I'm working through relativistic electrodynamics, but I'm stuck on determining the $00$-th component of the energy-momentun tensor. The (contravariant) 2-tensor (in Heaviside-Lorentz units) reads:


It's the first term that's causing me pain (the second is just a standard field invariant), so if we just look at that:


I want to show this to be the square of the electric field, but I'm confused about the indices. I would have thought that

$$F_{\alpha0}=\eta_{0 \sigma}F_{\alpha}^{\:\sigma}=\eta_{00}F_{\alpha}^{\:0}=F_{\alpha}^{\:0}$$

with no sign change as I'm lowering a temporal index. But when I swap the indices for $F^{0\alpha}$, I pick up a minus sign due to the antisymmetry of $F$. Since I want the whole thing to just be the EM energy density, I need to get rid of this minus sign. I presume that I must compensate for it in the index lowering process that I'm messing up somehow.

I presume I'm just making some novice mistake, so if someone could point it out and set me straight that would be much appreciated!


1 Answer 1


The term is$$F^{0\alpha}F_\alpha{}^0=F^{0i}F_i{}^0=F^{0i}\eta_{i\mu}F^{\mu0}=F^{0i}\eta_{ij}F^{j0}\stackrel{\ast}{=}\pm F^{0i}\delta_{ij}F^{j0}\stackrel{\ast}{=}\pm F^{0i}F^{i0}=\mp E^2$$if you work in $\mp\pm\pm\pm$, where each $\ast$ breaks the usual contraction rules for indices to emphasize the value.

  • $\begingroup$ Oh brilliant, that cleared it up! I prefer -+++ signature as I'm just learning this and differential geometry (with GR in view) for the first time. $\endgroup$ Dec 8, 2021 at 13:56

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.