I have a quick question, in the derivation of the Raychaudhuri equation in Wald: General Relativity on page 218, equation 9.2.10, an identity for $\xi^a \nabla_a B_{ab}$ is derived, where $B_{ab} = \nabla_b \xi_a$ and $\xi^a$ is a timeline geodesic. From the second to the third line, the expression $\nabla_b(\xi^c \nabla_c \xi_a)$ is assumed to vanish. Why is that? Is that just the geodesic equation? Does this hold even though $\xi_a$ is a covariant vector?

Any help will be appreciated.


1 Answer 1


That term vanishes for the reason that you suspect. It is simply the geodesic equation. A geodesic $\xi^a$ is a curve defined in such a way that $\nabla_\xi\xi^c=\xi^a\nabla_a\xi^c=0.$ If your concern is that the object inside the covariant derivative is $\xi_a$ rather than $\xi^a$ then you only need to use one more fact which is that the metric is preserved by the covariant derivative, i.e. $\nabla_a g_{bc} = 0$. Therefore: $\xi^a\nabla_a\xi_c = \xi^a\nabla_a g_{bc}\xi^b = g_{bc}\xi^a\nabla_a\xi^b = 0.$

  • $\begingroup$ Thank you! One additional question, the expansion is defined in Wald as $\Theta = B^{ab} h_{ab}$, where $h_{ab} = g_{ab}+\xi_a \xi_b$. Is this equivalent to the definition $\Theta = \nabla_a \xi^a$? $\endgroup$
    – eb1612
    May 8, 2018 at 18:34
  • $\begingroup$ I don't have Wald with me, and I don't recall these variables. It might be best to open another question rather than asking in comments. $\endgroup$
    – enumaris
    May 8, 2018 at 18:36

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.