91av

TO SEE how it all began, you have to go to the end of the Earth – a kilometre from the Amundsen-Scott station at the South Pole, to be precise. There, huddled against the great white wilderness, a telescope captures light from near-enough the beginning of time, interrogating it for answers to the oldest question of all: how the cosmos came to be.

One of the most famous answers says that the universe had a stupendous growth spurt in its first moments, expanding almost as much in a split second as it would in the following 13.8 billion years. It is quite a claim.

A few years ago, the researchers who built that Antarctic telescope were convinced they had found proof of this cosmic inflation, swirling patterns written in ancient light that signalled the universe’s early ballooning. Their discovery made headlines around the world. Then they realised there was an error in their analysis and the result turned to dust.

Except that wasn’t the end of the story. Those physicists got back to work and now, almost five years on, they are poised to catch their quarry for real. There is no guarantee, and they are bound to be more cautious this time. But with no other telescope capable of sighting these elusive signals due to come online for at least a decade, their upgraded detectors are the best shot we have at finding the truth about the origins of everything.

“This is probably the most exciting thing in cosmology today,” says Dan Hooper, a cosmologist at the University of Chicago, who isn’t involved in the search. “With these experiments, we are now starting to explore exactly the region we expect to discover the signature of inflation.”

The light this telescope seeks is known as the cosmic microwave background or CMB. For 380,000 years after the big bang, there was nothing but a ludicrously hot, dense soup of subatomic particles. Photons, the particles of light, pinballed between them, unable to escape. Then things cooled enough for hydrogen atoms to form, at which point photons could move unhindered. It was as if someone had flicked a switch: light shot out in every direction and the cosmos became transparent. The CMB is made up of those first liberated photons.

You can’t see this light with the naked eye because, as the universe expanded, it stretched out from visible wavelengths into microwaves, hence the name. But it permeates all space, raining down on Earth from every direction, and it has been a rich source of information. The CMB is very special, says at Paris Diderot University in France. “It is a backlight that shines on all of the structures between its emission and us, so it probes basically all the history of the universe.”

One of the most controversial ideas it can illuminate is inflation, the hypothesis that describes how the universe ballooned in its first moments. Although it sounds outlandish, inflation is a good explanation for the surprisingly smooth distribution of galaxies across the universe. Surprising, because quantum theory tells us that tiny energy fluctuations should have been bubbling away constantly in the universe’s first moments. These would have caused particles to pop in and out of existence in random places, resulting in an uneven distribution of matter that should have been amplified as the cosmos expanded. But we see no such imbalance. The smooth universe we observe might be explained by a rampant initial growth that ironed out the differences.

At least, that is what various models of inflation argue, and the CMB has already provided a degree of support. In July 2018, the European Space Agency’s Planck satellite released . The tiny temperature variations it charted are largely consistent with several of the key predictions arising from the most popular inflationary models.

Bang or bounce?

Fair enough, but cosmologist Paul Steinhardt at Princeton University, one of the original architects of the inflation hypothesis, is among those who now argue that it is inherently too flexible to be ruled in or out by observations. They reckon we should seriously consider the most prominent alternative: that our universe is the result of a previous one collapsing and rebounding. Recently, this big bounce scenario has itself enjoyed something of a bounce, with several groups demonstrating that it is theoretically possible without needing to invoke exotic physics.

The CMB could settle the matter, but only if we can enhance our current picture with a more detailed appreciation of how its microwaves corkscrew along their direction of travel. Planck did map this property, known as polarisation, but only with limited sensitivity. So even now we have barely scratched the surface. “This is where the last secrets of the CMB are hiding,” says Delabrouille.

One thing that could be concealed there is firm evidence of inflation. The idea is that a violent burst of expansion would have created turbulence in the fabric of the early universe, generating gravitational waves. These would be too weak to be picked up by the Laser Interferometer Gravitational-Wave Observatory (LIGO), the founders of which won the Nobel prize in physics in 2017 for detecting ripples in space-time produced by colliding black holes. Even so, these tiny waves would have left an imprint in the polarisation of the CMB, in the form of a telltale swirl pattern known as a B-mode (see “Diagram”).

“The amplitude of this signal can vary between different inflationary models,” says at Cardiff University, UK. “Measuring that signature would therefore not only be clear evidence of inflation, it would also inform us about how it happened.”

The trouble is that any signals left by primordial gravitational waves in the twisting of this ancient light would be devilishly difficult to pick out. For one, they are extremely subtle to begin with, far smaller than the temperature fluctuations mapped by Planck. Then there is the added complication that they must be teased out from similar-looking patterns created by all the other stuff in the universe that can twist light.

No one knows that better than the researchers behind the BICEP2 telescope, one of a small cluster of observatories at the South Pole. In March 2014, they announced they had detected the signature of primordial gravitational waves in the CMB polarisation. Newspaper headlines around the world heralded the discovery. A Nobel prize beckoned. But closer inspection made it clear that the signal was caused not by primordial ripples, but something altogether more mundane: the dust that fills our galaxy.

Microscopic grains tend to align with the local magnetic field such that microwaves scattering off them can generate B-mode patterns in the frequency range the BICEP2 telescope looked at. The BICEP team knew this, but grossly underestimated the contribution of dust to their signal.

“The swirls would be clear evidence for inflation – and tell us how it happened”

The climbdown was a chastening experience for everyone involved. But the storm of criticism blew over, and the researchers quietly got back to the job, heeding lessons about how to separate signal from noise. They return to the South Pole every austral summer to upgrade their instruments. “That’s one of the things I love about this,” says at Harvard University, who leads the renamed . “We built these telescopes and we still get to go down there and tinker with them. It’s a real adventure.”

Tinkering doesn’t quite do it justice. Since 2014, Kovac and his colleagues have dramatically boosted the raw sensitivity of their telescopes. BICEP2’s detector had 256 pixels but its successor, BICEP3, boasts 1280. They have also added the Keck Array: five detectors trained on the same patch of sky. “The improvement has been huge,” says , who helms the South Pole Telescope, one of a few other observatories at the South Pole that look at the CMB. “In terms of sensitivity, I don’t think anyone is even close to the BICEP/Keck people.”

Equally, if not more important, the Keck Array allows the team to tune into different specific wavelengths. That is useful because Planck recently showed that the dust in the Milky Way polarises some wavelengths of light more than others, whereas the polarisation resulting from primordial gravitational waves should remain uniform. “That was probably the most important lesson,” says Kovac. If you see a signal that doesn’t vary across wavelengths, you are onto something.

A further boost comes from the South Pole Telescope, which studies the CMB across a wider arc of sky than the BICEP/Keck array. This makes it a powerful tool for mapping another source of foreground contamination. It turns out that the gravitational heft of galaxies deforms the CMB to create a maddeningly convincing mimic of a primordial B-mode. Over the past two years, however, Carlstrom and his team have demonstrated that they can measure this distortion, which means they can remove it from the picture.

“Now we can combine these methods, we have a much more powerful way to separate signal from noise,” says Kovac. “We can dig deeper than ever before.” Indeed, they have already started. The latest BICEP/Keck Array set-up began gathering data in 2016 and improvements are ongoing. The question is, can it go deep enough to uncover inflation’s signal in the CMB, if indeed it is buried there?

Kovac is understandably cautious. Any detection would require all manner of cross-checks, he says, starting with a sighting of the same signal by other telescopes looking at different patches of sky. Likewise, much hinges on the size of the signal. “As always, a lot depends on nature,” he says. Even so, there is a detectable sense of optimism buried in Kovac’s carefully considered responses.

The reason for that optimism is probably the narrowing window in which B-modes might lurk. Whether we can detect them depends on how strong they are, which we can discern by measuring a statistical property of the CMB known as the tensor-to-scalar ratio. Planck’s map of the CMB published in 2018 showed this ratio must be above 0.001. And the latest measurements from BICEP, published in October, . The upshot is that if the most favoured models of inflation are right, primordial gravitational waves should be hiding somewhere between the two, just the energy window that the BICEP/Keck Array is beginning to probe.

“Now there is a very clear target,” says at Johns Hopkins University in Maryland. “If we don’t see something in that range in the next 10 years, it’s going to seriously piss us off as theorists.” Carlstrom agrees: “The predicted signal is within grasp.”

Scientists are already preparing for a scenario in which the BICEP/Keck array sees nothing, planning a next-generation collaboration of ground-based CMB telescopes that would collectively be 10 times as powerful as all the existing observatories combined. Some are even plotting a space telescope. But all such plans are still on the drawing board. There is, then, a big opportunity for a small group to transform our understanding of the cosmos. The Higgs boson was discovered by a team of hundreds of scientists. The BICEP/Keck team, on the other hand, could make a similarly momentous discovery with two telescopes and a team of 60-odd people.

If the primordial B-mode signal is at the top of the energy window, it could be sighted fairly soon. “If the signal was to emerge at the upper end of that range, things would get interesting quite a bit faster,” says Kovac.

It is also entirely possible that his collaboration has already detected something, even if few in the field dare dream it. The measurements released in October were collected in 2015 using BICEP alone. But both parts of the observatory have been taking data for two years now, none of it released. Who knows what it contains.

Kovac does, although he isn’t giving any hints. Then again, the one thing you can be sure of, given what happened last time, is that he and his team will want to be damn sure they have something real before bringing another discovery to the table.

“The burden of proof is enormous,” says Carlstrom. “We know the signal is buried deep. But we’ve got to make this measurement. We have a chance to see the dawn of time.”