[wolfram74]

Anyone familiar with this branch of astronomy want to explain why one detection in a volume on the order of 27 billion cubic light years is reasonable? Are they still processing data and will find more events? Is the sensitivity highly anisotropic so the detection volume is significantly smaller? Or are events like this just really conveniently rare that we get about 1 every data gathering interval?

[yodon]

Events big enough to be detected are quite rare, but the phenomenon you are asking about is more a financial reality than a "convenient coincidence" as you put it. There is a pretty steep curve connecting sensitivity and cost, so when the team that built LIGO was designing it, they used the best available models of colliding black hole event rates to estimate the sensitivity required to deliver a conclusive result in a reasonable amount of time.

If you're getting 10 events/second with a device like this, you probably overpaid for sensitivity and if you're getting 1 event per century you're probably not going to be able to maintain the operating expenses to still be running when the detectable event occurs (and, as critically, none of the people involved will be able to get the data they need in the time they need it to get their PhD's, assistant profesorships, or tenured positions, so you can't get the labor force you need for your experiment to work on it, which is really what sets the acceptable duration of most experiments in practice).

It looks like the original estimates were pretty good, so events are coming in at about the rate the experimenters hoped they would see them.

[mturmon]

Hmm...are the rates of black holes per volume well-constrained at all? I was under the impression that it's a possibility that dark matter consists prominently of primordial BHs? The truth or falsity of this would seem to have a big effect on rates.

[privong]

> are the rates of black holes per volume well-constrained at all?

There are estimates. But two of the three LIGO detections are of black holes that are more massive than we had expected to exist (~ a few tens of solar masses). Previously we had convincing examples of black holes with <~ 10 solar masses and others with >~ 1e6 solar masses. But since we didn't have any convincing observational detections of BHs with ~20-40 solar masses, it's safe to say that the volume density is poorly constrained for that mass range. At the high end we have a reasonable estimate of the volume density, because we think all galaxies with spheroid components have a black hole and that the black hole's mass is linked to the spheroid.

> I was under the impression that it's a possibility that dark matter consists prominently of primordial BHs?

It depends on what you mean by "primordial". Micro-lensing experiments (when a star is brightly made brighter by the gravitational focusing of light from an object passing between us and the star), mostly looking towards the LMC/SMC [e.g., 0] have tried to address this. My recollection is that there aren't enough stellar mass black holes around to account for all of dark matter. Assuming Hawking radiation exists, low-mass primordial black holes should have evaporated by now, leaving only the more massive ones. There's a range in between the two, but I'm not sure if you can fit enough of them in a galaxy to account for dark matter while still being consistent with the sensitivity of the microlensing surveys.

[0] https://en.wikipedia.org/wiki/Optical_Gravitational_Lensing_...

> The truth or falsity of this would seem to have a big effect on rates.

Possibly. Though in order to emit GWs, pairs of black holes have to become bound to each other. If black holes make up the dark matter halos, they probably have large velocities relative to each other, which would limit their ability to form bound pairs (though it is possible with 3-body interactions). I am not aware of estimates of the BH pair-formation rate in halos _if_ DM haloes are in fact made of black holes. But the event rate probably can't be extraordinarily high, otherwise we might expect to see dark matter halos becoming less massive as the Universe ages. Though there are many confounding factors that might hide such a signal.

[mturmon]

Thanks a lot for this reply.

[privong]

> If you're getting 10 events/second with a device like this, you probably overpaid for sensitivity

Aside from issues processing and disentangling the overlapping events in a situation with that high of an event rate, more events would not be bad, so I'm not sure I'd call it "overpaying". Imagine the kind of population demographics that could be built up if we were detecting that many events.

[specialp]

Keep in mind that when LIGO was built at tremendous expense, gravitational waves were never conclusively detected, and just to get to this sensitivity it was a feat of engineering. It was unknown how long it would be or if they would ever detect a wave.

Just detecting the initial wave was one of the most important measurements in the history of physics. Now that we know that gravitational waves exist, and can give us great insight into the mysteries of the universe, there will be more efforts to detect waves at smaller amplitudes and different frequencies. That is one reason the eLISA project is going on [1].

1. https://en.wikipedia.org/wiki/Laser_Interferometer_Space_Ant...

[privong]

I'm aware of the context for the GW detections and the expense, etc. I was making a general philosophical point about "overpaid", not commenting specifically on the LIGO cost-benefit analysis.

[yodon]

It's not that physicists wouldn't love to capture all those events, it's that the cost of building instruments like LIGO is nearly prohibitively high and the cost is a strong function of the sensitivity of the instrument. If you aim too high in your sensitivity aspirations, the cost hits a point where the experiment simply can't be funded.

[privong]

> If you aim too high in your sensitivity aspirations, the cost hits a point where the experiment simply can't be funded.

Agreed. I misunderstood your meaning then; I'd interpreted your wording to mean that "overpaid" was still within the bounds of reasonable expectations for funding. "Overpaid" didn't imply "too expensive to build", to me.

[davrosthedalek]

It also might not be possible to build the system more sensitive. Took them, what, 30 years or so, to get to this sensitivity.

[eanzenberg]

It's a balance between cost and sensitivity, and remember that 1 discovery would prove the experiment a "success". I think the way it was planned and executed was great. Also, I think the proof that gravitational waves is far more exciting than comparing gravity wave signatures among a sample of celestial collisions.

[dwaltrip]

This is sort of like saying that "proving that stars emit light" is more exciting than using better and better telescopes to compare electromagnetic wave signatures of different light emitting celestial objects.

We will likely learn a great many things over the coming decades with this new way of looking at the universe, many of which we couldn't have even guessed we would learn!

[eaq]

While there are many compact binary systems in the universe continuously emitting gravitational waves for a very long time, LIGO is only able to detect the most violent waves emitted by the final coalescence and merger. So, on the astrophysical side, there is some joint probability given by how common these systems are, and how likely they are to merge in a given time frame. (Space based observatories like LISA would be able to see the long-lived inspiral waves though.)

On the instrumental side, we've only just reached the sensitivity levels to make any detections in the first place, so it's not surprising that we're not getting a huge number of events (otherwise the previous generation of detectors would've seen something). In addition, each individual observatory has its own "antenna pattern", making us less less sensitive to certain sky locations. This will improve as VIRGO, KAGRA, and LIGO-India come online in the future.

[grigjd3]

It's not 100% accurate that "LIGO is only able to detect the most violent waves emitted" but that LIGO can only detect waves emitted in a certain frequency range (total mass) of binary mergers. For instance, LIGO pretty much cannot detect the merger of supermassive black holes.

[BurningFrog]

These waves are extremely small, and we've just managed to build instruments sensitive enough to detect the very biggest ones.

Earth bound instruments will no doubt get better, but to get a real jump in quality, you need instruments in the stillness of space: http://www.einstein-online.info/spotlights/eLISA

[ahnitz]

It's important to know that LISA and LIGO aren't really competing for sensitivity. Rather they complement each other by looking in different frequency ranges. The relationship between LISA and LIGO is analogous to a radio telescope and a gamma ray one. They observe different parts of the spectrum. At the frequencies that black holes merge for example, ground motion is not much of an issue, and other noise source dominate.

[ChuckMcM]

The info on eLISA is great.

I am just waiting for the $10,000 shielding for high end speakers to keep gravity waves from interfering with the acoustic purity of the sound they produce. :-)

[kevin_thibedeau]

You absolutely must keep your record player suspended from glass fiber in a vacuum chamber to avoid any unwanted coloration. Just takes 30 minutes to pump down when you want to flip sides.

[AnimalMuppet]

I will happily sell you such shielding for $10,000...

[InclinedPlane]

These events are actually not rare. There are roughly billions of them in that volume over the current lifetime of the Universe. They are comparatively rarer than many other kinds of events, like supernovae, because they involve the extremely massive stars that are themselves very rare.

[elorant]

I'm not a physicist/astronomer so someone could give you a more detailed explanation. From what I understand, LIGO has been designed to detect collisions of mid-sized black holes, and it looks like those aren't that common.

There's also a more paper on this issue which you can find here: https://arxiv.org/pdf/1704.04628.pdf