Gravitational Waves have been detected!

The LIGO-VIRGO consortium has today (2016 Feb 11) confirmed that they have directly detected, for the first time ever, gravitational waves passing through the Earth. If you're not a scientist, you may be wondering what all the hoo-ha is about, so let me put it like this: this is big. Really big. And then some. I guess you could say that this is astronomy's equivalent of the Higgs Boson discovery, although for gravitational waves the delay from when they were predicted (by Einstein) to detected is much longer than it was for the Higgs.

So why did it take so long to detect them?

First - I'm not going to go into details here of what gravitational waves are, because I'd just be duplicating what others have done better. A good quick intro is available via PhD Comics or at Universe Today, and of course the LIGO website. The very concise summary is: they are ripples in the fabric of space-time that was predicted by Einstein's theory of General Relativity.

These ripples in space-time are ludicrously hard to detect; the advanced LIGO facility is designed to notice if the gap between two mirrors, normally 4km, changes by something like 0.000 000 000 000 000 000 01 metres. That's not a lot. In fact, that's so little that all sorts of things not related to gravitational waves can shift the mirrors by a lot more than that: seismic activity, traffic on nearby roads; frankly, I'd be nervous to sneeze in the control room whilst the experiment was turned on. So detecting a gravitaional wave (which I'm going to call GW from now on) is extremely hard. And these tiny, tiny changes in distance caused by a GW are still too small for advanced LIGO to detect in most cases; the sort of powerful GW LIGO can spot are the sort that arise when a very massive object is violently accelerated, such as something spinning in a tight circle. The objects most likely to be detected by advanced LIGO would be a pair of compact stellar remnants - neutron stars or black holes - spiralling towards each other and coalescing. This doesn't happen very often, although exactly how often is not well constrained. Before advanced LIGO switched on the predictions for how many coalescing neutron star binaries should occur per year close enough for us to detect ranged from 0.0004 to 3 per year!

And what exactly did LIGO detect?

The LIGO team have reported finding a strong signal in their data, that matches what we'd expect from two ‘stellar mass’ black holes coalescing, at a distance of about 500 mega-parsecs from Earth. That's about 1.6 billion light years away. By ‘stellar mass’ black holes we mean those that have masses a few times that of the Sun; these are what's left behind when a really massive star ends its life as a supernova explosion.

OK, but what's so significant about this?

Up until now, we've never seen a gravitational wave. We thought they existed; we've seen things (like the way close binary stars behave) that can be explained if gravitational waves are emitted as we expected. But (as I regularly tell my children!), thinking something doesn't make it true. But now we've actually seen these waves directly.

But that's only the start. Of course, as a scientist it's really exciting to see a long-standing theory finally endorsed by direct observational evidence, but we want to do more than just prove Einstein right (it's not like his reputation needs enhancing), we want to do science! And being able to detect gravitational waves gives us a whole new set of tools to use. Although there's a lot we take for granted nowadays, astronomy is really hard: essentially we look at a tiny dot on the sky - often too faint to even see without expensive, large equipment - and try to work out incredibly intimate details about that dot (size, weight, age, composition, movement, what's around it, how it's interacting with that, etc). And we're getting pretty good at this, using light. But in the same way that a colour picture contains more information than a black and white one, or an audio-visual experience is often more complete than simply looking at pictures, being able to add Gravitational Wave information to what we get from light enables us to learn much more.

For example, the Swift satellite that I work for studies Gamma Ray Bursts or GRBs. One type of GRB, called ‘short’ GRBs (because they're, well, short) are believed to occur when two neutron stars merge together - if we're looking at it from the right direction. And as I mentioned just a few moments ago, that's the sort of thing that advanced LIGO should be able to detect GW from. From the observations with light we can get quite a lot of information: how much energy was radiated from the object, and over how long a time, where the object was in its galaxy and the redshift of the object (how fast it's moving away from us - which is related to its distance) for example. The GW observations can give us information about the masses of the stars that coalesced, potentially information about which direction they were orbiting in, and how far away they are. Each of these sets of information on their own is interesting and helpful, but the combination of the two gives us real diagnostic power, to do science and test our theoretical understanding of these extreme physical conditions. The combination of redshift (from light) and distance (from GW) is particularly noteworthy as it helps us constrain the cosmological parameters: the things that tell us how quickly the universe is expanding and accelerating.

So did Swift see a short GRB alongside this GW trigger?

Alas no. The likelihood of Swift detecting a GRB at the same time that LIGO detects the GW is pretty low. Swift-BAT (the bit that detects GRBs) can ‘only’ see 1/6 of the sky at a time, so for 5/6 LIGO detections we'll be looking the wrong way. Also, when two neutron stars merge, a jet of material is fired out of the ‘top’ and ‘bottom’ of the newly-forming black hole (rather like in the stylised image that forms the background of this page). We only detect a GRB if those jets point towards Earth; whereas we can detect the GW more or less regardless of the orientation of the system. We don't have a great handle of the size of the jets from short GRBs, but a reasonable estimate is that the opening angle is about 10°; if that's correct then only 1.5% of merging neutron stars will result in a short GRB we can detect from Earth.

But that doesn't mean we can't detect the light from a GW event. A GRB is followed by an ‘afterglow’ from material around the neutron stars which gets heated up by the GRB jet. This emission is also beamed at first, but later on it starts to be radiated in all directions. This emission is much fainter than a normal GRB, but we can only detect GW from nearby objects, and nearby means bright. Avid followers of my sporadic tweeting will know that I published a paper not long ago which investigated how to detect GRB afterglow emission with Swift, from objects both with the jet towards and pointed away from us.

To search for this afterglow emission is not straightforward. LIGO can't give very accurate information about where on the sky the GW came from. In the case of the detection announced today, the region of sky they were able to tell us the GW came from was about 600 square degrees, that's more than 3000 times the area covered by the full moon! For comparison, the X-ray telescope on Swift can see about 0.1 square degrees at any given time. There are things we can do to make this more feasible (only look where known galaxies are, focus on those galaxies with the most stars in them) which reduces the area we have to cover while not reducing too much the probability of looking in the right place. But this trigger caught us on the hop - it actually came right at the end of the LIGO engineering and testing phase, and we weren't yet ready to do the complicated, rapid-tiling observing we had planned. We did however look at a handful of locations on the sky, containing 18 galaxies. We also looked at a chunk of the Large Magellanic Cloud (a small galaxy orbiting our own Milky Way) which lay within the region LIGO said the GW may have come from. We didn't find anything (well, we found 3 X-ray sources, but they were known objects, not related to the GW).

As it turns out, for this particular event it's not that surprising that we didn't see anything. As I noted earlier, the LIGO team believe they've detected two stellar-mass black holes merging, 500 Mpc away. Unfortunately this information wasn't avialable at the time, and so the galaxies we chose to look at were all much nearer to Earth than that (for the merger of two neutron stars to be detected by LIGO they'd need to be less than about 100 Mpc from Earth). In other words - we looked in the wrong place! But also, the merger of two stellar-mass black holes is not expected to produce any light, so we wouldn't have seen anything even if we had got the right place!

Despite the lack of an X-ray detection from the GW event, this was a very valuable exercise for us at Swift, to prove that we can respond rapidly to GW triggers, and to find the bugs in the software I spent chunks of last year writing! But today, what Swift did or did not see is not important. Today belongs to LIGO, and it's a day that we'll all remember for a long time.