---
title: Universal-Horizon Thermodynamics
url: https://www.emergentmind.com/topics/universal-horizon-thermodynamics
type: topic
---

# Universal-Horizon Thermodynamics

Universal-horizon thermodynamics denotes a family of horizon-thermodynamic programs in which gravitational dynamics are rewritten as thermodynamic relations at distinguished causal or quasi-causal boundaries. In Lorentz-violating gravity, the term refers most directly to **universal horizons**—hypersurfaces defined by $(u\cdot\chi)=0$ that trap modes with arbitrarily high velocities and admit a first law, a temperature, and an area-proportional entropy. In cosmology and modified gravity, closely related literature uses **universal thermodynamics** to treat an FRW universe bounded by an apparent, event, or Ricci-scale horizon as a thermodynamic system obeying a Clausius relation, a unified first law, and the generalized second law of thermodynamics (GSLT). The common theme is that horizon entropy is theory-dependent, whereas the thermodynamic structure is comparatively robust [1309.0907][1110.6101][1503.03059].

## 1. Terminological scope and geometric definitions

Two distinct but connected usages appear in the literature.

| Usage | Horizon | Defining relation |
|---|---|---|
| Cosmological universal thermodynamics | Apparent, event, Ricci-scale horizon | $R_A=1/H$, $R_E=a\int_t^\infty dt'/a(t')$, $R_L=(2H^2+\dot H)^{-1/2}$ |
| Universal-horizon thermodynamics in Lorentz-violating gravity | Universal horizon | $(u\cdot\chi)=0$ |

In the cosmological usage, the system is the **matter inside a cosmological horizon plus the horizon itself**. For a flat FRW spacetime,
\[
ds^2=-dt^2+a^2(t)\bigl[dr^2+r^2d\Omega_2^2\bigr],
\]
the apparent horizon is at $R_A=1/H$, the event horizon is
\[
R_E=a(t)\int_t^\infty \frac{dt'}{a(t')},
\]
and the Ricci length scale is
\[
R_L=\left(2H^2+\dot H\right)^{-1/2}.
\]
These surfaces are used as thermodynamic boundaries for the universe as a whole [1110.6101].

In Einstein-æther theory and the IR limit of Hořava–Lifshitz gravity, the preferred timelike vector field $u^a$ defines a preferred foliation by æther-time hypersurfaces. In a static spacetime with Killing vector $\chi^a$, the **universal horizon** is the hypersurface where
\[
(u\cdot\chi)=u_a\chi^a=0.
\]
Because physical propagation is constrained to move toward the future in æther time, this hypersurface is a one-way causal boundary even for modes with unbounded speed relative to the metric light cone [1210.4940][1408.6479].

A recurrent source of confusion is the distinction between **universal thermodynamics** and **universal horizons**. The former concerns the thermodynamics of the universe bounded by cosmological horizons; the latter concerns the causal horizon that replaces the Killing horizon in Lorentz-violating black-hole spacetimes. The two programs share techniques—surface gravity, Clausius relations, entropy functionals, Smarr formulas—but not the same causal structures.

## 2. FRW horizon thermodynamics and the generalized second law

For an interacting $n$-fluid system in a flat FRW universe, the component conservation laws are
\[
\dot{\rho}_i+3H(\rho_i+p_i)=Q_i,\qquad \sum_i Q_i=0,
\]
which combine into the effective single-fluid conservation law
\[
\dot{\rho}+3H(\rho+p)=0.
\]
With Einstein equations
\[
3H^2=\rho,\qquad 2\dot H=-(\rho+p),
\]
the thermodynamic analysis depends only on the total effective $\rho$ and $p$, not on the individual fluids or interaction terms [1110.6101].

For a horizon of radius $R_h$, the energy crossing the horizon in time $dt$ is
\[
-dE=4\pi R_h^3H(\rho+p)\,dt,
\]
and the Clausius relation is written as
\[
-dE=T_h\,dS_h.
\]
Using the Gibbs equation for the interior matter,
\[
T_h\,dS_I=dE_I+p\,dV,
\]
with $V=\frac{4}{3}\pi R_h^3$, the total entropy change is
\[
\frac{dS_{\text{tot}}}{dt}
=
\frac{4\pi R_h^2}{T_h}\,\dot R_h\,(\rho+p).
\]
Hence GSLT requires
\[
\dot R_h(\rho+p)\ge 0.
\]
This gives the two standard cases: a quintessence-like era with $\rho+p>0$ and $\dot R_h\ge0$, or a phantom-like era with $\rho+p<0$ and $\dot R_h\le0$ [1110.6101].

The same work reformulates the condition in terms of the statefinder parameters
\[
r=\frac{1}{aH^3}\frac{d^3a}{dt^3},\qquad
s=\frac{r-1}{3\left(q-\frac12\right)},
\]
leading to
\[
\rho+p=3H^2\left[1+\frac{2(r-1)}{9s}\right].
\]
Then
\[
\frac{dS_{\text{tot}}}{dt}
=
\frac{12\pi R_h^2H^2}{T_h}
\left[1+\frac{2(r-1)}{9s}\right]\dot R_h.
\]
For the apparent horizon, $R_A=1/H$ implies
\[
\dot R_A=\frac32\left[1+\frac{2(r-1)}{9s}\right],
\]
and therefore
\[
\frac{dS_{\text{tot}}}{dt}
=
\frac{18\pi R_A^2H^2}{T_h}
\left[1+\frac{2(r-1)}{9s}\right]^2\ge0.
\]
In Einstein gravity, GSLT is therefore always satisfied at the apparent horizon. For the event horizon and Ricci-scale horizon, the allowed regions in the $\{r,s\}$ plane are conditional and horizon-dependent [1110.6101].

This cosmological program treats horizon thermodynamics as a diagnostic of the effective matter sector. A central result is that the interacting fluid system can be handled thermodynamically as a single effective fluid, so the validity of GSLT depends on total NEC-type information, not on the microscopic partition of the dark sector.

## 3. Unified first law, corrected temperatures, and modified entropy functionals

A broad modified-gravity literature rewrites the FRW dynamics as Hayward’s unified first law
\[
dE=A\psi+WdV,
\]
where $E$ is the Misner–Sharp energy, $W$ is the work density, and $\psi_a=T_a{}^b\partial_bR+W\partial_aR$. Projecting the unified first law along a vector tangent to the horizon yields a first-law-like identity. The **matter** contribution to the energy-supply vector is identified with the heat flow,
\[
\delta Q=\langle A\psi_m,\xi\rangle,
\]
and the remaining effective-gravity sector is absorbed into a modified entropy so that $\delta Q=T_h\,dS_h$ holds in equilibrium form [1503.03059][1503.07750].

In this framework, the apparent-horizon surface gravity is
\[
\kappa_A=-\frac{1-\epsilon}{R_A},\qquad
\epsilon=\frac{\dot R_A}{2HR_A},
\]
and the corresponding corrected or extended Hawking temperature is
\[
T_A=\frac{1-\epsilon}{2\pi R_A}.
\]
Related corrected temperatures are also used on the event horizon. The purpose of these dynamical factors is to preserve a consistent Clausius relation in nonstationary cosmological settings [1610.09283][1503.07750].

The resulting entropies are generally of the form
\[
S_h=\frac{A_h}{4G}+(\text{theory-dependent correction}),
\]
with the correction typically written as an integral over effective fluid variables. Representative examples include:

- In $f(R)$ gravity, the apparent-horizon entropy becomes
  \[
  S_A=\frac{A_A}{4G}
  -
  \frac{1}{4G}
  \int HR_A\left(\dot F-HF_1+2F_1\left(\dot H-\frac{k}{a^2}\right)\right)dt,
  \]
  where $F_1=dF/dR$ for $f(R)=R+F(R)$ [1503.03059].

- In Einstein-frame $f(R)$ gravity, the scalar degree of freedom produces
  \[
  S_A=\frac{A_A}{4G}-\frac{3}{4G}\int \dot\phi^2\,HR_A^3\,dt
  \]
  on the apparent horizon [1503.03059].

- In Einstein–Gauss–Bonnet gravity, the apparent-horizon entropy acquires a logarithmic correction,
  \[
  S_A=\frac{A_A}{4G}+\frac{4\pi\tilde\alpha}{G}\ln(R_A),
  \]
  while the event-horizon entropy carries an integral correction [1503.03059].

- In RS-II and DGP braneworld cosmologies, both apparent and event horizons acquire integral corrections proportional to the effective braneworld sector, while the extended Hawking temperature is retained [1503.07750].

- In Lanczos–Lovelock gravity, both $S_A$ and $S_E$ are given by $A/4$ plus Lovelock-dependent integral terms involving $\sum_{i=2}^m i\hat c_i/R_A^{2i}$ [1610.09947].

- In massive gravity with de Sitter reference metric, the apparent-horizon entropy takes the closed form
  \[
  S_A
  =
  \frac{A_A}{4G}
  -
  \frac{\pi m_g^2}{G H_c}
  \left[
  \beta R_A^3-\frac{\gamma}{H_c}R_A^2+\frac{3\delta}{H_c^2}\frac{1}{R_A}
  \right],
  \]
  while the event-horizon entropy is $A_E/(4G)$ plus an integral correction [1702.05372].

These results support an equilibrium interpretation of modified-gravity horizon thermodynamics: rather than introducing an entropy production term, one changes the entropy functional. A plausible implication is that the thermodynamic content of a gravity theory is encoded less by the formal existence of a first law than by the precise horizon entropy required to make the first law hold.

## 4. Universal horizons in Einstein-æther and Hořava–Lifshitz gravity

Einstein-æther theory is a generally covariant metric theory coupled to a dynamical timelike unit vector $u^a$ satisfying
\[
u^a u_a=-1.
\]
Its action is
\[
S=\frac{1}{16\pi G_{\ae}}\int d^4x\,\sqrt{-g}\,\bigl(R+L_{\ae}\bigr),
\]
with
\[
L_{\ae}
=
-
Z^{ab}{}_{cd}\,\nabla_a u^c\,\nabla_b u^d
+
\lambda(u^2+1),
\]
where the $c_i$ couplings enter through
\[
Z^{ab}{}_{cd}
=
c_1 g^{ab}g_{cd}
+
c_2 \delta^a_c\delta^b_d
+
c_3 \delta^a_d\delta^b_c
-
c_4 u^a u^b g_{cd}.
\]
Because the æther selects a preferred local rest frame, local Lorentz invariance is spontaneously broken while diffeomorphism invariance is preserved [1309.0907].

In these theories, different field excitations can propagate on different effective metrics; there is no universal maximum speed. Consequently, the metric Killing horizon does not define the true black-hole boundary, since sufficiently superluminal modes can escape through it. The relevant causal boundary is the universal horizon, defined in a static spacetime by
\[
(u\cdot\chi)=0,
\]
where $\chi^a$ is the time-translation Killing vector. In Painlevé–Gullstrand-type coordinates one may write
\[
ds^2=-dt^2+(v(r)\,dt+f(r)\,dr)^2+r^2d\Omega^2,
\]
while in ingoing Eddington–Finkelstein form one may write
\[
ds^2=-e(r)\,dv^2+2f(r)\,dv\,dr+r^2d\Omega_2^2.
\]
Both coordinate systems are used to analyze static, spherically symmetric universal-horizon spacetimes [1210.4940][1408.6479].

A universal horizon is a spacelike hypersurface, not a null Killing horizon. Its causal role follows from the requirement that physical propagation moves toward the future in æther time. Constant æther-time hypersurfaces bend downward and asymptote to $(u\cdot\chi)=0$; inside that surface, moving to larger æther time forces motion to smaller radius. Hence any future-directed æther-causal signal, even one with arbitrarily high local speed, cannot escape from inside the universal horizon [1210.4940][1408.6479].

For static, spherically symmetric solutions with maximally symmetric asymptotics, the universal-horizon sector has been classified in Einstein-æther theory, including asymptotically flat, de Sitter, and—when $c_{14}=0$—anti-de Sitter analogues. Moreover, any static, spherically symmetric Hořava–Lifshitz solution with a regular universal horizon is also a solution of Einstein-æther theory, independent of asymptotic boundary conditions [1408.6479].

## 5. First law, Hawking radiation, and entropy at universal horizons

A first-law-like relation for universal horizons was first derived in static, spherically symmetric Einstein-æther black holes for special coupling sectors. For the analytic families with either $c_{14}=0$ or $c_{123}=0$, the first law takes the form
\[
\delta M_{\ae}
=
\frac{\kappa_{\text{UH}}}{8\pi G_{\ae}}(1-c_{13})\,\delta A_{\text{UH}},
\]
where $M_{\ae}$ is the total mass, $\kappa_{\text{UH}}$ is the universal-horizon surface gravity, and $A_{\text{UH}}$ is the universal-horizon area [1210.4940].

A tunneling calculation with a Lorentz-violating scalar field then showed that the universal horizon radiates as a blackbody at a fixed temperature even when the scalar field equations also violate local Lorentz invariance. For the class of solutions studied, the temperature can be expressed as
\[
T_{\text{UH}}
=
\frac{(a\cdot s)_{\text{UH}}}{4\pi}\,
\frac{\mathcal{C}_{\ae}}{\mathcal{C}_{\text{UH}}},
\]
and also as
\[
T_{\text{UH}}=\frac{\kappa_{\text{UH}}}{4\pi\mathcal{C}_{\ae}}
\]
for those solutions. This supports assigning thermodynamic entropy to the universal horizon if the generalized second law is not to be violated [1210.4940].

The Noether-charge derivation sharpened this picture. Using the Wald–Iyer formalism, one finds for static, spherically symmetric Einstein-æther black holes
\[
\delta\mathcal{E}
=
\frac{1}{8\pi G_{\ae}}
\left[
\kappa_{\text{UH}}(1-c_{13})
+\frac{c_{123}}{2}K_{\text{UH}}\|\xi\|_{\text{UH}}
\right]
\delta A_{\text{UH}},
\]
and this is written as
\[
\delta\mathcal{E}=T_{\text{UH}}\,\delta S_{\text{UH}},
\qquad
S_{\text{UH}}=\frac{A_{\text{UH}}}{4}.
\]
The temperature extracted from the Noether charge matches the temperature found in tunneling calculations [1309.0907].

This result is technically significant because the usual Wald derivation at a Killing-horizon bifurcation surface fails in Einstein-æther theory: the æther cannot be both regular and Lie-dragged there. Replacing the inner boundary with a universal-horizon cross-section avoids this obstruction. The conclusion is explicit: in Lorentz-violating theories of this type, thermodynamic properties should be ascribed to the universal horizon, not to the Killing horizon [1309.0907].

The same sector also admits Smarr-type relations. In static, spherically symmetric Einstein-æther and Hořava–Lifshitz solutions, a divergence-free two-form leads to
\[
M_{\ae}=\frac{q_{\text{UH}}A_{\text{UH}}}{4\pi G_{\ae}},
\]
with $q_{\text{UH}}$ constructed from $\kappa_{\text{UH}}$, $K_{\text{UH}}$, and the cosmological contribution. For one-parameter asymptotically flat families this reduces to
\[
\delta M_{\ae}
=
\frac{q_{\text{UH}}}{8\pi G_{\ae}}\,\delta A_{\text{UH}},
\]
again giving a pure $T\,\delta S$ structure [1408.6479].

## 6. Universality, model dependence, and the limits of the area law

The literature does not support a naive claim that every horizon in equilibrium necessarily obeys the entropy-area law. In Schwarzschild–de Sitter spacetime, the black-hole event horizon and cosmological event horizon have different local Hawking temperatures,
\[
T_b=\frac{\kappa_b}{2\pi\sqrt{f_w}},\qquad
T_c=\frac{\kappa_c}{2\pi\sqrt{f_w}},
\qquad
\kappa_b>\kappa_c,
\]
so the total system is not in global equilibrium. A thermodynamically consistent treatment requires **three independent state variables**, one of which represents the external gravitational influence of the other horizon. In that framework, the cosmological horizon entropy generically fails to equal $A_c/4$, and for the black-hole horizon the area law holds if and only if
\[
\partial_M X_b\equiv 0.
\]
This establishes a limit of universality for the entropy-area law in multi-horizon spacetimes [1109.0801].

A related lesson appears in horizon thermodynamics for Lovelock black holes. Evaluating the Lovelock field equations at the horizon yields a universal horizon first law
\[
dE=T\,dS-P\,dV,
\]
where $P$ is the **total pressure of all matter in the spacetime**, including a cosmological constant if present. With
\[
V=\frac{\Sigma_{d-2}r_+^{\,d-1}}{d-1}
\]
and the Lovelock-Wald entropy, the horizon equation of state
\[
P=P(T,r_+)
\]
reproduces Hawking–Page-like behavior, Van der Waals–like behavior, and the presence of a triple point, provided there is sufficient non-linearity in the gravitational sector [1603.05689]. This suggests that the robust part of universality lies in the **thermodynamic form of the horizon equations**, not in a single entropy functional or a single notion of pressure.

The cosmological modified-gravity literature reaches a similar conclusion. In analyses using interacting holographic dark energy and Planck data, the detailed validity of GSLT and thermodynamical equilibrium depends on the gravity theory and parameter choice: for example, $f(R)=R+R^2$ fails GSLT in the parameter space studied, Einstein–Gauss–Bonnet gravity admits a range $0\lesssim b^2\lesssim 0.4$ where both GSLT and TE hold, RS-II generally fails to realize both simultaneously, and DGP admits a broad region $b^2\gtrsim 0.6$ where both hold for suitable model parameters [1610.09283]. The same study explicitly states that there is **no definite conclusion whether apparent or event horizon is more favourable** in modified gravity, although in Einstein gravity the apparent horizon is thermodynamically preferred because GSLT is automatically satisfied there [1110.6101][1610.09283].

Taken together, these results support a precise but limited notion of universality. Horizon thermodynamics is widely reproducible across GR, Lovelock gravity, braneworld models, massive gravity, and Lorentz-violating Einstein-æther/Hořava sectors. What is not universal is the detailed entropy functional, the relevant horizon in every theory, or the parameter regime in which GSLT and thermodynamical equilibrium are satisfied. A plausible synthesis is that the universal content lies in the persistence of first-law and entropy-balance structures, while the microscopic meaning of horizon entropy remains theory-specific.

Source: https://www.emergentmind.com/topics/universal-horizon-thermodynamics