Thursday, 16 November 2017

The Election Day Event

Today the LIGO and Virgo collaborations made their sixth announcement of a confirmed gravitational-wave detection. Once again the signal was from two black holes colliding — this time the black holes were each close to 10 times the mass of the sun, with the smaller object possibly as low as 5 solar masses, making this likely the lowest-mass binary-black-hole yet observed.

Monday, 16 October 2017

Kaboom! Two kinds of astronomy collide, and produce a burst of science

Holy crap! The LIGO and Virgo collaborations just announced their most incredible observation so far. After the first direct detections of gravitational waves from black holes colliding, that’s really saying something! And not just that — a string of other astronomical collaborations are part of it, making it even more incredible. If you have just heard the news, you may be in a whirl trying to make sense of it all. Even if you have spent the last two months scooping up all those scurrilous rumours, this may still be just too much. Hell, I have known about it since the observations came in on August 17th, and I am having trouble getting my head around it.

Ok. I need to calm down. Let’s take a step back, to the afternoon of August 17th (UK time).

Prelude


I was in the postdoc office, delivering a beautiful post-liquid-lunch soliloquy on the supremacy of binary black holes. All of the postdocs were listening intently, in that funny way they do, by looking directly at their computers and typing. The crux of my pontification was that so far the Advanced LIGO and Virgo detectors had found several gravitational-wave signals, but all of them were from the same kind of source: black holes colliding. For years people had been honking on about neutron stars, and how neutron stars spiralling together were going to be the prime LIGO source, and thumbing their noses at us binary-black-hole dreamers. And there was another crowd who were all worked up about “multi-messenger astronomy” — they hoped that gamma-ray satellites would observe a gamma-ray burst (which is an abstruse technical name for a burst of gamma rays) at the same time as we detected the gravitational waves from the neutron stars spiralling together. That would prove for the first time the hypothesis that neutron-star collisions are what produce short gamma-ray bursts. They were convinced that that was going to be what gravitational-wave astronomy was going to be all about.

No such luck. We had detected no binary neutron stars. There had been five months of the first LIGO-Virgo observing run, and then another almost nine months of the second observing run, and no neutron stars. Only black holes. Check out detections One, Two and Three. And just that Monday had come detection Number Four — black holes again.

Neutron-star fans were out of luck.

After a pleasant lunch, I was full of sympathy for them.

“They should stop calling it a gravitational-wave observatory!” I crowed. “They should rename it a black hole observatory!”

Some of the postdocs acknowledged this wisdom in their own special way, by putting on headphones.

The Big Moment


Five minutes later, Laura suddenly started yelling, “Send it! Send it!”

At first I did not pay much attention. She was on a conference call, and sometimes those things get a bit heated. Even when it comes to asking someone to send an email.

But she was not talking about emails. They were discussing whether to send a “circular”, an online notice to astronomers that we had found something interesting that they might want to point their telescopes at.

I now realised that three postdocs were yelling on three separate conference calls, and all of the rest were gathered around their desks.

Occasionally they tried to help me understand. There had been an alert that the Fermi gamma-ray satellite had found a gamma-ray burst. It was weak, so the gamma-ray people were unlikely to care much about it; they see these things all the time. Around the same time, the LIGO Hanford detector registered a binary-neutron-star signal. That sounds exciting, but the computer analysis codes had rejected it, because a confirmed detection needs a signal in at least two detectors, and the analysis codes decided that the Livingston detector data were poor, and there was no sign of a detection in Virgo. In other words, at first glance it might not sound interesting for gravitational-wave people, either.

But: any chance of a neutron-star detection is too good to pass up. Time to check by hand the data from Livingston and Virgo.

That was when I heard people periodically shouting “Koi fish!” around the room.

Laura explained to me, “A koi fish is a kind of glitch.”

I absorbed this information using the customary professorial response. “I know that.”

“There’s a great picture that just got sent around.”

(Yes, we were only minutes into this thing, and there was already a flood of emails, website links, data tables, and figures.)

This is something like the figure she showed me:

Signal plus "koi fish" glitch.
(Adapted from Fig.2 in the main LIGO-Virgo discovery paper.)

The big yellow splotch is the “koi fish” glitch — a patch of noise in the data that obscures any signal for a short time. But leading up to the glitch is the most incredible, perfectly clear signal any of us had ever seen: instantly recognisable to afficionados as a binary-neutron-star inspiral. The slowly rising line is the frequency of the gravitational wave increasing (“chirping”), as the two neutron stars orbit closer and closer, moving faster and faster. The full signal visible in the data was one hundred seconds long. ONE HUNDRED SECONDS! That’s over a MINUTE AND A HALF! The longest previous detection was the binary-black-hole signal GW151226, which lasted only ONE second! The movie illustrates the point best.

[Credit: LIGO/University of Oregon/Ben Farr]

Once the glitch was cleaned out of the data (you can read all about it in the paper), the combined data from all of the detectors added up to the loudest signal we have ever observed. (The signal was weak in Virgo because of the source's location in the sky. The fact that the signal was so weak in Virgo helped us to localise the source to somewhere over south-Eastern Africa.)

And of course there was also the gamma-ray burst. So we now know for sure that short gamma-ray bursts are indeed produced by colliding neutron stars.

Twenty minutes earlier I thought we were going to have to wait another five years, or longer, to answer that question, and most likely slowly and hesitantly, after a series of weak detections and “possible”, not completely certain, gamma-ray-burst counterparts. Now it was done.

A Rare Gift


I am a devoted disciple of black holes, yet I will gleefully admit that this was the most incredible detection so far. A few reasons:

Neutron-star mergers are still going to be rarer sources than binary black holes. We know this from the fact that in the last two years we have not observed any others. That is what made this observation so astounding: the two neutron stars were ten times closer than any of the black holes we have observed so far. Even the most evangelical and caffienated neutron-star lover never dared to suggest we would observe a signal this close; it would have been a joke. Only about 1 in 50 neutron-star detections will be this strong — so it could be a long while before we see anything like this again. (I explain this back-of-the-envelope estimate here.) That makes this a very rare and precious signal. Factor in that satellites will see the gamma-ray burst in only a handful of binary-neutron-star detections, and we see that it is a truly remarkable signal indeed.

When we observed binary black holes, we learned nothing new about the nature of black holes. We already knew from Einstein’s equations what the signals would look like. That was why we were confident that we had squeezed as much information as we could out of the signals: we could match the signals we had calculated against what we measured. In the case of black holes, the theory was well ahead of observations.

Neutron-star mergers are the other way around. People considered them a likely source of gamma-ray bursts, but no-one could confirm it with theoretical calculations. People have struggled to produce a gamma-ray burst from computer simulations of merging neutron stars for years, but without conclusive results. Even when they artificially cooked up the most promising conditions, they could only manage a hint of the beginning of a burst. This is because computer simulations of neutron-star mergers are much more difficult than for black holes — there is all that nuclear physics to take into account, for a start, and besides all of the extra (in some cases speculative) physics involved, calculating what is going on in every centimetre of ten-kilometre-wide objects takes a lot of computing power!

So now we have that ideal situation in science: the observations are guiding the theory. Combined gravitational-wave and gamma-ray measurements have nailed the question of the source of short gamma-ray bursts. Now it is for the theorists to catch up.

There was a huge amount of other information in these observations. I am not qualified to talk about the measurements from EM astronomers, but even from gravitational waves there was much more to do. One lovely example was to make good on Bernard Schutz’s thirty-year-old idea that we could use gravitational waves to measure the expansion of the universe. The calculation is not yet as accurate as those from conventional astronomy, but since the two current measurements disagree, this gives us hope that one day gravitational waves could nail this problem, too.

The future


GW170817 is going to be a goldmine of science for a while to come, and it should be clear from everything I have said that it will be a long time before we find a richer signal. That is a rock-solid prediction for the next year, because the LIGO and Virgo detectors will be switched off for another upgrade. When they switch back on in 2019, we may find one or two neutron-star mergers each year that coincide with gamma-ray bursts — but, again, it is unlikely that any of the signals will be as strong as GW170817. For that we may well have to wait until the next decade.

Or I could be wrong, and we get another surprise. It has happened before.



More Gravitational-Wave Stories

February, 2016:
The Discovery
How it Felt
How We Squeezed Out the Juicy Science

March, 2016:
Trying to Explain Gravitational Waves (Part I) (Part II)

June, 2016:
Book Review: Black Hole Blues
Detection Number 2 -- Black Holes Rule!
Rumours, Secrets and Other Sounds of Gravitational Waves

February, 2017:
One Year Anniversary (of being world famous)

June, 2017:
Detection Number 3 -- Nothing to see here: they are black holes
A hint of controversy

September 2017:
Detection Number 4 -- Virgo nails it

October 2017:
Did I just win the Nobel prize?

November 2017:
The Election Day Event

Tuesday, 3 October 2017

Woohoo! I just won the Nobel Prize!

Today I won the Nobel Prize in Physics. Yeah, I am pretty excited about it. To be honest, though, I was not surprised.

The prize was awarded for LIGO's first direct detection of gravitational waves, which we all know was the biggest scientific discovery of the century. That is not me showing off. That is what other people have said. It is such an undisputedly incredible achievement that all of us LIGO scientists can afford to be perfectly humble about it. We knew the Nobel was coming. We can take it in our stride. Hell, after all of those telecons of the Nobel Prize Acceptance Working Group, we were starting to get a bit sick of the whole thing.

Just a second. What is that you say? The prize is not for me? Nonsense. Take a look at the citation.