Skip to content

Covariant Phase Space, Boundaries, and Charges

Scrollable formula. Use Left and Right arrow keys to reveal hidden content.

Scrollable table. Use Left and Right arrow keys to reveal additional columns.

Scrollable figure. Use Left and Right arrow keys to inspect hidden content.

Covariant phase space begins with the boundary term in a Lagrangian variation and ends, when the necessary conditions hold, with Hamiltonians for spacetime symmetries. The route is boundary sensitive: a locally closed current can carry flux, a diffeomorphism can cease to be pure gauge at a boundary, and a finite surface variation can fail to integrate to a charge. This chapter keeps those logical gates separate and fixes the sign convention used throughout.

Helpful background. The Einstein Equation supplies the solution space; Hypersurface Normals and Tangential Projectors supplies orientations and cuts; The Gibbons–Hawking–York Boundary Term supplies the boundary variational problem; Presymplectic Potentials and Currents supplies field-space forms; Killing Fields and Isometry Groups supplies symmetry generators; and Formulations of General Relativity orients the volume.

For dynamical fields ϕ\phi and a covariant Lagrangian nn-form,

δL=E ⁣⋅ ⁣δϕ+dΘ(ϕ;δϕ).\delta L=E\!\cdot\!\delta\phi+d\Theta(\phi;\delta\phi).

This chapter adopts the Iyer–Wald ordering

ω(ϕ;δ1ϕ,δ2ϕ)=δ2Θ(ϕ;δ1ϕ)−δ1Θ(ϕ;δ2ϕ).\omega(\phi;\delta_1\phi,\delta_2\phi) =\delta_2\Theta(\phi;\delta_1\phi) -\delta_1\Theta(\phi;\delta_2\phi).

For an allowed, field-independent vector ξa\xi^a, the Einstein Noether construction leads to

Jξ=Θ(Lξϕ)−ξ⌟L,Jξ≐dQξ,J_\xi=\Theta(\mathcal L_\xi\phi) -\xi\mathbin{\lrcorner}L, \qquad J_\xi\doteq dQ_\xi,

and the candidate Hamiltonian variation on a cut SS is

̸ ⁣δHξ=∫S(δQξ−ξ⌟Θ).\not\!\delta H_\xi =\int_S (\delta Q_\xi-\xi\mathbin{\lrcorner}\Theta).

Iyer and Wald derive this chain in Iyer and Wald 1994, §§3–6, pp. 6–16, PDF. With the displayed ordering, the corresponding bulk identity is ̸ ⁣δHξ=ΩΣ(Lξϕ,δϕ)\not\!\delta H_\xi=\Omega_\Sigma(\mathcal L_\xi\phi,\delta\phi) when S=∂ΣS=\partial\Sigma and the stated on-shell hypotheses hold. Reversing the arguments reverses this sign; it does not change the surface-charge normalization. Each arrow adds hypotheses: equations and linearized equations for current closure, boundary conditions for a phase space, falloff for finiteness, field-space closure for integrability, and normalization for a finite charge.

The six pages form the following sequence:

  1. Covariant Presymplectic Current derives Θ\Theta, ω\omega, its closure, and its exact-form ambiguities.
  2. Diffeomorphism Noether Currents and Charges decomposes JξJ_\xi into constraints and dQξdQ_\xi and derives the Einstein charge form.
  3. Gauge Degeneracies and Boundary Symmetries tests when Lξg\mathcal L_\xi g is a degenerate direction and when it changes a boundary charge.
  4. Symplectic Flux and Boundary Conditions relates hypersurface independence to lateral flux and source–response conditions.
  5. Charge Integrability and Ambiguities checks whether a surface variation is exact on solution space and records its remaining choices.
  6. Komar, ADM, and Covariant Charges Compared identifies the stationary and asymptotic hypotheses under which familiar formulas agree.

The ordering prevents a common shortcut: writing down a Komar-like surface integral before defining the allowed perturbations, flux, and normalization that give it a Hamiltonian meaning.

PropertyRequired checkFailure means
DifferentiabilityBoundary variation is fixed or canceledNo well-defined variational principle
ConservationNet symplectic flux vanishes or is includedBulk form depends on the slice
IntegrabilityAntisymmetrized field variations vanishNo finite Hamiltonian function on that phase space
Gauge reductionPairing vanishes for every tangent perturbationThe transformation is boundary-active, not pure gauge

Finiteness is a fifth requirement: every displayed surface integral must converge. These properties constrain one another but none substitutes for all the others. Compère states the charge variation, integrability, and conservation criteria in Compère 2019, §§1.4.1–1.4.3, pp. 22–23, PDF.

Boundaries determine the physical question

Section titled “Boundaries determine the physical question”

The same bulk Einstein Lagrangian supports different phase spaces at spatial infinity, null infinity, a horizon, or a finite timelike wall. The boundary conditions select allowed perturbations and symmetries; counterterms select a compatible potential representative; corner data control exact-form shifts; and flux determines whether the bulk is closed or open. Consequently there is no boundary-independent list of gravitational charges.

This chapter develops the machinery, not the classification of every asymptotic symmetry algebra or the thermodynamic laws of particular spacetimes. Field-dependent generators, anomalous boundaries, and nonintegrable radiative charges require extensions of the fixed-generator formulas stated here.

For an exact symmetry ξa\xi^a and a linearized solution δg\delta g, state the surface variation that appears on two boundary components, then list the two additional checks needed before calling it a conserved finite charge.

Solution

The exact-symmetry identity gives the closed surface form

kξ(δg)=δQξ−ξ⌟Θ(g;δg),dkξ=0.k_\xi(\delta g) =\delta Q_\xi -\xi\mathbin{\lrcorner}\Theta(g;\delta g), \qquad dk_\xi=0.

Its integrals on homologous cuts agree when there is no intervening singularity or unaccounted boundary flux. One must still check that the integral is finite under the declared falloff and that the field-space one-form is integrable. After those checks, a reference solution fixes the additive constant. If ξa\xi^a is field dependent, further variation terms are required before any of these conclusions are used.

  • Compère, Geoffrey. Advanced Lectures on General Relativity. Lecture Notes in Physics 952. Cham: Springer, 2019. DOI; Open PDF.
  • Iyer, Vivek, and Robert M. Wald. “Some Properties of Noether Charge and a Proposal for Dynamical Black Hole Entropy.” Physical Review D 50, no. 2 (1994): 846–864. DOI; Open PDF.