4
$\begingroup$

I have seen quite a bit online about how there are only 20 independent components for the (lowered) Riemann tensor $R_{abcd}$ for the Schwarzschild metric. I've been told this follows from the symmetries of the tensor, i.e.:

$R_{abcd}=-R_{bacd}=-R_{abdc}=R_{badc}$ and $R_{abcd}=R_{cdab}$

Now if the indices in the tensor can all run from 1 to 4, then $R_{abcd}$ has 256 components. These symmetries seem to reduce our need to calculate components, but why only 20?

Note: I am aware there are some similar questions on the stack exchange. I have read these, but none of them very clearly explained this specific point, so I decided to ask it directly so hopefully I can get my head around it.

$\endgroup$
1
$\begingroup$

Just to summarise the calculation FrodCube mentioned, $R_{abcd}\ne 0$ only when $a\ne b$ and $c\ne d$. In $n$-dimensional spacetime there are $\binom{\binom{n}{2}+1}{2}$ ways to choose the pairs $ab,\,cd$ (they're allowed to be the same, hence the $+1$). We then lose $\binom{n}{4}$ degrees of freedom to the first Bianchi identity $R_{abcd}+R_{acdb}+R_{adbc}=0$. A bit of algebra says that leaves $\frac{1}{6}\binom{n^2}{2}$ DOFs, which for $n=4$ is $20$ as required.

$\endgroup$
1
$\begingroup$

The comment by FrodCube points to the December 1997 Lecture Notes on General Relativity by Sean M. Carroll equation (3.85) which when evaluated for n=4 gives 20.

For those who don't want to open the link/it does not exist/want to see it quickly, my answer is for them.

First note that $^nC_{k}$ is given by (https://en.wikipedia.org/wiki/Binomial_coefficient),

\begin{equation} ^nC_{k} = \frac{n!}{k! (n-k)!} \end{equation}

The number of independent components of Riemann tensor ($IRC$) in $d$ space-time dimensions is simply,

\begin{equation} IRC = \frac{~^dC_{2} (^dC_{2} + 1)}{2} - ~^dC_{4} \end{equation}

$\endgroup$
  • $\begingroup$ You probably ought to explain how the expression for $IRC$ comes about from $^nC_k$ to make this a better answer. $\endgroup$ – StephenG Sep 13 '18 at 17:49

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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