Gravitational Wave Background (Script) (Patreon)

It was pretty impressive when we built this giant machine that spotted gravitational waves from colliding black holes. Well we’ve just taken that to the next level with a galaxy-spanning gravitational wave detector that may have detected a foundational element of space itself - the gravitational wave background.

—-

When the Laser Interferometer Gravitational Wave Observatory - LIGO - turned on in 2015 it quickly spotted exactly what it was built to detect - vast ripples in the fabric of spacetime produced by colliding black holes a billion light years away.  Since then, LIGO and its partner VIRGO have detected 50 similar signals from merging black holes and neutron stars across the universe. These signals sweep past the Earth every few days, causing space itself to oscillate. Lengths and widths contract and expand in turn by the tiniest fraction - one one-thousandth of the width of a single proton over the 4km arms of the LIGO antennas. And these are only the most powerful gravitational waves that we can detect - produced by only relatively nearby mergers.

The reality of these “strong” spacetime ripples reveals an awesome truth - we know that black holes and neutron stars have been merging for all of cosmic history, and supernovae have been exploding, and stellar remnants have been spinning, and probably at or before the instant of the big bang, insanely energetic events took place - and all of these produced gravitational waves. That means the entire universe must be thrumming with these vibrations all the time, and everywhere. We call this underlying buzz the gravitational wave background, and it could reveal the lives of the most massive black holes in the universe, provide a glimpse into the time before the big bang, and reveal the structure of spacetime itself. Oh, and we may just have detected it.

LIGO itself may be able to detect a small part of the gravitational wave background - the GWB - but it’s not there yet. But it turns out we already have another device

that may have done the job. It’s a gravitational wave observatory that spans the entire galaxy and recruits some of the most bizarre stars in the universe as our measuring devices. I’m talking about the pulsar timing array, and about the first tentative detection of the gravitational wave background revealed in early January by the NANOGrav collaboration.

But it’s going to take a minute to get to that. First let’s talk about pulsar timing arrays and why we need them. LIGO has a fundamental limit - it can only see those gravitational waves with frequencies of around 1 to 10,000 Hz - corresponding  to relatively compact waves from 10s to 100s of thousands kilometers in wavelength. LIGO can’t see ripples with either smaller or larger wavelengths.

LIGO is like a rowboat on the ocean. You’re sitting in the rowboat, eyes closed. You can feel the rise and fall and rocking of the boat  in the choppy ocean - you and your boat are sensitive to those waves. Very tiny ripples on the ocean surface just lap at the hull, but you don’t notice them. And there are also vast, rolling waves that are so big that the rowboat feels steady - eyes closed, you don’t notice those either. OK, now you’re on a giant cargo ship - it tips forward and backwards in the ocean due to those giant waves - eyes closed on deck, you notice that tipping, but are oblivious to the little waves that rocked the rowboat. And there are also much larger waves - a tsunami perhaps, or the planet-sized wave of the tide itself. There’s no human-built ship that is rocked by those waves.

It’s the same with gravitational waves. LIGO is rocked by the little ripples produced at the instant of merger of stellar remnants like black holes and neutron stars. Or potentially in supernova explosions.  But we know the space-time ocean must have swells that are millions to billions of kilometers long - produced by the merger of the most massive black holes across the universe, or by the slow inspiral of smaller compact objects in our galaxy. For that we’re building our giant ocean tanker - the Laser Interferometer Space Antenna, or LISA.

But even tankers are oblivious to the ocean tide. And ridiculously colossal gravitational waves must also exist - waves whose single oscillations span of the solar system, or even the distance to the nearest stars. Such gravitational waves are expected to be produced in a few different ways. Black holes with masses of millions to billions of suns live in the cores of most galaxies - these supermassive black holes can end up orbiting each other when galaxies collide, and that slow, fatal inspiral produces enormous gravitational waves. And there are theoretical sources also. Cosmic strings - topological seams in the fabric of space - would produce waves when their kinks become un-kinke� In the speculative time before the big bang, cosmic inflation was driving exponential accelerating expansion - and huge gravitational waves would have been produced in the phase transition that ended inflation. All of these spacetime tsunamis would be invisible to our tiny human-made machines. To see them, we’d need a gravitational wave observatory spanning the galaxy.

And fortunately, nature has provided us with just such a network: a system of buoys scattered across the Milky Way, each fitted with the most perfect measuring devices known to science. Obviously I’m talking about pulsars.  To remind you - a pulsar is the collapsed core of a massive star - masses of at least one and a half suns packed into balls 20 km across, and that extreme density converts all their matter into neutrons. They are on the edge of absolute collapse into black holes, supported only by weird quantum forces.

Neutron stars tend to channel jets of high-energy particles due to their intense magnetic fields. They also rotate rapidly, with rotational axis offset from the magnetic poles, causing those jets to sweep through space - to precess. When the jets sweep past the Earth we see flashes of electromagnetic radiation - and we call those flashing sources pulsars. Some flash every few seconds, but the fastest - millisecond pulsars - can rotate 100s of times per second. Just to give you feel for that - this is what a “typical pulsar sounds like. That’s the sound of a trillion-trillion-trillion tons rotating 1.4 times per second. And this is the crab pulsar, with its 30 rotations per second And this is true millisecond pulsar - at 174 rotations per second. And the fastest known pulsar with its 642 Hz rotation.

These pulses are not just fast, they’re exquisitely regular. With the most stable millisecond pulsars, the exact instant of one of these ticks can be predicted to within one ten millionth of a second - several years prior to that individual pulse. So nature has graced us with these perfect clocks scattered across the galaxy. How do we use them to detect gravitational waves? The whole thing about gravitational waves is that they change distances - which can change the travel time of the pulses from a pulsar, leading to a difference in the arrival time of those pulses compared to what was expected.

Here’s a simple example. Imagine a pair of supermassive black holes are orbiting each other on the other side of the universe. As their gravitational waves pass through the Milky Way they cause space to stretch and contract in succession at right angles to the direction of the wave’s motion in what is known as a quadrupole wave.

 If the wave passed through a ring of points they’d move in and out like this. OK, so imagine those points are pulsars, and Earth is at the center of the ring. As the wave passed by, the signal from these pulsars would have to travel a little further - there’d be a lag in the arrival time during this part of the oscillation. For these pulsars the distance traveled is shorter, so the signal arrives early. After one oscillation cycle, the situation reverses. If we had pulsars in all directions we could record the so-called timing residuals - the lateness or earliness of pulses - and we’d record this sort of shamrock shape where the petals are large timing residuals and the diagonals show very little timing residual.

Of course space is 3 dimensional. A pulsar signal coming from the direction as the gravitational wave will actually ride the wave, keeping pace with it - and so may spend a lot of its journey either in a stretched-out region of space or a compacted region. Those guys have large timing residuals, while a pulsar on the opposite side will skip through many peaks and troughs of the gravitational wave so that its timing residual cancels out.

The end result is a very predictable pattern to the lag experienced by pulsar signals as a function of their position. It’s called the so-called antenna pattern. A gravitational wave coming from here towards the Earth, here, causes the largest timing offsets for pulsars within this structure.

Now a single pulsar isn’t enough to map this structure - and it’s not enough to even be sure we’re seeing a gravitational wave. Other things can cause shifts in pulsar signals - for example neutron stars experience star quakes which can glitch their rotational rates. The magnetized plasma between the stars also slows radio waves from pulsars, and the relative motion of the Earth and Sun and Milky Way shift the frequency of pulsars. But if you have an array of pulsars - a pulsar timing array, in fact - then this extremely characteristic pattern leads to an unmistakable correlation in the arrival time of the many pulsar signals.

So this is where you probably want me to tell you that our galactic array of quantum star-clocks  has revealed gigantic gravitational waves from some colossal black hole dance or from the big bang or something. Well, maybe, but not quite so simple - at least not yet. Instead, it may have detected the sum of all such waves - the cosmic gravitational wave background.

Let’s get to the actual result. NANOGrav - the North American NanoHertz Observatory for Gravitational Waves - made an announcement at the American Astronomical Society meeting in early January. NANOGrav is a collaboration of over 100 US and Canadian researchers, and itself is one of several groups that together make up the international pulsar timing array, which combines many of the largest radio telescopes on the planet to monitor around 100 millisecond pulsars.

NANOGrav’s own pulsar timing array consists of 47 pulsars which it monitors with the Green Bank Observatory in West Virginia, and with Arecibo Observatory in Puerto Rico. I should say it previously monitored with Arecibo - you might have heard, the iconic Arecibo is no longer with us after its hurricane-damaged cables snapped, causing a catastrophic collapse on December 1st. We had the great honor to visit Arecibo in 2019 and shot some 180 video there. I invite you to climb the catwalk with me. It’s a tragedy to have lost such an important workhorse for our exploration of the universe, but NANOGrav’s findings are now part of Arecibo’s broad and enduring legacy.

While NANOGrav didn’t detect a single, resolvable gravitational wave, its 12.5 years of observations have tentatively detected the gravitational wave background. This is the sum of all gravitational waves too weak to be detected individually, but which together form this universe-wide thrum of activity

The “GWB” doesn’t produce the tell-tail quadrupole antenna pattern - rather it results in more subtle correlations in the timing residuals between pulsars. Individual fluctuations in this background will result in a correlation across all pulsars in the array, and the entire background results in this overlapping set of correlations that can be extracted with sophisticated statistical techniques. Ultimately, the result will be a distribution of wave frequencies - a spectrum for the background - and the shape - the slope of that spectrum will help tell us whether the gravitational wave background is produced by supermassive black holes, cosmic strings, or events during the inflationary epoch - or a combination of the above. But the current NANOGrav result isn’t sensitive enough to measure the shape of that spectrum yet. It’s just revealed a signal correlation in pulsar timing residuals that can’t be explained by other known processes.

If this is real then there’s no question that the gravitational wave background will become better and better resolved as we add pulsars to the timing array, and stare at it for longer with more and larger radio telescopes. The promise is a deeper understanding of the weirdest things in the universe - and if we manage to prove the existence of cosmic strings, inflation, or other weird physics this way - then we have opened a window into realms of physics beyond our current imagination. All it takes is our galactic fleet of star-clocks, bobbing in the ripples across the ocean of spacetime.