▲ 80 points
back
52 comments
LIGO is very serious about doing blinded test runs, even going as far as writing the whole paper before they check whether the signal was real. There's only a couple of people in the entire collaboration who even know when the blinded test runs occur. In any case, this rumor has been going around for a couple of months, so if it's true, we should know soon. Between this and the LHC blip, I'm quite excited about the next couple of months.
Another level of experimental blinding would be to avoid conferring with the other observatory (there are two components to LIGO, one in Hanford, WA and one in Livingston, LA) for confirmation/correlation until after one site's evaluation is complete. I don't know whether this is the current practice.
In any case, if the observatories are actually able to detect gravitational waves and if there are no theoretical obstacles, eventually two black holes will merge, which will produce a completely fantastic observation consisting of a rapidly rising frequency (as the two BHs circle each other closer and closer, shedding their orbital energy as gravitational waves), followed by a very loud cosmological pop as the BHs finally collapse into each other.
My impression (from well off to the side in a tangentially related field, mind you) has been that false positives are common enough in the individual detectors that they really do need the comparison with the other location's data as a filter. (Not common enough for an actual coincidence to be at all likely, though.) I'm happy to be corrected if I'm wrong, of course! But I think the system has only been designed to be reliable as a whole.
> But I think the system has only been designed to be reliable as a whole.
Yes, absolutely, I only meant that initial evaluations might be more objective if the two observatories cannot compare notes (at first) -- that might lead to a kind of confirmation bias. Only after one observatory satisfies itself that they have a sufficiently unambiguous signal should they be allowed to correlate with the other observatory's results.
Considering the absence of plausible signals to date, this idea might be bending over backwards to avoid observational bias.
I see your point, but I still have a sense that the false positive rate is just too high. I don't know the actual false positive rate, but I've somehow pictured maybe a dozen or so very brief (<1 sec?) signals per year. Under anything like those circumstances, an accidental coincidence would still be vanishingly unlikely (since there are 3x10^7 sec/year, it would be something like a one in a million chance), but the collaboration can't be expected to do full, detailed analysis of false signals a dozen times a year.
> ... I still have a sense that the false positive rate is just too high.
That might be true in principle, but considering what's at stake, I think the investigators will bend over backward to avoid making a claim that isn't ironclad. Remember that in real science, investigators look for explanations that will falsify their results, and only give credence to results that resist sincere efforts at falsification. (Contract this to the outlook of a pseudoscientist, who only looks for reasons to confirm his results.)
Remember the incident in which some investigators claimed that neutrinos were exceeding the speed of light? In truth, they were looking for reasons to confirm their result, not falsify it. They were wrong, the real cause was a bad connection and ... they were pressured to resign their positions.
https://en.wikipedia.org/wiki/Faster-than-light_neutrino_ano...
> ... the collaboration can't be expected to do full, detailed analysis of false signals a dozen times a year.
That's true, but they could instead choose a handful of genuinely peculiar signals, signals unlikely to have natural causes, and try to falsify those.
As above, there are a handful of cosmological sources that would produce -- not a brief signal -- but a pattern of signals, like a rising tone created by merging black holes. Those would be more interesting, as well as being easier to compare to false terrestrial sources.
> eventually two black holes will merge ... as the two BHs circle each other closer and closer
I've heard this so many times, and I still don't see how it's possible! Because of time dilation this merger will take infinite time.
No one has ever explained this satisfactorily to me. Why do people just seem to completely ignore relativity when dealing with black holes?
This is the exact same conceptual issue as the formation of a single black hole (BH). Consider a spherical shell of matter (i.e. GR dust) collapsing inwards to form a BH. As the dust particles get closer and closer together, one can roughly say that time slows down for them and that none of the particles ever pass through their mutual horizon. Rather, the argument goes, the particles just becomes exponentially close to where that horizon would form. From the outside, this is indistinguishable from a BH for all practical purposes, since the only way to isolate the still-free dust particles is to zoom exponentially close to the horizon and then drag yourself out (which requires an exponential amount of energy).
The problem with this argument is that there is no preferred global time coordinate with respect to which you declare that "the BH has formed". Rather, the spacetime events corresponding to "each dust particle crosses their mutual horizon" (i.e., "the formation of the BH"), along with the rest of the interior of the BH, comprise an additional spacetime region you can choose to attach to your manifold or not. This is a free choice because causality flows one-way across the boundary; no experiments outside the BH can be sensitive to whether the region "really" exists. There are multiple ways to extend the normal time coordinates (at infinity) into the interior of the BH, and a natural choice is the proper time of in-falling particles/observers. In that case, the BH definitely forms at a finite values of that time coordinate.
Merging two existing BHs doesn't introduce new fundamental issues. Quantum mechanics does introduce new fundamental issues, but formally the jury is still out on whether it resolves this question. The majority opinion is that the interior of the BH exists just as much as regions beyond your cosmological horizon (in an expanding cosmology) exist, from which you are likewise causally disconnected
> This is the exact same conceptual issue as the formation of a single black hole (BH).
I know. No one has ever explained how those can form either.
> From the outside, this is indistinguishable from a BH for all practical purposes
Except without the physics breaking properties of black holes. So not the same at all.
> The problem with this argument is that there is no preferred global time coordinate with respect to which you declare that "the BH has formed".
I don't need a global time, my own time frame is perfectly fine with me. If a black hole can never form, as viewed by me, then black holes don't exist.
> This is a free choice because causality flows one-way across the boundary
This presupposes the black hole exists already, but we have not gotten there yet. First it has to form, and it can not.
> and a natural choice is the proper time of in-falling particles/observers. In that case, the BH definitely forms at a finite values of that time coordinate.
It gets interesting here - what does happen from the POV of that particle? I suspect it never sees a black hole. This particle is moving exponentially close to the speed of light, relative to the center of mass of the forming black hole.
This means that time is stopped for the center of mass of the black hole from the POV of the particle, which means the center of mass can never form a black hole since it can never do anything at all (since time is stopped for it).
> If a black hole can never form, as viewed by me, then black holes don't exist.
But they do -- the evidence is copious:
http://www.galacticcenter.astro.ucla.edu/animations.html
Arguing that, because the event horizon collapses the time dimension (from the perspective of a distant observer), therefore BHs aren't real, is like saying that sum(2^-n,(n,1,oo)) can't equal 1 because it's an infinite series and an infinity of additions takes too long.
Think of black holes as objects of great mass, that have very bright accretion disks, surrounding an event horizon that is perpetually almost forming a singularity but never quite completes it (from the perspective of someone far from the horizon).
> This means that time is stopped for the center of mass of the black hole ...
From the perspective of a distant observer, the time horizon is located at the event horizon.
> But they do -- the evidence is copious:
The evidence is non-existent. The only evidence is for supermassive objects, not event horizons.
> Arguing that ... BHs aren't real, is like saying ... an infinite series and an infinity of additions takes too long.
That's a terrible argument. One is an actual physical process, the other is a mathematical abstraction. Do you not understand infinity? Infinity means you can't actually do it. It doesn't mean you can't calculate it.
> Think of black holes as objects of great mass, that have very bright accretion disks, surrounding an event horizon that is perpetually almost forming a singularity but never quite completes it (from the perspective of someone far from the horizon).
Eh? If it never forms a singularity, it has no event horizon. i.e. it has none of the physic breaking properties of black holes.
And accretion disks don't have to be bright BTW, as stuff falls in time dilation makes them much less bright (they "slow down", and emit photons less often).
> From the perspective of a distant observer, the time horizon is located at the event horizon.
When I say center of mass of the black hole I am naming the object I am talking about, not describing the location of the time horizon. Sorry if it was confusing.
I appreciate that you refuse to accept an explanation until it completely makes sense to you, but it sounds like you're attached at a gut level to some basic misconceptions. (In particular: it's not just that local clocks appear to run at different rates, it's that ticks of two separated clocks do not always have a meaningful one-to-one mapping, so it's often not possible to assign a sensible time that a particular event happens according to a clock held by an outside observer.) Those misconceptions won't be cleared up unless you're staring the math in the face.
Grab an intro GR textbook and work through it. There are free ones online.
I did, after I studied it more is when I realized black holes can't form. That's when I found that others have noticed it to, and called it an "eternally collapsing object" (ECO).
> so it's often not possible to assign a sensible time that a particular event happens according to a clock held by an outside observer.
That's not completely accurate. The two can not assign the SAME time to some event happening - it might happen at different times for the two.
But it IS possible to say when the even will happen from ONE of their points of view. It just might not be the same time as what the other one sees.
> Why do people just seem to completely ignore relativity when dealing with black holes?
It's true that, at and below the event horizon, spacetime curvature is such that time hasn't its usual meaning, but this doesn't mean two BHs cannot orbit closer and closer and eventually merge. Such events take place primarily outside the event horizon. This video --
http://www.galacticcenter.astro.ucla.edu/animations.html
-- diagrams stars orbiting close to the BH said to be at the center of our galaxy (in fact these orbits give us an estimate of the BH's mass).
So if stars can orbit a BH, then a BH can orbit a BH, and the latter would orbit in normal spacetime, at least until the merger.
> Because of time dilation this merger will take infinite time.
Only from the perspective of the event horizon and below. Remember that "above" the event horizon, time still has its usual meaning, even though it's greatly modified compared to time on Earth's surface (which is also slightly dilated in GR).
They can orbit, yes. But they will take infinite time to actually merge though.
> Remember that "above" the event horizon, time still has its usual meaning
The closer they get to that event horizon the slower time goes, so they will take longer and longer to orbit.
> But they will take infinite time to actually merge though.
Yes, from the perspective of a distant observer. From the perspective of an observer falling into the BH, time will pass in the usual way, indeed it must.
> The closer they get to that event horizon the slower time goes, so they will take longer and longer to orbit.
Only from the perspective of a distant observer. This might explain why black holes last as long as they do, from the perspective of a distant observer. For all we know, they all blow up in short order from the perspective of an observer falling into one.
I am speaking about the distant observer though. I am speaking the search for gravity waves.
If, us, that distant observer, never see the black holes merge, then what does that mean for the gravity waves?
It would take infinite time to an observer outside the horizon. It would probably look something like a cell dividing, in reverse, over infinite time, such that the end result is indistinguishable from a single hole.
As I understand it, time only "slows" from the BH frame. From an observer POV it would happen as described.
Other way around actually.
Can you explain that a bit more? I'm fascinated but don't quite get the rationale behind having not even the authors of the paper know the results. What is their process?
It means you have to develop your experimental method prior to seeing the data, so you can't try a dozen different ways of slicing that data up until you get the result you want.
...and now knowing this, I wish it were the approach for _all_ scientific research.
Yeah. This is what's absolutely missing in many scientific fields, and seems to be the reason for so many results turning out to be utter bullshit (hello, psychology). There's a strong push coming from research community to make this approach, which is known as 'study pre-registration', mandatory.
Presumably pre-registration could also help with positive results bias.
Assuming the registration was made public, one could mandate that something had to be produced from the study, if only a statement that nothing was found.
Wouldn't solve the problem but would presumably aid meta-studies?
It's not really feasible for most research (although there are plenty of needed improvements along these lines). Big physics projects are unusual in that they are enormous undertakings of 100s or 1000s of PhDs, and few or no other experiments will take data that can cross-check the results. (For instance, no other machine will produce the conditions at the LHC for several decades at least, which is why they go to the fantastic expense of building two completely separate general purpose detectors.) Devoting ~5 full-time PhDs for the sake of super-duper methodological rigor is doable for LIGO, but not for smaller experiments.
Well, keep in mind that if hundreds of scientists take their chance on the same data using different methods, we basically have the same problem. One of them will be the lucky one that took the approach that shows the favorable results, whether the results are truly there or not.
Source for the blind injection of signals:
"In 2014, researchers on another US experiment, called BICEP2, [0] called a press conference to announce the discovery of gravitational waves, but others have since pointed out that the signal could be due entirely to space dust."
The BICEP2 team thought they detected gravitational waves with their Cosmic Microwave Background (CMB ) measurements from their Antarctic observatory.[0],[1],[2] By chance the ESA had the Planck observation satellite measuring a broader spectrum. Matching the two observations the BICEP2 measurements were capturing light interference from space dust in-between the observatory and the source. [3]
If gravitational waves are found, this will be the biggest thing since Smoot and COBE [4] and confirmation of the Gurth Inflation theory. [5]
[0] http://www.theguardian.com/science/2014/mar/17/primordial-gr...
[1] https://www.cfa.harvard.edu/CMB/bicep2/
[2] https://www.cfa.harvard.edu/CMB/bicep2/science.html
[3] http://www.theguardian.com/science/2014/sep/22/gravitational...
[4] http://aether.lbl.gov/www/projects/cobe/ and https://en.wikipedia.org/wiki/George_Smoot
[5] http://www.nytimes.com/2014/03/18/science/space/detection-of...
> The BICEP2 team thought they detected gravitational waves with their Cosmic Microwave Background (CMB ) measurements from their Antarctic observatory. By chance the ESA had the Planck observation satellite measuring a broader spectrum. Matching the two observations the BICEP2 measurements were capturing light interference from space dust in-between the observatory and the source.
This is a bit different, though. BICEP2 was claiming a detection of a polarization signal in the cosmic microwave background, which if present, would point to existence of primorial gravitational waves resulting from cosmic inflation. But, as you note, the multi-wavelength observations of Planck (not taken by chance, but part of a carefully planned all-sky survey with somewhat coarser resolution compared to BICEP2) showed that the polarization signal measured by BICEP2 was consistent with foreground dust, rather than primordial gravitational waves.
> If gravitational waves are found, this will be the biggest thing since Smoot and COBE and confirmation of the Gurth Inflation theory.
It will be huge, but the gravitational waves LIGO would detect are not primordial or from inflation, but rather from supernovae and compact binary mergers (neutron star pairs, stellar-mass black hole pairs, etc.). So a LIGO gravitational wave detection would not provide any evidence for or against inflation. But it would tell us a lot about stellar explosions and the evolution of compact binaries.
EDIT: I should note that LIGO could detect or place constraints on some "exotic" gravitational wave sources – cosmic strings, for example.
"It will be huge, but the gravitational waves LIGO would detect are not primordial or from inflation, but rather from supernovae and compact binary mergers (neutron star pairs, stellar-mass black hole pairs, etc.)."
I missed this. How do they tell the difference? (Interferomety?)
Generally, the difference is direct vs indirect detection. LIGO aims to directly detect gravitational waves as they pass Earth by measuring the change in length of two parallel arms (using interferometry). In contrast, BICEP2 was looking for a polarization signal in the cosmic microwave background (CMB) which would occur if there were primordial gravitational waves (but this signal could also be mimicked by foreground dust).
So BICEP2's claimed detection wasn't a direct detection of gravitational waves. BICEP2 detected polarization (specifically B-mode polarization[0]) in the CMB. The BICEP2 team initially thought the amount of foreground dust was too low to explain the degree of polarization they observed, so they attributed the polarization to primordial gravitational waves affecting the cosmic microwave background. As you noted, Planck had multiple bands and so could more effectively determine the amount of foreground dust. Using a joint BICEP2+Planck dataset, the teams determined the signal could be explained by foreground dust.
In contrast, LIGO is attempting to directly measure gravitational waves from:
1) discrete, individual events, such as a merger of neutron stars. As the objects inspiral, they emit gravitational waves, with the emission becoming stronger (and, presumably higher frequency) as the objects near final merger. At final coalescence, a burst of gravitational waves is emitted. It is hoped/expected that this burst (plus some GW emission leading up to the merger) will be detected by LIGO for individual sources.
2) A gravitational wave background from the ensemble of events too far away to detect individually, but whose sum adds up to enough to be detectable. In contrast to the above, this will be some linear combination of all the passing waves from distant individual events, from a variety of directions.
The frequency of the gravitational waves depends on the bodies involved. With LIGO's 2km arms, it is sensitive to a range of frequencies which correspond to expectations for stellar-mass compact binaries. The LISA project[1] would have longer baselines and so would be sensitive to mergers of more massive objects, such as supermassive black hole pairs. So, the baseline of your gravitational wave experiment determines what you can possibly measure. This graphic shows at which frequencies various phenomena emit[2]. Based on that graphic, LIGO may be sensitive to the frequencies corresponding to primordial gravitational waves, but they are likely far far below LIGO's detection ability. My impression is that direct detection of primordial gravitational waves will not be possible for a while, if ever. Rather, they may be detected indirectly, via the imprint they leave on the CMB polarization (as BICEP2 tried).
[0] https://en.wikipedia.org/wiki/Cosmic_microwave_background#Po...
[2] https://en.wikipedia.org/wiki/File:The_Gravitational_wave_sp...
EDIT: Added new first paragraph and reworded the rest for clarity/flow.
thanks @privong, this is a first rate explanation of LIGO/BICEP2 and CMB and a major reason I hang out on HN.
Sure thing, @bootload.
By the way, I should have mentioned that we already have indirect evidence for gravitational waves, from the orbital decay of binary pulsars such as PSR B1913+16 [0]. The observed orbital decay has an excellent match with the orbital decay predicted using general relativity and gravitational waves. [1] shows the agreement, with the line being a theoretical prediction and not a fit.
[0] https://en.wikipedia.org/wiki/PSR_B1913%2B16
[1] https://en.wikipedia.org/wiki/PSR_B1913%2B16#/media/File:PSR...
Off-topic, but if anyone has a chance to go tour the LIGO facilities, definitely take it! It is a very interesting experience, and highly educational, even for those that don't have a physics background. I've been to the one in Livingston while studying at LSU, and even after several years, I'm still astounded at the precision of the technology used there.