In general $d$ dimensional setup ($d-1$ spatial coordinates and $1$ temporal), with a nontrivial metric $g_{\mu\nu}(t,\vec{r})$, having for simplicity a static electric potential $\phi(\vec{r})$. How to calculate the electric charge that is the source of this potential?
I know that when considering classical physics, the charge $Q$ is given by the spatial integration on the laplacian of the potential over all the space $V$:
$$Q=-\int_V \nabla^2\phi(\vec{r})d\vec{r}$$
Three options make sense to me when moving to curved spacetime:
- There is no change at all and we can calculate it as the classical integration above
- All the components of the metric should be taken into account according to $$Q = -\int_V d^{d-1}r \sqrt{-g}g^{\mu\nu}\nabla_\mu\partial_{\nu}\phi(\vec{r})$$ Here $\mu,\nu\in\{\vec{r},t\}$ Where $\nabla_\mu$ is the covariant derivative which operate on a vector field $\partial_{\nu}\phi(\vec{r})$ as follow: $$\nabla_\mu\partial_{\nu}\phi(\vec{r})=\partial_\mu\partial_{\nu}\phi(\vec{r})-\Gamma^{\alpha}_{\mu\nu}\partial_{\alpha}\phi(\vec{r}) $$ and $g=\det{g_{\mu\nu}}$
- Only the spatial part of the metric is joining the spatial integration: $$Q = - \int_V d^{d-1}r \sqrt{-\tilde{g}}\tilde{g}^{\mu\nu}\nabla_\mu\partial_{\nu}\phi(\vec{r})$$ Here $\mu,\nu\in\{\vec{r}\}$ Where $\tilde{g}$ represents the spatial part of the metric, such that, the full metric is given by: $$g_{\mu\nu}=\begin{pmatrix} g_{tt} & g_{t\vec{r}} \\ g_{\vec{r}t} & \tilde{g}_{\vec{r}\vec{r}} \\ \end{pmatrix}$$
From (1)-(3) what is the right way to calculate the total charge?