jump to navigation

The search for gravitational waves July 2, 2007

Posted by dorigo in astronomy, news, physics, science.

The title of this post is borrowed from today’s talk by Jim Hough at the plenary session of PASCOS 2007. So here follows my attempt to justify my presence in London today.

Jim gave a review of ground-based interferometers aimed at detecting gravitational waves with frequencies above a Hz (I sort of understood you cannot get below a few Hz with ground-based experiments due to an effective cutoff from seismic backgrounds).

I must say from the outset that what impressed me most in his talk was his first slide, which showed how LISA will have such excellent sensitivity to all the things he took as benchmarks in his later discussion, that one question remained in the back of my mind: if LISA works, will LIGO, VIRGO, and their friends be justified ?

Anyway, back to his talk. Gravitational waves are rippes in the curvature of space-time caused by violent acceleration of mass in events such as neutron star coalescences, black hole formation and interactions, cosmic string vibrations, or less violent ones like pulsars or binary stars.

The strength of the signal is the amplitude of the strain of space caused by the dynamics. Strains are extremely small: we are talking about effects of the scale of one part in 10^24.

Jim was clear that most everybody believes in the existence of gravitational waves, and he seemed to imply that their detection is only a matter of money and time. He stressed that there have been indirect detections of gravitational waves, one striking example being PSR 1913+16, a binary pulsar system which showed a loss of orbital energy via gravitational waves in a way so perfectly in agreement with that predicted by general relativity that no more discussion seemed necessary.

He stressed that gravitational waves have a small effect on atoms on earth. One expects movements of less than a trillionth of the wavelength of light (10^-18 m) over a 4 km baseline. The LIGO interferometer, with 4km arms, has a quite challenging task. However, there is actually a network of interferometers: LIGO, GEO600, VIRGO. LIGO is built with 3 interferometers with two 4 km arms in Hanford, Washington, and Livingston, Louisiana. Geo600 is a 600m interferometer near Hannover, comparable in sensitivity to LIGO because of a better laser technology. Then VIRGO is a 3 km device near Pisa, Italy.

Others around the world are tama300, in Japan (300m long, now undergoing upgrades). Aigo is another one in Australia.

These experiments are not small business: LIGO is run by 500 people from 55 institutions. LIGO and VIRGO actually signed an agreement to doing a joint analysis of their data. This increases the science
potential and can guarantee observation of the gravitational sky over the next 10 years, without down time.

If one considers the network sensitivity from these experiments, it is good, peaking at frequencies of 100 hertz, to amplitudes of some parts in 10^-23 /Hz^(1/2).

Jim said that these machines have already ran for some years. Upper limits have been set for a range of signals, and a lot of papers are out for you to read. We did not really expect to find signals with the limited exposure and sensitivity currently at hand, but black hole binary sistems could indeed have given detectable signals, with models predicting rates of up to 1 /(4 years).

Data analysis is tough: you need to decide on the shape of a signal you are looking for and do an optimal filtering.
Despite the limited sensitivity, one can observe hundreds of galaxies in search for black hole signals, since if the system is optimally located and oriented, the detection range of the ground-based inteferometers is long.
Other interesting sources are neutron stars and pulsars. Rapidly spinning ones provide a potential source of continuous gravitational waves.

97 pulsars have been looked at for 10 months. Experiments set limits on lowest ellipticity of neutron stars at 1.1E-7 for a particular pulsar, PSR J2124-3358, which is located 0.25 kparsecs away from us.

Jim then discussed the prospects for enhancing and upgrading the current experiments. The plan is to make upgrades interleaved with periods of data taking, to avoid missing a supernova signal. Sensitivity improvements are aimed at reducing all sources of noise.

LIGO and VIRGO plan to do incremental enhancements in 2007-2009. LIGO will most of all enhance its laser power and get a better optical readout. This buys a x2 sensitivity. VIRGO will also get a higher laser power, and install silica suspensions to reduce thermal noise (he did not explain what is the effect of silica suspension, nor where they are suspended, so I am as much in the dark as you are). Between 2009 and 2011 they will run, happily upgraded.

The “final” expectations of the detector upgrades are to see black hole-black hole coalescences at a rhythm of up to 500 per year with sensitivity improvements of 10 to 15, foreseen in the future. To this aim, in Japan the LCGT (large cryogenic gravitational telescope) is exploiting the Kamioka mine, a place that will arise nice warm feelings in neutrino enthusiasts. To get x10-15 improvements, the plan is to take GEO technology and apply to LIGO.silica suspensions, more sophisticated interferometry, more powerful lasers.

In all cases, the sources of noise are hard to avoid, and they will limit the sensitivity in the 10^-25 amplitude range for gravitational wave frequencies around 100 Hz. Nonetheless, with x10-x15 improvements, the range of detection of binary black holes could go from 100 Mpc up to z=2, the sensitivity to ellipticity of known pulsars could go from 3E-6 down to 2E-8, and binary neutron stars could be detected all the way to 350 Mpc, at rates going from 1 per 100 years to 40 per year. Please don’t ask me how a x15 improvement can buy you that much, I did not get it.

Jim Hough concluded by noting that gravitational wave detectors systems are now reaching levels where they may see signals associated with gamma ray bursts within the next few years. The essentially guaranteed detection of compact binary systems by the advanced detectors early in the next decade is likely lead to further understanding of the nature of the gamma ray bursts. Once gravitational waves are established, we will enter an era where GW astronomy will be just another way of observing the universe.

… All good, but I will still fancy my dobsonian scope better.



1. Robert Spero - July 3, 2007

I must say from the outset that what impressed me most in his talk was his first slide, which showed how LISA will have such excellent sensitivity to all the things he took as benchmarks in his later discussion, that one question remained in the back of my mind: if LISA works, will LIGO, VIRGO, and their friends be justified ?

The justification is that ground-based detectors are more sensitivity at higher frequency (> 1 Hz), since they can use much more powerful lasers. High-frequency sources — such as solar-mass neutron star inspirals — are different from LISA’s low-frequency sources, such as massive black hole inspirals.


at rates going from 1 per 100 years to 40 per year. Please don’t ask me how a x15 improvement can buy you that much, I did not get it.

Assuming sources are homogeneously distributed (a not bad assumption in a volume including many galaxies), the event rate is proportional to the cube of the “range.” Range is proportional to strain sensitivity.

2. dorigo - July 3, 2007

Hi Robert,

yes, I understand ground-based and airborne detectors are complementary – I think what I am trying to ask is what sources can hide in one part of the spectrum and give no signal in the other ?

As for the x15: duh!


3. Scott H. - July 3, 2007

I think what I am trying to ask is what sources can hide in one part of the spectrum and give no signal in the other ?

Quite a few. A source’s wave spectrum is more or less inversely proportional to its mass. LISA sources tend to be things involving the kind of massive black holes that we find at the cores of galaxies (as well as a large number of less massive objects that are widely separated – Kepler’s law in action). LIGO/Virgo/GEO/TAMA/etc sources tend to be things involving stellar mass objects — neutron stars and smaller black holes, for example.

A good analogy is to the bands of the electromagnetic spectrum. If you were to ask “What source could be visible in gamma-rays but give no signal in optical?” we could come up with a pretty long list!

4. Robert Spero - July 3, 2007

Hi, Tomasso.

As in radio vs. optical electromagnetic astronomy, 5 orders of magnitude difference in frequency reveals dramatically different astronomy. Think supernovae — stellar mass energy at modest red shift — vs. galactic core supermassive black holes at cosmological distance. That said, orbital detectors should have higher event rate and signal-to-noise than ground-based detectors, which is the main reason that I’ve switched from LIGO to LISA.


5. Ponderous - July 4, 2007

There are two LIGO upgrades planned. What’s running right now is “Initial LIGO”. They might, in a year, see a clear signal, but it was the most advanced detector that folks felt sure they could build. And after a few years of tweaking, it’s now performing almost at design sensitivity.

The big planned upgrade is “Advanced LIGO”, which will be 10x more sensitive. That means that it can detect the same events 10 times further away, which is a 1000x larger volume of space. So you’ll see in 1/3 of day (8 hours) what you’d have to wait 333 days (1 year) for in initial LIGO. They’re SURE to see lots of interesting stuff.

Advanced LIGO will replace initial LIGO in the same facility, but there’s a gap in the schedule. People are proposing to do a semi-upgrade, using some of the cheaper Advanced LIGO technologies, to create an intermediate “Enhanced LIGO” about 2x as sensitive as initial LIGO, before it has to be torn down to make room for AdvLIGO.

Now this whole system is detecting motions that are not only smaller than an atom, they’re smaller than a single proton! This is done using laser light bouncing off mirrors. The mirrors must be suspended with the most exquisite care, so that nothing disturbs them but gravity. They’re basically pendulums hung in a vacuum from very very thin wires.

Initial LIGO’s mirrors are hung using thin steel wires. But for the upgrade, they want to use something lighter and stronger, namely silica (quartz) fibres.

The reason that you can’t detect low frequencies on earth is that the earth moves. Distant earthquakes, as well as waves breaking on shores worldwide, cause low-frequency sound waves to travel through the earth. Those waves are made up of moving atoms. Which is moving mass, which causes gravity. This miniscule “gravity gradient” signal, because it is so close, drowns out the one coming from the stars.

The only way to “hear” well at frequencies much below 10 Hz is to get away from the earth.

6. dorigo - July 4, 2007

Hi Robert,

yes, it is no surprise to me that we want to explore the whole spectrum. My point was that the motivation for ground based experiments was sort of deceiving the way I heard it, since it was based on proving the existence rather than on measuring effects that LISA can see with more ease in another frequency range.

Ponderous, thank you for your contribution, which is entirely in the style of this blog – trying to make things easy to understand (and even succeeding in doing so sometimes). I will promote your comment to a post, if you send me some links and pictorial material to add to it. A non-anonymous id would be good too.
Let me know, my email is dorigo(at)pd(dot)infn(dot)it.

Cheers all,

7. Robert Spero - July 4, 2007

Tommaso —

Right, if the purpose of building these detectors were to confirm the wave prediction of general relativity, source frequency wouldn’t matter. From the start, however, the driving force behind the effort has been astronomy, not physics.


8. dorigo - July 4, 2007

Hi Scott,

thanks for your comment, which I have only now recovered from the spam filter. (I think it was tagged because of the list of unknown words “ligo/virgo/tama/…”.

It is nice to be found wrong in so many ways… One learns a lot very economically!


9. Serious News from Outer Space [Uncertain Principles] · Articles - July 9, 2007

[…] Tommaso Dorigo has a summary of the current sate of gravitational wave detection experiments. These are really cool stuff, and I may have more to say about the topic […]

10. carlbrannen - July 16, 2007

The latest word from GRG18 is that the gravitational wave people don’t see any signal coincident to GRBs or neutrinos. One of the GRBs is supposed to come from M31 which would mean that it should have a detectable gravity wave. Detecting the wave by coincidence of course assumes that the speed of gravity is c. I’ve written a quick blog post on the history of the theory of the speed of gravity. There was quite a ruckus on arXiv about using VLBI to measure the Shapiro delay of a pulsar near Jupiter recently.

And, more importantly, shows that WordPress allows latex in both blogs and comments. So we can \int_0^1x^2\;dx = \frac{1}{3} to our heart’s content.


P.S. When I read the above comments, I see “Tommaso” also spelt “Tomasso”. This problem would go away if you signed your posts with something longer than “T”.

Sorry comments are closed for this entry

%d bloggers like this: