I'm trying to derive Ricci scalar with FRW metric, but additional $c^2$ makes me confused.

The book by D. Baumann says \begin{align} R &= g^{\mu\nu}R_{\mu\nu} \\ &= -R_{00}+\frac{1}{a^2}R_{ii}=\frac{6}{c^2}\left[\frac{\ddot{a}}{a}+\left(\frac{\dot{a}}{a}\right)^2+\frac{kc^2}{a^2R_0^2}\right],\tag{2.133} \end{align} where $$R_{00}=-\frac{3}{c^2}\,\frac{\ddot{a}}{a}$$

And for the calculation of the Ricci scalar, since $g^{\mu \nu}$ is inverse of $g_{\mu\nu}$ so I thought if $g_{00} = -c^2$ then inverse of it should be $g^{00}=-1/c^2$ because the FRW metric says $ds^2 = -c^2dt^2 + a^2(t)\gamma^{ij}dx^idx^j$.

My assumption on Ricci scalar was $R = g^{00}R_{00} + g^{ij}R_{ij} = -1/c^2 R_{00} + 1/a^2 R_{ii}$(in cartesian coordinate with $x=0$.)

But I don't understand why the additional $1/c^2$ factor is missing on the text book.

Or is it conventional not to right speed of light $c$ in $g_{00}$ component of the metric? So that even it is in SI unit, not natural unit, I should right down $g^{00} = -1$, not $-c^2$?

  • $\begingroup$ Images of mathematical expressions are very strongly discouraged here. Please use Mathjax instead. $\endgroup$ Dec 20, 2022 at 13:43
  • $\begingroup$ $g^{\mu\nu}$ being the inverse of $g_{\mu\nu}$ does not mean that $g_{00}= 1/g^{00}$. It is a matrix inverse. $\endgroup$
    – mike stone
    Dec 20, 2022 at 14:13
  • 1
    $\begingroup$ Since $ds^2=g_{\mu\nu}dx^\mu dx^\nu$ it's usually convenient to choose coordinates for which $[x^\mu]=\mathsf{L}$ for all $\mu$, so $[ds^2]=\mathsf{L}^2$ while $g_{\mu\nu}$ is dimensionless. For example, in this convention we take $x^0=ct$ rather than $x^0=t$. (However, it gets awkward if the space coordinates might require e.g. $x^2=R\theta,\,x^3=R\phi$ for a suitable length $R$, rather than $x^2=\theta,\,x^3=\phi$; one might even take $R=r=x^1$.) Then $[\Gamma^\mu_{\nu\rho}]=\mathsf{L}^{-1}$, while Riemann/Ricci terms are $\mathsf{L}^{-2}$. $\endgroup$
    – J.G.
    Dec 20, 2022 at 14:24
  • 1
    $\begingroup$ I'm sorry for uploading image instead of equations. I'll make sure not to upload image next time Thank you for letting me know. And now i understand what was wrong with my assumption. Thank you. $\endgroup$
    – hwan
    Dec 20, 2022 at 15:22
  • 3
    $\begingroup$ @mikestone Not in general, but if the metric is diagonal (like the FLRW is), then it is correct. $\endgroup$
    – Javier
    Dec 20, 2022 at 15:39

1 Answer 1


I don't have the required 50 reputation to write this as a comment so this will be an answer instead.

It is a common convention to write $c=1$ (and often $\hbar=1$) since we can use dimensional analysis at the end to redimensionalise our equations (there will be a unique way of writing the powers of $c$ and $\hbar$). The lecture notes by Baumann use exactly this convention (so I assume the book does too). As such, the metric would read $ds^2=-dt^2+a^2(t)\gamma_{ij}(x^\mu)dx^idx^j$.

It is worth noting that, under this convention, time now has the unit of length.

A note on metric inversion: When calculating the inverse, as others have noted, this will be a matrix inverse, however often the metric is already diagonal and of the form (in FRW)) $ds^2 = -dt^2 + a^2(t)f^i(x^\mu)dx_i^2$. Then the components of the inverse metric are indeed just the reciprocals of the coefficients of $dt^2$, $dx_i^2$ etc.

Punchline: Yes, this should just be a matter of convention with the author writing $c=1$.


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.