We know that gravitational waves are emitted (at least in GR) when the system has a time-varying quadrupole (or higher) moment. My question is

Is it possible to easily tell (e.g. just by looking) if a system has such moments?

Is there some kind of symmetry to look for that lets you know that the system will emit such radiation?


2 Answers 2


I don't think it possible to easily tell, but here is how you might approach this. The multipole moments $T_{\ell, m}(t)$ of the energy density $T^{00}(t, \mathbf x)$ of a source can be written on surfaces of constant time in terms of spherical harmonics as follows: $$ T_{\ell m}(t) = \int d^3x \, Y^*_{\ell, m}(\theta, \phi)r^\ell\,T^{00}(t, \mathbf x) $$ Notice that if we write the energy density in terms of spherical harmonics; $$ T^{00}(t,\mathbf x) = \sum_{\ell, m}c_{\ell, m}(t,r)Y_{\ell, m}(\theta, \phi) $$ then performing the integration in spherical coordinates and using orthogonality of spherical harmonics (I can give details if you want), we find $$ T_{\ell, m}(t) = \int dr\, r^{\ell+2} c_{\ell, m}(t,r) $$ Therefore, we see that if certain of the coefficients $c_{\ell, m}$ in the expansion of the density in terms of spherical harmonics vanish, then the corresponding multipole moment vanishes. The monopole moment corresponds to $\ell = 0$, the dipole moment to $\ell = 1$, and the quadrupole moment to $\ell = 2$.

So say, for example, that the energy density has spherical symmetry, namely $$ T^{00}(t,\mathbf x) = f(t,r), \qquad r = |\mathbf x| $$ Then its expansion contains no spherical harmonics with $\ell>0$, so all multipole moments other than the monopole moment vanish. In particular, there is no quadrupole.

So the best advice I can give is to look at the pictures of spherical harmonics, and attempt to get some intuition for when certain energy distributions will contain them in their expansions.

I hope that helps! Let me know of any mistakes and/or typos.


  • $\begingroup$ I like the idea of using the pictures* of spherical harmonics but I'm a little confused on one point. Take a binary system for example, to me that looks like a dipole, am I thinking of this in the wrong way? *en.wikipedia.org/wiki/File:Harmoniki.png $\endgroup$
    – user12345
    Commented Feb 7, 2013 at 11:23
  • 1
    $\begingroup$ Yeah I don't have very good intuition for this myself; have you tried calculating the $c_{\ell,m}$ for an example binary system for some small values of $\ell,m$ to see what you get? That's what I would do (and I might if I have time). $\endgroup$ Commented Feb 7, 2013 at 18:33
  • $\begingroup$ I haven't calculated any spherical harmonics, no. Although I think I could, I don't see how it would help me with developing an intuition, if that's even possible. Thanks for interest and suggestions though! $\endgroup$
    – user12345
    Commented Feb 7, 2013 at 19:05
  • 2
    $\begingroup$ My pleasure. My thought process is that computing them for a few simple systems might begin to reveal qualitative features of distributions with each kind of moment. Then, one could make some sort of a hypothesis in general about what which multipole moments do and don't vanish and under which conditions this is the case. If I end up doing this; I'll let you know what I learn. $\endgroup$ Commented Feb 7, 2013 at 19:22

The simplest way to construct a quadrupole is to have two identical dipole displaced by some distance opposed to each other, there is a quadrupole. In some sense, a dark mode dipole (with dipole canceling each other in far field, but not canceling in near field), gives you a quadrupole. An octupole is constructed by placing quadrupoles in a similar fashion.

Or actually, as long as the source is smeared in a finite volume and oscillating with time, you will have all multipole modes up to arbitrary order (just increasingly vanishing). This is the result of multipole expansion.

  • $\begingroup$ Taking the case of a binary system of stars, it almost sounds like what you describe as "to have two identical dipole displaced by some distance opposed to each other, there is a quadrupole" except I can't see why a star would be a dipole, or am I thinking of this in the wrong way? $\endgroup$
    – user12345
    Commented Feb 7, 2013 at 19:08
  • $\begingroup$ That part I also don't know since gravity has only one kind of pole (not like charge or magnet). But is mass is not concentrated into a singular point, then Multipole Expansion is useful to get dipole, quadrupole etc. even though there is one type of "mass". The dipole component should look like $$ mr/R^2 $$ while quadrupole looks like $$ mr^2/R^3 $$ etc. $\endgroup$
    – Bo Zeng
    Commented Feb 7, 2013 at 21:29

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.