1. Limited time only! Sign up for a free 30min personal tutor trial with Chegg Tutors
    Dismiss Notice
Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Homework Help: Determinant of metric tensor

  1. Apr 12, 2009 #1
    1. The problem statement, all variables and given/known data
    Prove that

    where g is the determinant of the metric, and [itex]\Gamma[/itex] are the Christoffel connection coefficients.

    3. The attempt at a solution
    From the general definition of the coefficients I got:
    [tex]\Gamma^\mu_{\mu\lambda}=(1/2)g^{\mu\rho}\partial_\lambda g_{\rho\mu}[/tex]

    But I have no idea how to work with the determinant of the metric. I'm not sure if I'm allowed to use this:

    det(g)=exp[Tr ln G]

    And if I did, would I have to use the GR definition of the trace?

    [tex] Tr R = R^\mu_\mu [/tex]

    I cleaned it up a little bit with the chain rule:

  2. jcsd
  3. Apr 12, 2009 #2


    User Avatar
    Science Advisor
    Homework Helper

    The last part you wrote can't be right.

    I'd start with

    [tex] \delta |g| = \delta \[\exp \(\mbox{Tr} \ln |g|\)\] [/tex]

    and then see what the 'delta' of the RHS brings me.
  4. Apr 12, 2009 #3


    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    Isn't there an index-notation expression for the determinant?
  5. Apr 12, 2009 #4
    Thanks for the tips. Turns out it's a bit simpler:

    1. From Jacobi's formula we get:
    [tex] dg=A^{ik}dg_{ik}[/tex]
    where [itex]A^{ik}[/itex] are elements of the adjugate of the metric.

    2. Standard differential expansion:
    [tex]dg=\frac{\partial g}{\partial g_{ik}}dg_{ik}[/tex]
    [tex]\Rightarrow A^{ik}=\frac{\partial g}{\partial g_{ik}}[/tex]

    3. From Laplace's formula for the determinant we get:

    Putting this all together we have:
    [tex]g^{ik}=\frac{1}{g}\frac{\partial g}{\partial g_{ik}}[/tex]

    Substitute this into my expression for [itex]\Gamma^\mu_{\mu\lambda}[/itex] above and the result follows.

    I'm still working on the proofs to steps 1 and 3 above, but essentially the problem is solved. Feels good to brush up on my linear algebra, haven't used it for years.
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook