Static Limits and Ergoregions

An ergoregion is the domain where the Killing field that represents time translation at infinity becomes spacelike. No future timelike observer can then follow an orbit of that field, so every observer must rotate relative to infinity. The outer boundary is the stationary-limit surface. For nonzero-spin Kerr it lies outside the event horizon away from the rotation axis; at $a=0$ it coincides with the horizon at every angle and the ergoregion is absent. It is not itself a horizon on the rotating branch, because causal worldlines can cross it in both directions.

Required background. Spacetime Symmetries and Killing Fields supplies conserved Killing energy and axial symmetry. Helpful background. Exact Solutions of Einstein’s Equations supplies the Kerr metric; Causality and Global Structure separates stationary limits from event horizons; Einstein Equations fixes the rotating vacuum family; Black-Hole Geometry and Mechanics supplies horizon generators; Black-Hole Perturbation Theory develops wave amplification beyond the geometric threshold used later.

Timelike stationarity and its limit

Let $t^a$ be the stationary Killing field normalized to unit time translation at the asymptotically flat end. A stationary observer follows an orbit of $t^a$, with four-velocity proportional to $t^a$. Such an orbit can be timelike only where

$$t^2=g_{ab}t^at^b\lt0.$$

The stationary-limit surface is $t^2=0$. The ergoregion is the black-hole exterior subset in which $t^2\gt0$, or one of that subset’s connected components when the distinction matters. There $t^a$ is spacelike, so no normalization makes its integral curves physical observer worldlines.

This statement is tied to a chosen asymptotic time translation. In a rotating asymptotic frame, the corresponding Killing combination changes. For asymptotically flat Kerr, unit time at infinity gives the canonical choice. The definition is coordinate invariant once that Killing field is fixed: $t^2$ is a scalar.

The event horizon is different. Its stationary generator is

$$\chi^a=t^a+\Omega_H\phi^a,$$

where $\phi^a$ is the axial Killing field. The horizon is where this particular combination is null and normal, not generally where $t^a$ is null.

Kerr stationary-limit calculation

Use Boyer–Lindquist coordinates on Kerr with physical mass $M\gt0$ and write $a=J/M$. With

$$\Sigma=r^2+a^2\cos^2\theta, \qquad \Delta=r^2-2GMr+a^2,$$

the stationary Killing norm is

$$g_{tt}=-\left(1-\frac{2GMr}{\Sigma}\right).$$

The outer stationary-limit radius is found from $g_{tt}=0$:

$$r_{\mathrm{s}}(\theta) =GM+\sqrt{G^2M^2-a^2\cos^2\theta}.$$

The outer horizon is the larger root of $\Delta=0$,

$$r_+=GM+\sqrt{G^2M^2-a^2}.$$

Since $\cos^2\theta\leq1$, $r_{\mathrm{s}}(\theta)\geq r_+$. For $a\ne0$, equality holds only at the poles; for $a=0$, equality holds at every angle. At the equator,

$$r_{\mathrm{s}}\left(\frac\pi2\right)=2GM,$$

while $r_+\lt2GM$ for nonzero spin. Thus the ergoregion has finite exterior thickness away from the rotation axis. Visser derives and interprets the Kerr horizons and stationary limits in Visser 2009, §§5–6, pp. 20–27.

Three limits check the calculation. At $a=0$, both radii equal $2GM$ at every angle and the exterior ergoregion disappears. For nonzero spin they still coincide at the poles. In the extremal limit $\lvert a\rvert=GM$, the equatorial static limit remains $2GM$ while the horizon is at $GM$, so the equatorial separation is maximal in this family.

Why all observers must rotate

A worldline at fixed $r$ and $\theta$ can rotate with angular velocity $\Omega=d\phi/dt$. Its tangent is proportional to

$$k^a=t^a+\Omega\phi^a.$$

It is timelike when

$$k^2=g_{tt}+2\Omega g_{t\phi}+\Omega^2g_{\phi\phi}\lt0.$$

Outside the axis, $g_{\phi\phi}\gt0$, so the allowed interval lies between the two roots

$$\Omega_\pm =\frac{-g_{t\phi}\pm \sqrt{g_{t\phi}^2-g_{tt}g_{\phi\phi}}}{g_{\phi\phi}}.$$

Outside the stationary limit, zero may lie in this interval and an observer can remain fixed relative to infinity. Inside the ergoregion, $g_{tt}\gt0$, so $\Omega=0$ is forbidden. In Kerr with positive $a$, both permitted roots have the co-rotating sign sufficiently near the horizon. Frame dragging is then a causal requirement, not merely a force on an otherwise stationary observer.

At the horizon the discriminant collapses and the allowed interval shrinks to

$$\Omega_+=\Omega_-=\Omega_H.$$

The limiting null direction is the horizon generator. This collapse is a useful check that the stationary-limit analysis and Killing-horizon normalization agree.

The static limit is traversable

The surface $t^2=0$ is not generally null as a hypersurface. It marks a change in the causal character of one Killing field, not the boundary of causal escape. An infalling observer crosses the stationary limit while remaining timelike and can, with sufficient outward motion, cross it back before reaching the horizon.

Nor does the ergoregion require divergent curvature or a coordinate failure. Boyer–Lindquist $g_{tt}=0$ is a regular invariant locus for the selected Killing field. The coordinate singularity of that chart occurs at $\Delta=0$, the horizon, and horizon-penetrating coordinates remove it.

The ergoregion permits negative energy relative to $t^a$ because a spacelike Killing vector has no fixed sign when contracted with future causal momentum. That possibility powers Penrose extraction and wave superradiance, but it does not mean a local observer measures negative energy density in an orthonormal frame. The particle and wave mechanisms are developed in Brito, Cardoso, and Pani 2020, Chapter 4, §§4.2–4.3.1 and 4.7.1–4.7.3, pp. 39–44 and 57–60.

Compare pole, equator, and static limit

For a Kerr black hole with $a=\tfrac35GM$, calculate $r_+$, $r_{\mathrm{s}}(0)$, and $r_{\mathrm{s}}(\pi/2)$. Where is the exterior ergoregion thickest?

Solution

The horizon radius is

$$r_+=GM\left(1+\sqrt{1-\frac{9}{25}}\right) =\frac95GM.$$

At the pole, $\cos^2\theta=1$, so $r_{\mathrm{s}}(0)=r_+=9GM/5$. At the equator, $r_{\mathrm{s}}(\pi/2)=2GM$. The coordinate radial separation is therefore zero at the poles and $GM/5$ at the equator, where it is largest. This is a statement about the invariantly defined surfaces expressed in this coordinate radius, not a Euclidean proper thickness.

Negative Killing Energy States explains how future-directed motion can carry negative energy relative to infinity inside this region.

References

• Brito, Richard, Vitor Cardoso, and Paolo Pani. Superradiance. 2nd ed. Lecture Notes in Physics 971. Cham: Springer, 2020. DOI.
• Visser, Matt. “The Kerr Spacetime: A Brief Introduction.” In The Kerr Spacetime: Rotating Black Holes in General Relativity. Cambridge: Cambridge University Press, 2009. arXiv.