Universal Horizons and Preferred-Foliation Black Holes

A universal horizon is a proposed causal boundary for a theory whose preferred foliation permits modes with unbounded group velocity. Its defining equation is simple, but its physical force is conditional: the foliation must remain timelike and regular, propagation must respect its future direction, and the black-hole branch must survive rotation, nonspherical perturbations, and the higher-spatial-derivative regime. Static existence is well supported in restricted models; universal regularity and thermodynamics remain active research questions.

Required background. Frames, Tetrads, and Spinors supplies the æther frame and its hypersurface-orthogonality condition.

Helpful background. Einstein Equations supplies stationary field equations; Gravitational Actions and Boundary Terms supplies variational charges; Formulations of General Relativity supplies foliation and constraint language; and Gravity Beyond General Relativity supplies branch and cutoff discipline.

A causal boundary tied to preferred time

Use signature $(-,+,+,+)$ and a hypersurface-orthogonal unit æther

$$u_a=-N_T\nabla_aT,\qquad u^a u_a=-1,$$

where increasing $T$ defines the allowed future. Let $\chi^a$ be the stationary Killing vector, normalized to unit time translation at spatial infinity. On the static spherical branch, a universal horizon is a limiting preferred-time leaf with compact two-sphere sections satisfying

$$u_a\chi^a=0.$$

Because $u_a$ is normal to $T=\text{constant}$, this equation makes $\chi^a$ tangent to the leaf. In the static spherical solutions, exterior leaves arriving from infinity accumulate at the outermost such surface; their future side points toward decreasing areal radius. A signal may have arbitrarily large spatial velocity relative to $u^a$, yet it cannot cross outward without moving toward decreasing preferred time. This is the causal construction in Barausse, Jacobson, and Sotiriou 2011, end of §V and §VI, PDF.

The statement assumes more than $u\cdot\chi=0$. The scalar $T$ must define a global spacelike foliation outside the boundary; its gradient must not become null; the high-frequency dispersion relation must preserve positive preferred frequency; and the equations must admit evolution without data supplied from the hidden side. If instead the theory has a finite maximum speed $s_{\max}$, its relevant causal boundary is the horizon of the widest finite characteristic cone. If the foliation becomes singular before the root, the root is not a regular causal horizon.

The sources Barausse–Jacobson–Sotiriou and Blas–Sibiryakov use $(+,-,-,-)$, while Berglund–Bhattacharyya–Mattingly use $(-,+,+,+)$. The equation $u\cdot\chi=0$ is signature invariant, but normalizations, effective metrics, surface gravity, and action signs are not; the discussion here translates each full sign package to the site convention.

Static branches and horizon regularity

For static, spherical, asymptotically flat Einstein–æther black holes, finite-speed spin sectors generally possess distinct characteristic horizons. Requiring regularity at the spin-0 horizon removes one shooting datum, leaving a one-parameter family labeled by mass for fixed couplings. Numerical solutions then contain an outermost leaf with $u\cdot\chi=0$ inside the metric and spin-0 horizons. Barausse, Jacobson, and Sotiriou 2011, §§III and V–VI, PDF establishes this result for the infrared two-derivative equations.

That construction is not a solution of the full ultraviolet Hořava action. The higher-spatial-derivative terms are suppressed for a large black hole where the universal horizon lies at low curvature, but precisely the modes invoked to test “arbitrarily fast” propagation probe increasingly short preferred-frame wavelengths. Existence of an infrared background and validity of geometric-optics propagation through it are separate approximations.

A mechanics relation exists for regular static spherical families at fixed æther couplings. To fix its conventions, let $s^a$ be the outward unit spacelike vector orthogonal to $u^a$, set $K_{ab}=h_a{}^ch_b{}^d\nabla_cu_d$ and $K=h^{ab}K_{ab}$, and define the source’s Killing-derivative surface gravity by $\nabla_a\chi_b=-\kappa(u_as_b-s_au_b)$. At the universal horizon, $|\chi|=s\cdot\chi$ on the chosen orientation. This $\kappa$ is a specified geometric quantity; other proposed universal-horizon temperatures need not use it. Defining

$$q_{\mathrm{UH}}=(1-c_{13})\kappa_{\mathrm{UH}} +\frac{c_{123}}2K_{\mathrm{UH}}\lvert\chi\rvert_{\mathrm{UH}},$$

Berglund, Bhattacharyya, and Mattingly derive

$$M_{\mathrm{ae}}=\frac{q_{\mathrm{UH}}A_{\mathrm{UH}}}{4\pi G_{\mathrm{ae}}}, \qquad \delta M_{\mathrm{ae}}=\frac{q_{\mathrm{UH}}\,\delta A_{\mathrm{UH}}} {8\pi G_{\mathrm{ae}}}$$

for one-parameter families of regular, static, spherical solutions Berglund, Bhattacharyya, and Mattingly 2012, §II.A, Eqs. (19)–(24), and §III, Eqs. (39)–(43), PDF. The first law can be checked from scaling: the only length $r_0$ gives $A_{\mathrm{UH}}\propto r_0^2$ and $q_{\mathrm{UH}}\propto r_0^{-1}$, hence $A_{\mathrm{UH}}\delta q_{\mathrm{UH}}=-q_{\mathrm{UH}}\delta A_{\mathrm{UH}}/2$. Differentiating the Smarr relation yields its factor $1/(8\pi G_{\mathrm{ae}})$. This argument fixes the couplings; varying them introduces further terms. The extra extrinsic-curvature contribution means $q_{\mathrm{UH}}$ is not merely the usual Killing-horizon surface gravity. A Smarr relation and first law of mechanics do not by themselves establish entropy, a Hawking state, or a universal temperature.

Rotation separates the theory classes

Static spherical æther is automatically hypersurface orthogonal, which hides a difference between unrestricted Einstein–æther theory and khronometric Hořava gravity. At first order in slow rotation, the infrared Hořava solution retains the spherical khronon profile and is therefore expected to retain its universal horizon Barausse and Sotiriou 2013, §§III–V, PDF. This is an infrared, slow-rotation result, not a construction at arbitrary spin.

Slowly rotating Einstein–æther black holes behave differently. Regular asymptotically flat solutions form a mass-and-angular-momentum family, but the æther develops nonzero vorticity; no hypersurface analogous to the static universal horizon exists in that calculation Barausse, Sotiriou, and Vega 2016, §§III–V, PDF. Finite-speed spin horizons can still exist. This adverse branch is a direct counterexample to promoting the static universal-horizon claim from hypersurface-orthogonal theories to all vector–tensor theories.

Results and limitations through 2026-08-19

The evidence cutoff for this frontier assessment is 2026-08-19.

Formulation and assumptions	Calculation or result	Evidential force	Limitation and exact locator
Two-derivative Einstein–æther; static, spherical, asymptotically flat; regular spin-0 horizon	Numerical one-parameter black-hole branches contain an outermost universal horizon	Strong branch-level existence evidence	No rotation and no full higher-derivative solution; Barausse, Jacobson, and Sotiriou 2011, §§III, V–VI, PDF
Infrared khronometric theory; spherical background followed by nonspherical perturbations	The spherical universal horizon is regular, but instantaneous modes are argued to generate derivative blow-up and a nonlinear finite-area singularity	Serious negative result for regularity under generic collapse	Perturbative argument; ultraviolet operators may change the endpoint; Blas and Sibiryakov 2011, §3.3 and §4.1, pp. 24–27, PDF
Einstein–æther; static, spherical, special coupling families	Exact solutions obey a Smarr formula and a first law involving area and preferred-foliation extrinsic curvature	Establishes classical mechanics on those branches	Radiation spectrum, entropy, and unique thermodynamics remain unresolved; Berglund, Bhattacharyya, and Mattingly 2012, §§III–V, PDF
Infrared Hořava gravity; stationary axisymmetry to first order in spin	Slowly rotating solutions keep the spherical khronon profile and are expected to keep the universal horizon	Permissive evidence beyond spherical symmetry	Slow rotation and infrared order only; Barausse and Sotiriou 2013, §§III–V, PDF
Einstein–æther; stationary axisymmetry to first order in spin	Regular rotating æther has nonzero vorticity and no universal-horizon-like leaf	Explicit theory-class limitation	Does not exclude finite-speed characteristic horizons or a different rapidly rotating structure; Barausse, Sotiriou, and Vega 2016, §§III–V, PDF

Competing interpretations and negative limits

The strongest permissive position is that a preferred foliation supplies an absolute causal order, static infrared solutions realize its compact boundary, slow Hořava rotation preserves it, and classical mechanics provides a coherent horizon law. On this view, ultraviolet dispersion changes the rays but not the one-way character of preferred time.

The strongest conservative position is narrower: universal horizons are established structures of particular stationary infrared branches. Nonspherical instantaneous interactions may turn the surface singular, unrestricted æther rotation removes hypersurface orthogonality, and no unique ultraviolet matter dispersion or quantum state has established universal thermodynamics. These positions disagree about extrapolation, not about the static equation $u\cdot\chi=0$.

Two limiting checks prevent overstatement. In the GR limit with the æther decoupled and all physical fields having finite luminal speed, the metric event horizon is the operative boundary; an auxiliary $u\cdot\chi=0$ leaf has no independent observable status. At the other extreme, if $\nabla_aT$ becomes null or the higher-derivative evolution requires boundary data at the proposed surface, the assumptions behind preferred-time trapping fail before the geometric condition can do its job.

Empirical bounds on couplings, images, or waveforms belong to Tests of Gravity. A stable low-energy quantum-gravity embedding and microscopic horizon state belong to Quantum Gravity. This page assesses the classical theory claim rather than performing either neighboring appraisal.

The conclusions could change after any of the following: a nonlinear nonspherical collapse calculation changes the regularity conclusion; a rapidly rotating solution establishes or excludes a preferred-time boundary; a full higher-spatial-derivative analysis changes the ultraviolet causal structure; or a controlled quantum calculation derives a spectrum and entropy consistent across matter dispersions.

Common pitfalls

Calling every inner characteristic surface universal. A finite-speed spin horizon depends on one effective metric. A universal horizon is tied to the preferred foliation and the unbounded-speed limit.

Deriving thermodynamics from a first-law shape alone. The coefficient contains extrinsic-curvature data, and a radiation state has not been supplied. Classical mechanics is necessary evidence, not a complete statistical interpretation.

Exercises

For a mode of speed $s$ relative to the æther, let its effective covariant metric be $g^{(s)}_{ab}=g_{ab}+(1-s^2)u_a u_b$. Show that the stationary finite-speed horizon approaches $u\cdot\chi=0$ as $s\to\infty$, and state what this limit does not prove.

Solution

The stationary horizon is where the Killing vector is null in the effective metric:

$$0=g^{(s)}_{ab}\chi^a\chi^b =\chi^2+(1-s^2)(u\cdot\chi)^2.$$

Divide by $s^2$:

$$0=\frac{\chi^2}{s^2}+\left(\frac1{s^2}-1\right)(u\cdot\chi)^2.$$

If $\chi^2$ remains finite, the $s\to\infty$ limit gives $(u\cdot\chi)^2=0$, hence $u\cdot\chi=0$. This independently recovers the candidate universal-horizon condition from the nested finite-speed cones. It does not prove that such a root exists, that the foliation is regular there, that ultraviolet dispersion is captured by a constant-$s$ metric, or that nonspherical perturbations preserve the surface.

References

• Barausse, E., and T. P. Sotiriou. “Slowly Rotating Black Holes in Hořava–Lifshitz Gravity.” Physical Review D 87 (2013) 087504. DOI. Open PDF.
• Barausse, E., T. Jacobson, and T. P. Sotiriou. “Black Holes in Einstein-Aether and Hořava–Lifshitz Gravity.” Physical Review D 83 (2011) 124043. DOI. Open PDF.
• Barausse, E., T. P. Sotiriou, and I. Vega. “Slowly Rotating Black Holes in Einstein-æther Theory.” Physical Review D 93 (2016) 044044. DOI. Open PDF.
• Berglund, P., J. Bhattacharyya, and D. Mattingly. “Mechanics of Universal Horizons.” Physical Review D 85 (2012) 124019. DOI. Open PDF.
• Blas, D., and S. Sibiryakov. “Hořava Gravity versus Thermodynamics: The Black Hole Case.” Physical Review D 84 (2011) 124043. DOI. Open PDF.