"We shall not cease from exploration, and the end of all our exploring will be to arrive where we started and know the place for the first time." -T.S. Eliot
Now that not just one but two gravitational wave events have been directly detected, we're officially in the era of true gravitational wave astronomy. LIGO has taught us something unique about stellar mass black holes, and will continue to teach us about these objects, their population statistics, and their merger rates as time goes on.
Image credit: Bohn et al 2015, SXS team, of two merging black holes and how they alter the appearance of the background spacetime in General Relativity.
But beyond that, we're poised to learn the origin of gamma-ray bursts, to observe neutron star quakes, and once we go to space, to observe supermassive black holes and possibly even the relic gravitational waves from cosmic inflation. The latter would do more than confirm what set up the Big Bang; it would prove that gravity is an inherently quantum force!
Illustration of the density (scalar) and gravitational wave (tensor) fluctuations arising from the end of inflation. Image credit: National Science Foundation (NASA, JPL, Keck Foundation, Moore Foundation, related) – Funded BICEP2 Program.
- Log in to post comments
More like this
A minor correction:
If one divides a length by a speed, the result is in units of time, not frequency. Indeed, 'millisecond' is a period, not a frequency. Ethan means either 'a period in the range of milliseconds' or 'a frequency in the range of hundreds of Hertz'. If the latter, then the first sentence needs to be 'dvide the speed of light by the path length of the detector'
Hi Ethan,
I've loved your column for a couple years now and you continue to write about so many great topics each week. I'm not sure of the best place to ask this but hopefully this is it. Perhaps you've even answered this in another column I've somehow missed.
I am in awe of the stat about more energy radiating out for a few milliseconds than is output in the rest of the observable universe when 2 black holes collide.
My question is this: assuming there would be a distance safe enough from the black hole merger that you wouldn't be fried by the 'normal' radiation from this merger (X-rays etc as well as being well outside the event horizon), what would a casual observer experience from the ripples in the gravitational waves in space-time itself? If this safe distance is possible, would there be sudden dramatic changes in gravity as the peaks and troughs of the gravitational waves pass through, tearing you apart, or would you be able to even perceive space time compressing in one direction, since you are a part of that space time? Forgive me if I am missing something basic in my understanding.
Thank you for considering my question!
Kind regards,
O
“LIGO has taught us something unique about stellar mass black holes, and will continue to teach us about these objects, their population statistics, and their *merger rates* as time goes on.”
After the second wave detection in December, LIGO taught us the black hole merger rate was one every 66 days, approximately.
But I guess the rate has dropped substantially since then. How low do they think the BH merger rate will go, and why?
@#3
You seem to have not noticed that LIGO is running at 30 odd% of sensitivity so far in its present state. Results are going to change as that sensitivity increases. The future of the project will bring more data as it is moved into space, where the sensitivity can be increased immensely over earthbased equipment.
@See Noknowledge #3: You continue to demonstrate your ignorance, of both statistics and of how scientific research, analysis, and publication works. I recommend you read a simple article on the Poisson distribution, then come back with a more intelligent question or comment.
To PJ #4:
“You seem to have not noticed that LIGO is running at 30 odd% of sensitivity so far in its present state.”
That is correct, I hadn’t noticed that, because I hadn’t read Ethan’s Forbes piece at the time of my comment.
And as of right now, I’ve only skimmed it.
But I have some more questions.
You say “Results are going to change as that sensitivity increases.”
If when running at only 30% sensitivity LIGO picked up waves from two BH mergers that occurred only 66 days apart,
then the rate of black hole mergers would be expected to increase with increased LIGO sensitivity, yes?
Would LIGO be surprised by a BH merger rate of, say, one every 10 days?
…………
You say “The future of the project will bring more data as it is moved into space, where the sensitivity can be increased immensely over *earthbased* equipment.”
And Ethan says “But the biggest advances will come from taking our gravitational wave ambitions into space. In space, you’re not limited by *seismic noise, rumbling trucks or plate tectonics*; you have the quiet vacuum of space as your backdrop.”
But surely the GW measurements coming from the sensitive detection equipment are not bothered by earth-based “seismic noise, rumbling trucks or plate tectonics.”
LIGO must know with certainty that none of those things impacted its news making measurements. Doesn't it?
And if being earth-based had no effect on LIGO nailing two waves in 103 days, when running at only 30% sensitivity, why the rush, why the need, (and why the cost), to put LIGO into space?
@#6
Yes, there could be an increase in detections, dependent on their quantity and period; remembering these are going to be totally random in nature. I cannot answer for LIGO regarding any 'surprises' at a higher rate of detection. You would have to refer to their site for deeper information.
Regarding earthbased equipment, it has been desensitized to outside interference by the design of its footings, mounting methods, etc. The inference of moving it into space alleviates the necessity to isolate the equipment from these earthly problems. There will be no need to compensate for earth curvature in each arm. The list goes on.
I do not believe LIGO is in a rush, just looking forward to the future of the science to further our knowledge.
SN,
Your conclusions about the rate of mergers are way off base. Consider an experiment. Roll a normal 6-sided die a million times. Observe how many rolls there are between each time you roll a 6. We know that the average rate of 6-rolling is once out of every six rolls.
However, that's an average rate. Individual observations will vary quite greatly around that average. There will almost certainly be occasions where in a million rolls, a six will be rolled on consecutive rolls. There will be, with probability of around 92%, occasions where a six will be rolled and then the next one will occur 65 rolls later (see below for calculation). Obviously, the estimate we make on the rate of six-rolling would be highly dependent on which of these situations we happened to observe if we only observed one data point. We would, in the first instance, conclude that sixes would be rolled EVERY roll. In the second, we would conclude that a six being rolled is a very rare event indeed. The truth can only be ascertained by observing MANY rolls to determine the frequency of six-rolling.
I hope the analogy is clear. We are now in the position of having made a single observation of black hole mergers occurring 66 days apart. That could be analogous to the case of consecutive sixes and be an unusually closely-spaced occurrence of the mergers or it could be analogous to the case of 65 rolls passing between sixes, in which case we would be underestimating the frequency of black hole mergers based on our meager data. Only continued observation of black hole mergers, possibly with better detectors can truly shed light on the frequency of these events.
--Note - the calculation of a 92% probability of observing the next six 65 rolls later is pretty straightforward. The probability on any given roll of rolling a non-six is 5/6. The rolls are independent of each other, so the probability of rolling a non-six on 65 consecutive trials is (5/6)^65, or about 1 in 140,000. The probability of this occurring in 1 million trials can be calculated from the binomial distribution using this probability. The probability of zero occurrences of 65 straight non sixes in a million rolls is about 8%, so the probability of at least one such occurrence is 92%. (I am simplifying here; I realize that there aren't really one million chance to roll 65 consecutive non sixes in a million rolls, but my result should be a decent approximation).
@See Noevo
Based on your many comments on Ethan's articles you have a bit of a reputation as a troll, so I hope I am not contributing to the feeding.
A good troll will do at least a tiny bit of work before commenting (reading the article, for example). A lazy troll won't bother before asking his own questions.
I want to assume you are asking genuine questions, but not reading the article before commenting casts doubt on that genuineness.
Others have done a great job of trying to explain the probability of events, so I will focus on your other questions.
It's really easy to spend a short time googling to see how LIGO works, which would answer many of your questions. For example, LIGO detects compressions in space-time due to gravitational waves of a width of 1/1000 of a proton. A passing truck would generate many time that disturbance, and it is impossible to isolate the measurement system (LIGO's laser that travels 4,000km each measurement) from these disturbances. So to correct for them, there are many external measurements that are taken so the data from the laser can be 'cleaned' of that disturbance. Yes, the LIGO scientists have thought of this and so you don't need to be skeptical that the data might be corrupt. [side note: a related topic for how astronomers use data correction techniques to improve data collection is called 'adaptive optics'. Feel free to google]
I'm not sure why you ask 'what's the rush' to put this in space. I don't think there are concrete plans to do so in the immediate future; Ethan was just talking about what is possible in the future and what we could gain by doing so. As noted in your quote, the point of a space-based LIGO is to increase sensitivity over earth-based sensors (limited by said disturbances) by factors of thousands, to be able to detect the gravitational waves leftover from the Big Bang. Reading more about how LIGO works would have answered that for you.
To Sean T #8:
Thanks for trying to explain dice and probability to me.
I think I remember seeing such stuff in the 3 or 4 statistics courses I took in undergrad and grad many black hole mergers ago.
I also like to play craps!
Now, since Ethan may never answer me on this, maybe you could. Ethan says “LIGO has taught us something unique about stellar mass black holes, and WILL CONTINUE TO TEACH US ABOUT these objects, their population statistics, and THEIR MERGER RATES as time goes on.”
What has LIGO taught us about their merger rates SO FAR?
..........
P.S.
As I recall, the first wave was detected shortly after LIGO went online. (Actually, LIGO may not even have been fully online. I think I remember reading LIGO was still in ‘engineering’ phase or a test phase or something.) So, LIGO goes partially online, and at only 30% sensitivity, and almost immediately… BINGO!
Or rather, snake eyes!
You know. You don't have much time and you walk up to the table to make one bet, on the 2, and it hits!
Your $20 becomes $620!
(In millions, that would be enough to cover the cost of LIGO.)
What a fortunate roll of the dice, huh?
P.P.S.
I myself never bet on the 2.
@Seeno
Ok, I take my previous comments back; your troll level is pretty high. After 3 or 4 statistics courses you still don't seem to understand the nature of probability!
Prior to LIGO we had no way of directly detecting gravitational waves. So we had estimates for what the expected rates of events large enough to produce detectable gravitational waves would be.
SO FAR, LIGO has taught us that two confirmed events occurred 66 days apart. It has taught us much more about those specific events *themselves*, but not much else about the frequency of such events, as Sean T. tried to explain. You say you took many stats courses; therefore you must know that small sample sizes are not very useful in predicting anything. A sample size of two is completely useless. So if you have so many stats courses under your belt, why are you asking these questions?
If you are really just being skeptical that these events are genuine, solely based on the fact that expected probability predicts them to occur far less frequently than what has been observed, then that is a conscious choice to ignore how the scientific method works. Or you are trolling.
Continuing to feed, I'll use another example. Southern Alberta has experienced 2 '100 year' flood events in the past 20 years. Does that mean we need to change the definition of '100 year' flood event? No, and if you don't understand why, then I don't believe you passed your stats courses. [in this example I'm purposefully ignoring the potential need to revise event frequency based on climate change, which is a separate issue].
"Rush"? Plans for a space-based interferometric array have been in the works for 30 years.
You might be surprised what a sample size of 1 can tell you.
Poor choice of words then. I am not a scientist myself, and this paper doesn't have Ethan's plain-language interpretation to help us laymen. However, you are correct; I am surprised.
From the sample size of 1, they improve on earlier estimates of between 0.1-300, to between 2-600, annual binary black hole mergers per cubic Gpsc. And 2-53 is their 90% range.
Which is exactly what Seenoevo asked for when he said 'what has LIGO taught us about merger rates so far'.
__________________________________________________
Following up on my #2 comment, I believe I've found (an) answer. A large gravitational wave would pass through very quickly, and would momentarily cause different parts off your body to experience different levels of gravity, potentially similar to the spaghettification experienced close to a black hole.
I'd then like to refine my question: given what we assume is an 'average' BBH merger (or even taking the aspects of the December merger LIGO observed), is there a distance where the casual observer would be safe from electromagnetic radiation, far enough outside the normal event horizon for the binary pair before and after that relativistic speeds are not required to escape, but the GW from the merger would spaghettifi them? How far would they need to be to survive the gravitational waves passing through? At that point would it feel something like a carnival ride for a short time, as the changes in Gravity passed through, or would it be too fast to notice?
Thanks!
"I think I remember seeing such stuff in the 3 or 4 statistics courses I took in undergrad and grad many black hole mergers ago."
You've repeatedly demonstrated you don't understand any of it, even if we believe you actually took such classes.
And the classes you would have taken, given what you've said about your background, would top out at undergrad complexity anyway, even if you would have received graduate credit.
@See Noevo
For your craps example:
LIGO isn't walking up to one table and betting on snake eyes on the first roll. Instead, it walks into one casino, and has the ability to detect when any craps game in the casino rolls snake eyes. It doesn't know how many games are happening at the same time, but it can infer that based on enough data. Eventually it will have the ability to detect any snake eyes rolled at any table in Nevada. And the space-based version might be able to tell when a 2 is rolled from any pair of dice on the planet. This isn't the type of 'random chance' that you are thinking when you compare it to betting on a dice roll, it is all about probability.
Would these happen to have been the same course over and over again? Quick: Can a Box–Cox transform be validly applied after z-scoring? I had to ascertain this for a freaking invoice, sans the "3 or 4 statistics courses."
Oh, and what was your "Ivy League" master's degree in?
To Outtatheblu #11:
“Prior to LIGO we had no way of directly detecting gravitational waves. So WE HAD ESTIMATES for what the EXPECTED rates of events large enough to produce detectable gravitational waves would be.”
And what were those estimates for the expected rate?
[Feel free to google.]
“You say you took many stats courses; therefore you must know that small sample sizes are not very useful in predicting anything. A sample size of two is completely useless.”
Well, let’s not get into a lot of conehead math and stats right now. Let’s just consider a small sample of readily accessible, everyday observations, and see if they might imply anything.
1)Prior to LIGO’s news-heard-round-the-world, LIGO, according to you, had estimates for the expected rates of gravitational wave-producing black hole mergers.
2)Neither you nor anyone else in this thread has yet to say what those estimates were. [Feel free to google.]
3)$620 million is spent trying to make LIGO operational.
4)Within a short time of becoming just partly operational, LIGO says it detected not one but two GWs, which it claims are due to black hole mergers of very specific masses and at very specific (and very great) distances.
5)We’re told that neither snow nor rain nor heat nor gloom of night stays the USPS from delivering your junk mail, and that neither will those things (along with earth-based “seismic noise, rumbling trucks or plate tectonics”), stay LIGO’s detection of BHMGWs.
6)No one here has said whether LIGO was surprised by the plentitude of its discoveries.
7)And no one here has said what the estimated rate of BHMGWs was or is, but only that the estimate is bound to increase with space-based LIGO.
I’m not sure what these observations imply, but it’s getting to be time for dinner anyway.
It’s tough to “troll” on an empty stomach.
Trivially false.
SN
Just to extend my previous analogy: even a small number of observations can provide some information about frequencies. Since you've had 3 stats courses, I don't need to explain to you what the chi squared distribution is. One can observe a couple of spacings between sixes in my analogy, set a value for the actual probability of rolling a six, and then use chi squared to calculate the probability that you'd get the observed results given the selected value for true six-rolling probability. You could then find a confidence interval for the true probability.
In any case, I am unsure what your point is. Suppose that the observed merger rate turns out to be much higher than the theoretically expected rate. So what? The theoretical rate is wrong, therefore God did it all and we might as well just give up on acience? I'm sure that astronomers studying the phenomenon will come up with some idea that explains the observation.
In case you haven't been paying attention during your time here, that's the way science is SUPPOSED to work. Come up with some idea. Make observations to test it. If the observations show it to be false, modify the idea or abandon it. Sometimes this is indicative of major changes in our understanding of the universe (eg. unexpected perhelion shift of the planet Mercury, the details of black body radiation spectra, or the photoelectric effect), but mostly such instances only change some of the finer details of our theories. I am no expert in astrophysics, but I suspect a higher than expected black hole merger rate probably only modifies the fine detail, not our basic understanding.
For an analogy you'd relate to better, suppose a new archeological study found that the entire New Testament was true in every detail except one. An example of the first kind would be that Jesus was not actually resurrected. If you would know definitively that the resurrection dod not occur, it would certainly cause major changes to Christian theology. An example of the second type would be an error in the geneology of Jesus outlined in the Gospels. Yes, that would be an error, and Bibles would need revision, but this error would have no effect on the general understanding of Christian theology.
Then again, if the "Fritz papyrus" weren't bogus, it would have a quite an effect on the church that S.N. nominally affiliates himself with.
I take it that boxcars are only enough for one casino hooker.
"After 3 or 4 statistics courses you still don’t seem to understand the nature of probability! "
See Nowt has never taken a course.
It has a Walter Mitty existence. After all, it believes in a fairy tale as absolute truth because its entire personnae is a fiction.
Re: 17, Narad, no need to go so deep. High School maths is sufficient (and early high school at that).
See Nowt, how do you calculate a trend from a set of datapoints? What is the mathematical formula for the simplest Least Squares Fit line's trend?
This is something any schoolchild should know and would have been a required prerequesite for any further courses in the subject, and would have been continually reinforced by those courses, so you cannot have forgotten it.
So I haven't been to this website in a while and I wondered "I wonder if that obvious virgin, Wow, is still crying like a bitch in the comments section. (I wasn't disappointed)
Dude, go find out what pussy is. You'll like it. I promise.
In other words:
Bri: Whaaaaa waaaaa waaaaa!