutrons. Soon all traces of ordinary matter will be expunged for ever from the Universe.
The fireball radiation will have returned whence it came. No longer the afterglow of creation, it will now have transformed itself into the deadly aura of destruction.
1. These are rather like the magnetic field lines revealed when iron filings are sprinkled about a bar magnet.
2. To make the numbers simple, I have ignored the fact that the expansion of the Universe is actually speeding up. This is because of the ‘dark energy’, invisible stuff that fills all of space and whose repulsive gravity is driving the galaxies apart. The dark energy was discovered only in 1998.