Gravitational waves travel at the speed of light. A photon travelling in exactly the same direction as a gravitational wave will therefore remain in exactly the same position relative to the wave - at the peak, the trough, or somewhere in between. So my question is: if we're observing light emitted from the same direction as a source of gravitational waves (or even light emitted from the GW source), would we be able to detect any difference compared to light observed later, after the GW strength had dropped below detectable level?

The answers to another question (Effect of Gravitational Waves on light?) mention "riding" a GW and that "for the impacted time, [the GW] should impact speed of light, time, wavelength etc", but none of the answers provide specific detail on what the actual effect would be.

If the stretching/squeezing of spacetime was in the same direction as the GW, I'd expect the light to be alternatively redshifted and blueshifted in pulses of the same frequency as the GW. [Whether we could develop the technology to detect this is another matter]. However, as Paul T notes in his answer, the stretching/squeezing occurs in the transverse directions. I'm at a loss to picture what this does to the light when we observe it.

So I'm interested to know what the effect of light "riding" gravitational waves would be, and whether any effect would actually be observable.

I'm assuming the effect would be ridiculously tiny; but on the other hand, if LIGO can detect an oscillation in spacetime the size of a proton...


1 Answer 1


Good question.

If you ignore the effect of the light on the gravitational wave (it has energy momentum so it'll do something, maybe very minor), and we treat the problem in the linear domain first, the GW wave would distort the space transverse to the propagation, so the speed of light would not change but the polarization would be changing some as the electric and magnetic field would also change. Not obvious to me whether they stretch/squeeze or something else, one would have to take the $F^{\mu\nu}$ tensor and solve the source free GR Maxwell's equation in the gravitational linear domain, with the metric terms in the linearized space.

Other than more nonlinearities it seems to be this would then be calculable. I've never seen a solution or it posed.

  • $\begingroup$ Thanks Bob for your answer, which certainly clarifies the nature of the problem. I was hoping someone might actually do the calculations and describe what comes out. I'll give it a few more days, but if no one improves on this, you win the tick. $\endgroup$ Commented Nov 2, 2017 at 7:58
  • $\begingroup$ I have not tried, but just maybe, in the linear domain, the first approximation might be just the EM wave, and you probably have to figure out the right boundary conditions to make it appropriate to having initially the EM wave. Maybe at t =0 it's a simple sinusoid (I'd do it all in complex exponential), but am not sure t =0 conditions would work. The fun will be in setting it up, not necessarily solving it. Keep just the lowest interaction terms. Without actually doing it, not sure that one gains much by seeing it. $\endgroup$
    – Bob Bee
    Commented Nov 3, 2017 at 3:14

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.