≡ Menu

Effects of Einstein’s Elusive Gravitational Waves Observed

Two white dwarfs similar to those in the system SDSS J065133.338+284423.37 spiral together in this illustration from NASA. Credit: D. Berry/NASA GSFC

Locked in a spiraling orbital embrace, the super-dense remains of two dead stars are giving astronomers the evidence needed to confirm one of Einstein’s predictions about the Universe.

A binary system located about 3,000 light-years away, SDSS J065133.338+284423.37 (J0651 for short) contains two white dwarfs orbiting each other rapidly — once every 12.75 minutes. The system was discovered in April 2011, and since then astronomers have had their eyes — and four separate telescopes in locations around the world — on it to see if gravitational effects first predicted by Einstein could be seen.

According to Einstein, space-time is a structure in itself, in which all cosmic objects — planets, stars, galaxies — reside. Every object with mass puts a “dent” in this structure in all dimensions; the more massive an object, the “deeper” the dent. Light energy travels in a straight line, but when it encounters these dents it can dip in and veer off-course, an effect we see from Earth as gravitational lensing.

Einstein also predicted that exceptionally massive, rapidly rotating objects — such as a white dwarf binary pair — would create outwardly-expanding ripples in space-time that would ultimately “steal” kinetic energy from the objects themselves. These gravitational waves would be very subtle, yet in theory, observable.

Read: Astronomy Without a Telescope: Gravitational Waves

What researchers led by a team at The University of Texas at Austin have found is optical evidence of gravitational waves slowing down the stars in J0651. Originally observed in 2011 eclipsing each other (as seen from Earth) once every six minutes, the stars now eclipse six seconds sooner. This equates to a predicted orbital period reduction of about 0.25 milliseconds each year.*

“These compact stars are orbiting each other so closely that we have been able to observe the usually negligible influence of gravitational waves using a relatively simple camera on a 75-year-old telescope in just 13 months,” said study lead author J.J. Hermes, a graduate student at The University of Texas at Austin.

Based on these measurements, by April 2013 the stars will be eclipsing each other 20 seconds sooner than first observed. Eventually they will merge together entirely.

Although this isn’t “direct” observation of gravitational waves, it is evidence inferred by their predicted effects… akin to watching a floating lantern in a dark pond at night moving up and down and deducing that there are waves present.

“It’s exciting to confirm predictions Einstein made nearly a century ago by watching two stars bobbing in the wake caused by their sheer mass,” said Hermes.

As of early last year NASA and ESA had a proposed mission called LISA (Laser Interferometer Space Antenna) that would have put a series of 3 detectors into space 5 million km apart, connected by lasers. This arrangement of precision-positioned spacecraft could have detected any passing gravitational waves in the local space-time neighborhood, making direct observation possible. Sadly this mission was canceled due to FY2012 budget cuts for NASA, but ESA is moving ahead with developments for its own gravitational wave mission, called eLISA/NGO — the first “pathfinder” portion of which is slated to launch in 2014.

The study was submitted to Astrophysical Journal Letters on August 24. Read more on the McDonald Observatory news release here.

Inset image: simulation of binary black holes causing gravitational waves – C. Reisswig, L. Rezzolla (AEI); Scientific visualization – M. Koppitz (AEI & Zuse Institute Berlin)

*The difference in the eclipse time is noted as six seconds even though the orbital period decay of the two stars is only .25 milliseconds/year because of a pile-up effect of all the eclipses observed since April 2011. The measurements made by the research team takes into consideration the phase change in the J0651 system, which experiences a piling effect — similar to an out-of-sync watch — that increases relative to time^2 and is therefore a larger and easier number to detect and work with. Once that was measured, the actual orbital period decay could be figured out.

Comments on this entry are closed.

  • John Stock August 29, 2012, 3:27 PM

    NASA cancelled ExoMars too.. Leaving ESA to do it alone.

    • Aqua4U August 29, 2012, 10:06 PM

      The shortsightedness of our political minions approaches infinity.. given what’s being spent for other questionable exertions and what might be gained by continuing our quest for knowledge…

  • CJSF August 29, 2012, 3:39 PM

    I am confused by the rate of reduction. In the article, it is stated that there is a 0.25 millisecond reduction per year, but that after a year, there was a 6 second reduction. Did the author mean 0.25 milliseconds per orbit of the star(s)?

    • Jason Major August 29, 2012, 4:09 PM

      Yes. Orbital period is shrinking (decaying) at a rate of 0.25 milliseconds per year.

      • spell7 August 29, 2012, 5:48 PM

        Is the decay rate of 0.25 milliseconds per year based on an Earth year or the binary system year of 12.75 minutes?

        • Deejay Nunti August 30, 2012, 3:46 AM

          that decay was measured in 1 Earth year…

  • Jon Souter August 29, 2012, 3:50 PM


    Originally observed in 2011 eclipsing each other (as seen from Earth) once every six minutes, the stars now eclipse six seconds sooner — an orbital period reduction of about 0.25 milliseconds each year.”

    Does that maths seem wonky to anybody else? Or am I missing something?

    • Steven Reid August 29, 2012, 6:37 PM

      Two stars eclipse eachother, one star a will eclipse star b, then star b will eclipse star a. One eclipse happens every 6 minutes, ergo 2*6 = 12

    • Jason Major August 29, 2012, 7:05 PM

      Check out the asterisk addendum above. – J

  • MVK August 29, 2012, 4:30 PM

    So how long until the stars come together and what would happen then? I calculated it as 3.06 million years until they collide but thats if the decay isn’t accelerating

  • Aqua4U August 29, 2012, 5:12 PM

    I really like that second image! USUALLY when one sees an image of something orbiting a gravity well, it is depicted only in two dimensions. The massive body is shown ‘denting’ or making an impression in a 2D wireframe representation, while the obiting object makes a smaller ‘dent’ in the same. I’ve always thought that imagery misleading.. Yes the basic concept is there for visualization purposes, but the 4 dimensional aspect of a warped space time is not complete in that reference.

    Then again.. were someone to more accurately portray a like event in 4 dimensions, it would still be lacking due to the possibility that even more dimensions are effected? Which, of course, would be even harder to depict from our four dimensional point of view. (Assuming time as the 4th dimension)

  • Jay Cross August 29, 2012, 8:15 PM

    Not the first such measurement… there is a pair of co-orbiting pulsars that were also measured as radiating orbital energy somehow (presumed to be gravitational waves).

    • Jason Major August 29, 2012, 9:59 PM

      This is the first time, though, that measurements were made with optical data.

  • lcrowell August 29, 2012, 9:28 PM

    This is a variant of the Taylor-Hulst measurement of the frequency change in neutron star orbits. It is good to get a repeat of this measurement. The search fro gravitational waves has been slow and frustrating. The problem is that the wave equation for weak gravity waves is

    ?h_{ab} = (16?G/c^4)T_{ab},

    and the coupling term 16?G/c^4 is extremely small. The h_{ab} is the wave perturbation on a flat spacetime metric and ? is the d’Alembertian operator ? = ?^2 – ?_t^2. T_{ab} is the response of material to a gravity wave, or equivalently the matter source of a gravity wave. The result is the LIGO detector has a tiny response to a gravity wave.

    LC

  • Rick Holcomb August 30, 2012, 3:06 PM

    I wonder what frame dragging looks like around such a system.

  • softeky August 31, 2012, 1:34 AM

    I too find the article very confusing. The footnote does not help me understand anything and is quite jumbled.

    From TFA (http://xxx.lanl.gov/pdf/1208.5051v1) I do not see a reference to the 6 second or 20 second period change numbers reported here (but there are a lot of numbers present and I’ll have to read more thoroughly). A 20 second change in two years off a 12.75 minute orbital period means the phenomenon is short lived and the report of a {6, 14, …} second decay-period-change-per-earth-year series is suspect for that reason alone (I don’t see any reference to it on a second read of TFA).

    I don’t see any discussion of the possibility that these white dwarfs are losing mass over time which might also contribute to their rotation rate change. Does orbital mechanics and conservation rule that (mass loss) out for some reason?

    Universetoday, where did you get the 6 and 20 second numbers from?

hide