Metric Counterterms and the Boundary Stress Tensor

The boundary stress tensor of an asymptotically locally AdS solution is the variation of the renormalized gravitational action with respect to the boundary metric. The Gibbons–Hawking term makes the Dirichlet problem well posed; intrinsic curvature counterterms remove radial divergences; finite terms and anomalies fix the remaining convention-dependent contacts. The resulting Brown–York tensor is finite and obeys the correct Ward identities, but its vacuum energy can depend on dimension, boundary geometry, and scheme.

Required background. Currents, stress tensor, and bulk metric fields fixes the dictionary. Fefferman–Graham expansions identifies the metric response. Helpful background. Boundary stress and surface counterterms and stress-tensor curvature ambiguities give the curved-spacetime comparison.

First application. Renormalize the on-shell action of global AdS and compute the boundary stress tensor and vacuum energy in a fixed dimension.

A finite gravitational variational problem

Use the Euclidean action convention

$$I_{\mathrm{reg}} =-\frac{1}{2\kappa^2}\int_M\mathrm d^{d+1}x\sqrt G \left(R+\frac{d(d-1)}{L^2}\right) -\frac{1}{\kappa^2}\int_{\partial M}\mathrm d^d x\sqrt\gamma\,K,$$

where $K_{ij}=\tfrac12\mathcal L_n\gamma_{ij}$ and $n$ is the outward unit normal of the regulated region. A source using the opposite curvature, normal, or Euclidean-action sign must be translated as a complete set.

For $d>2$, the first intrinsic counterterms are

$$I_{\mathrm{ct}} =\frac{1}{\kappa^2}\int_{\partial M}\mathrm d^d x\sqrt\gamma \left[ \frac{d-1}{L} +\frac{L}{2(d-2)}R[\gamma] +\cdots \right].$$

Higher-curvature terms cancel subsequent divergences. When their recursion denominator vanishes, a logarithmic counterterm encodes the gravitational Weyl anomaly. The ellipsis cannot be truncated solely by counting powers of curvature if the boundary dimension requires the next term.

Brown–York tensor and its renormalized limit

At the cutoff, variation defines

$$T^{ij}_{\epsilon} =\frac{2}{\sqrt\gamma} \frac{\delta(I_{\mathrm{reg}}+I_{\mathrm{ct}})}{\delta\gamma_{ij}}.$$

The Gibbons–Hawking contribution gives the Brown–York combination $K^{ij}-K\gamma^{ij}$ in the convention above; counterterms add local tensors such as $-(d-1)\gamma^{ij}/L$ and the boundary Einstein tensor. After the conformal rescaling to $g_{(0)}$,

$$\langle T_{ij}\rangle =\lim_{\epsilon\to0} \epsilon^{2-d}T_{\epsilon,ij},$$

with the power adjusted if another defining function is used. In Fefferman–Graham variables it has the form

$$\langle T_{ij}\rangle =\frac{dL^{d-1}}{2\kappa^2}g_{(d)ij} +X_{ij}[g_{(0)}],$$

where $X_{ij}$ is a local curvature functional fixed by lower coefficients, anomaly terms, and finite scheme. The precise prefactor follows from the convention for the $z^{-2}g_{ij}$ metric and must be checked against the stress-tensor two-point normalization.

Bulk momentum constraints yield

$$\nabla^i\langle T_{ij}\rangle =\langle\mathcal O\rangle\nabla_j\phi_{(0)} +F_{ji}^{(0)}\langle J^i\rangle+\cdots,$$

while radial rescalings yield the trace identity and anomaly. A finite tensor that violates these identities signals a missing boundary term or inconsistent source convention.

Global AdS as a normalization fixture

Poincaré AdS with flat boundary and its vacuum state has vanishing renormalized stress tensor in the standard flat-space scheme. Global AdS has boundary $\mathbb R\times S^{d-1}$ and may carry a curvature-induced Casimir energy. For global AdS$_5$ in the standard Balasubramanian–Kraus normalization,

$$E_{\mathrm{vac}}=\frac{3\pi L^2}{32G_5}.$$

The result agrees with the Casimir energy of the dual four-dimensional CFT after using the AdS/CFT central-charge map Balasubramanian and Kraus 1999. This is a useful round trip: the same counterterm signs must give zero on the flat Poincaré boundary, the stated nonzero value on the cylinder, conservation, and the correct trace anomaly.

Finite local curvature counterterms can shift scheme-dependent pieces in dimensions where allowed. A universal anomaly coefficient or a charge defined by a fixed variational principle cannot be changed arbitrarily under the name of scheme.

Stress tensor, charge, and scope

Given a boundary Killing field $\xi^i$ and spatial slice $\Sigma$ with unit normal $u^i$, the conserved charge is

$$Q[\xi]=\int_\Sigma\mathrm d^{d-1}x\sqrt\sigma\, u^i\langle T_{ij}\rangle\xi^j.$$

The formula assumes vanishing flux through the remaining boundary and a fixed conformal representative. For time-dependent sources, nonreflecting boundary conditions, or anomalies, the corresponding balance law replaces conservation.

This construction determines finite one-point functions and charges of a declared asymptotic problem. It does not decide which bulk saddles belong to a nonperturbative path integral or prove that the boundary stress tensor generates a complete microscopic theory.

Exercises

Why is canceling the leading volume divergence insufficient to define $\langle T_{ij}\rangle$ on a curved boundary?

Solution

Subleading divergences depend on intrinsic curvature and require curvature counterterms. Their variations contribute to the stress tensor. In even boundary dimension a logarithmic term also supplies the anomaly. Omitting these terms can leave $T_{ij}$ divergent or violate its trace and conservation identities even if the total on-shell action’s leading divergence is gone.

The chapter overview contains the structure diagram and validity and failure diagram. They are embedded there once so that their shared chapter-level context is not repeated on every article.

For the chapter-wide comparison of assumptions, counterevidence, falsifiers, and claim ceilings, see the claim-domain table.

References

• Balasubramanian, V., and Kraus, P. “A Stress Tensor for Anti-de Sitter Gravity.” Communications in Mathematical Physics 208 (1999): 413–428. DOI. arXiv.
• de Haro, S., Solodukhin, S. N., and Skenderis, K. “Holographic Reconstruction of Spacetime and Renormalization in the AdS/CFT Correspondence.” Communications in Mathematical Physics 217 (2001): 595–622. DOI. arXiv.
• Henningson, M., and Skenderis, K. “The Holographic Weyl Anomaly.” Journal of High Energy Physics 1998, 023 (1998). DOI. arXiv.