When two black holes eat each other, they release a lot of energy.
A lot. A significant fraction of the mass of the black holes is converted into energy, radiating away as gravitational waves, ripples in the fabric of spacetime. These can have as much energy as — and I am not kidding you here — tens of thousands of times as much as the Sun will radiate over its entire lifetime… and they blast away in a few seconds. Put another way, during that time the black holes emit hundreds of millions of times more energy than all the stars in the galaxy combined.
That kind of power is terrifying. Mind crushing. Yet, for all that, the event itself is entirely invisible, giving off no light at all. It's possible under some specific circumstances some light will be emitted, but generally speaking the two black holes merge, scream out their gravitational waves, and become a single, somewhat larger black hole… and all this without a photon emitted.
So no light is produced directly. But indirectly, it turns out things can be different.
On 21 May, 2019, the LIGO and Virgo gravitational wave observatories detected a pair of black holes merging somewhere in the Universe — the event was given the designation S190521g. Triangulating the signal, it was found to come from a spot in the sky the direction of the constellation Coma Berenices. Within moments, an automated alert was sent out to telescopes all across the planet to look for some sort of optical signal.
No flash of light was seen. At least, not right away.
On Mt. Palomar in California there is a 1.2-meter telescope with a 600 megapixel camera that can see a whopping 47 square degrees of sky at a time, a huge field of view. Called the Zwicky Transient Facility it scans the sky looking for things that go flash (or just change brightness) in the night.
34 days after the LIGO-Virgo gravitational wave detection, it was looking toward Coma Berenices and saw a flare, a brightening, coming from the distant galaxy AGN J124942.3 +344929 (let's call it J1249 for short). This is an active galaxy, one with a supermassive black hole in its heart (with a mass of 100 million times the Sun, so decently huge) that's actively eating material. As it falls in, this material gets very hot and emits quite a bit of light, making this galaxy visible even from its forbidding distance of 4.6 billion light years.
The flare detected lasted for about 50 days. Over that time, the amount of energy it emitted was huge, about 1044 Joules — more energy than the Sun will emit over its entire lifetime!
Active galaxies are known to flare, but this particular one had been quiet, or at least had a steady luminosity, for a year around the flare. Looking at the galaxy's history (and the histories of other galaxies) and applying some statistics, astronomers found the likelihood that this flare is due to some intrinsic variability in the galaxy as less than 0.01% — one chance in ten thousand. So, very low.
What does fit, however, is a "small" black hole careening through space at high speed and slamming into the disk of material around the supermassive black hole in the center of J1249. An event like this would create huge shock waves in the disk, causing it to heat up substantially, blasting out radiation for weeks. Gas falling into the smaller black hole would also pile up around it, heat up, and give off a lot of light as well. These are extremely powerful events, releasing about as much energy as the Sun will over its entire lifetime.
OK, you're possibly thinking, so what? What does this have to do with S190521g, the black hole merger that happened a month earlier?
Ah, here's the very fun part. If two black holes of unequal mass merge, the blast of gravitational wave energy is emitted in an asymmetric way: More of it is sent to one side than the other. This produces a huge force on the remnant merged black hole, giving it an extremely powerful kick — remember the energies we're talking about here are vast — and it can be accelerated to high velocities through space.
And if that black hole merger happened near the center of a galaxy, say one with a supermassive black hole and a huge disk of matter swirling around it, then it could cannonball right through that disk at high speed, creating shock waves that could heat the material up and create a gigantic two-month-long flash of light, one bright enough to travel over 4 billion light years and still be seen by a telescope in California.
Incredibly, the math works out. If the gravitational wave event S190521g was from two black holes in the active galaxy J1249 merging to form a single bigger black hole of about 100 times the Sun's mass, the kick from the powerful waves could accelerate it to a staggering velocity of 200 kilometers per second — over 700,000 kilometers per hour! Then, a month or so later this black hole plunged through the disk around the supermassive black hole, and the above events ensue. The amount of energy emitted from such a catastrophe would be about what was seen, and last for as long as the flare from the galaxy.
This gets better. You need a pair of binary black holes to merge, and black holes form when massive stars explode. But it's hard to make black holes with around 50 times the Sun's mass each (or say 80 and 20). You'd need a star bigger than any we've ever seen. So maybe the two black holes formed separately, from two independent stars, grew over time by eating stuff around them, and then somehow came together to form a binary system later. Good idea, but that's almost impossible to do out in empty space. It's incredibly unlikely, and even then it's hard to get them to orbit one another.
There is a place in the Universe where something like that is way easier, though: Near a supermassive black hole! There can be lots of massive stars orbiting it that turn into black holes, and plenty to feed them. Getting them to orbit one another is easier when there's lot of other stuff around that can nudge them together safely. In fact, it's likely that black holes that massive are themselves the result of mergers of previous black holes. Even if they got one of those off-center kicks, the gravity of the much larger supermassive black hole nearby would keep them from shooting out of the galaxy.
So this story hangs together from start to finish. It actually could be correct.
Now mind you, this is all circumstantial. The math works, and the physics too, but that doesn't mean this is what happened. I was very skeptical when I saw the press release. But reading the paper, I have to think this Rube Goldberg series of events is actually the most likely scenario.
I'd say "incredible" but it's literally credible. Astonishing, certainly.
So the merger didn't produce a flash of light directly, but may have started a series of event what resulted in a whopping huge flash of light a month later. Amazing. And if true, it means this could be happening many times in the Universe, which means we can't just look for immediate flashes from mergers, but also ones that happen weeks or months later.
Incidentally. Given all the physics involved, the astronomers think that the final black hole from the merger is still in orbit around the supermassive black hole, which means it may plunge through the disk again in the future. And not that long, either: Just 1.6 years after the first event, so late 2020/early 2021. You can bet astronomers will keep their telescopic eyes on the galaxy J1249. If the fireworks start up again, we'll have a pretty good guess as to their cause.