betaFeedback

WorksPhysics and Astronomy

Regular/Singular Black Hole Thermodynamics from Renormalised Volume/Area

Abstract

We study the finite mismatch between the holographically and topologically renormalised Euclidean actions of the four-dimensional AdS C-metric. Starting from the relation between Renormalised Volume and the Einstein sector of Conformal Gravity, we separate the regular curvature contribution from the contributions supported on the two conical axes. After introducing the usual Euler normalisation for smooth manifolds, the residual difference between the regular action evaluations is determined by the tension-weighted Euler terms of the renormalised worldsheet area. We identify this difference with the Gauss–Bonnet entropy contribution at the polar conical points of the horizon. The ordinary Euler characteristic remains fixed, while the deficits determine how its curvature representation is divided between regular and singular contributions. At fixed total string tension, the difference of the renormalised worldsheet areas per unit Euclidean time determines the relative tension work, giving a geometric interpretation of the relative thermodynamic length.

Introduction

The notion of Renormalised Volume arises naturally in differential geometry when studying noncompact manifolds whose ordinary Riemannian volume is infinite, but whose asymptotic geometry is sufficiently controlled to extract a finite, geometrically meaningful quantity.

The principal setting is that of asymptotically hyperbolic manifolds, particularly Poincaré–Einstein manifolds, whose metrics approach hyperbolic space near a conformal boundary. Although the volume diverges, its asymptotic expansion contains a finite part that can be interpreted as a renormalised geometric invariant (with qualifications depending on the dimension and the choice of boundary representative).

The construction connects conformal geometry, Einstein manifolds, Chern–Gauss–Bonnet formulas, holographic renormalisation, conformal anomalies, and renormalised areas of minimal surfaces.

In the mathematical literature, the concept of renormalised Volume VrenV_{\mathrm{ren}} for asymptotically hyperbolic spaces is closely related to the existence of bulk conformal invariants in even-dimensional manifolds [1]. Indeed, in four spacetime dimensions, a theorem due to Anderson [2] states that the renormalised Volume is proportional to the integral of a Weyl-squared term

132π2∫M4d4x−g WαβμνWαβμν=χ(M4)−34π2Vren(1)\frac{1}{32\pi^{2}}\int_{M_{4}} \mathrm{d}^{4}x\sqrt{-g}\,W^{\alpha\beta\mu\nu}W_{\alpha\beta\mu\nu}=\chi(M_{4})-\frac{3}{4\pi^{2}}V_{\mathrm{ren}} \tag*{(1)}

up to the Euler characteristic of the manifold χ(M4)\chi(M_{4}).

Thought of as a gravity theory, the left-hand side of eq. (1) is proportional to the action of Conformal Gravity, ICG=∫M4−g W2I_{\mathrm{CG}}=\int_{M_{4}}\sqrt{-g}\,W^{2}, whose field equations are of higher order in derivatives of the metric tensor.

In point of fact, an arbitrary variation of the previous action produces the equations of motion in the bulk plus a surface term

δICG=∫M4d4x−g Bμνδgμν+∫∂M4nμΘμ,(2)\delta I_{\mathrm{CG}}=\int_{M_{4}}\mathrm{d}^{4}x\sqrt{-g}\,B^{\mu\nu}\delta g_{\mu\nu}+\int_{\partial M_{4}}n_{\mu}\Theta^{\mu}, \tag*{(2)}

where BμνB^{\mu\nu} is the Bach tensor, which is a covariantly conserved, symmetric and traceless object. This tensor is conformally covariant under Weyl rescalings of the metric only in four dimensions, and it is defined as

B νμ=S λσW  νσμλ+□S νμ−∇σ∇μS νσ,(3)B^{\mu}_{\ \nu}=S^{\sigma}_{\ \lambda}W^{\mu\lambda}_{\ \ \nu\sigma}+\Box S^{\mu}_{\ \nu}-\nabla_{\sigma}\nabla^{\mu}S^{\sigma}_{\ \nu}, \tag*{(3)}

with SμνS_{\mu\nu} being the Schouten tensor, which, in terms of the Ricci tensor and the Ricci scalar, is written as

Sμν=12(Rμν−16Rgμν).(4)S_{\mu\nu}=\frac{1}{2}\left(R_{\mu\nu}-\frac{1}{6}Rg_{\mu\nu}\right). \tag*{(4)}

Θμ\Theta^{\mu} is a boundary term and nμn_{\mu} is a normal vector to the boundary.

In six dimensions, a formula for Renormalised Volume was proposed by Chang, Qing, and Yang [3].

It is straightforward to check that Einstein-AdS spaces

Rμν=−3ℓ2gμν,(5)R_{\mu\nu}=-\frac{3}{\ell^{2}}g_{\mu\nu}, \tag*{(5)}

satisfy the field equations of Conformal Gravity, defining a generic sector of the theory. The length ℓ\ell is a characteristic scale given by the AdS radius.

From a holographic standpoint, in Einstein gravity, the description of an asymptotically AdS space is given in terms of the Fefferman–Graham coordinate frame [4]

ds2=ℓ2z2(dz2+gˉij(z,x)dxidxj),(6)ds^{2}=\frac{\ell^{2}}{z^{2}}\left(dz^{2}+\bar{g}_{ij}(z,x)dx^{i}dx^{j}\right), \tag*{(6)}

which endows the boundary with a conformal structure. The boundary metric gˉij(z,x)\bar{g}_{ij}(z,x) is a series in the zz holographic coordinate, which starts with the corresponding source g(0)ijg_{(0)ij} and, in the case of four dimensions, takes the particular form

gˉij(z,x)=g(0)ij(x)+z2ℓ2g(2)ij(x)+z3ℓ3g(3)ij(x)+⋯ .(7)\bar{g}_{ij}(z,x)=g_{(0)ij}(x)+\frac{z^2}{\ell^2}g_{(2)ij}(x)+\frac{z^3}{\ell^3}g_{(3)ij}(x)+\cdots. \tag*{(7)}

The mode at the holographic order z3z^3, i.e., g(3)ijg_{(3)ij}, represents the holographic stress tensor, whose vanishing trace is determined by the field equations. The geometry characterised by the holographic data g(0)ijg_{(0)ij} is assumed to be smooth, i.e., free of higher codimension defects.

In turn, the holographic consistency of the equations of motion in Conformal Gravity requires an additional coefficient g(1)ijg_{(1)ij}

gˉij(z,x)=g(0)ij(x)+zℓg(1)ij(x)+z2ℓ2g(2)ij(x)+z3ℓ3g(3)ij(x)+⋯ ,(8)\bar{g}_{ij}(z,x)=g_{(0)ij}(x)+\frac{z}{\ell}g_{(1)ij}(x)+\frac{z^2}{\ell^2}g_{(2)ij}(x)+\frac{z^3}{\ell^3}g_{(3)ij}(x)+\cdots, \tag*{(8)}

where the holographic stress tensor also contains a response associated with the new source. It was argued in Ref. [5] that Neumann boundary conditions for the metric gˉij\bar{g}_{ij} provide a mechanism to turn Conformal Gravity into Einstein theory. Indeed, the condition ∂zgˉij=0\partial_z\bar{g}_{ij}=0 implies that g(1)ij=0g_{(1)ij}=0, such that the metric (8) reduces to the expansion (7) in Einstein gravity.

At the level of the action, the Einstein condition on the spacetime may be implemented either directly by Eq. (5) or by the vanishing of the traceless Ricci tensor

Hμν=Rμν−14gμνR.(9)H_{\mu\nu}=R_{\mu\nu}-\frac{1}{4}g_{\mu\nu}R. \tag*{(9)}

The latter equation readily implements the Neumann boundary condition by Maldacena [6]. It also implies the rest of the holographic relations for asymptotically AdS spaces in Einstein gravity, including the fact that the holographic stress tensor g(3)ijg_{(3)ij} is transverse and traceless.

From the general expression for the Weyl tensor

Wμναβ=Rμναβ−Sμαδνβ+Sναδμβ+Sμβδνα−Sνβδμα,(10)W^{\alpha\beta}_{\mu\nu}=R^{\alpha\beta}_{\mu\nu}-S^\alpha_{\mu}\delta^\beta_{\nu}+S^\alpha_{\nu}\delta^\beta_{\mu}+S^\beta_{\mu}\delta^\alpha_{\nu}-S^\beta_{\nu}\delta^\alpha_{\mu}, \tag*{(10)}

for Einstein-AdS, one obtains a reduced form of this tensor

W(E)μναβ=Rμναβ+1ℓ2δμναβ.(11)W^{\alpha\beta}_{(E)\mu\nu}=R^{\alpha\beta}_{\mu\nu}+\frac{1}{\ell^2}\delta^{\alpha\beta}_{\mu\nu}. \tag*{(11)}

From here on, we use the superscript (E)(E) to denote the Einstein sector. Therefore, the Renormalised Volume for Euclidean Einstein-AdS is given as

132π2∫M4d4xg W(E)αβμνW(E)αβμν=χ(M4)−34π2ℓ4Vren(E).(12)\frac{1}{32\pi^2}\int_{M_4}d^4x\sqrt{g}\,W^{\alpha\beta\mu\nu}_{(E)}W_{(E)\alpha\beta\mu\nu}=\chi(M_4)-\frac{3}{4\pi^2\ell^4}V^{(E)}_{\mathrm{ren}}. \tag*{(12)}

The above expression is connected to the definition of the MacDowell-Mansouri action [7, 8].

IMM=−ℓ2256πG∫M4d4xg δμ1⋯μ4ν1⋯ν4(Rν1ν2μ1μ2+δν1ν2μ1μ2ℓ2)(Rν3ν4μ3μ4+δν3ν4μ3μ4ℓ2).I_{\mathrm{MM}}=-\frac{\ell^2}{256\pi G}\int_{M_4}d^4x\sqrt{g}\,\delta^{\nu_1\cdots\nu_4}_{\mu_1\cdots\mu_4}\left(R^{\mu_1\mu_2}_{\nu_1\nu_2}+\frac{\delta^{\mu_1\mu_2}_{\nu_1\nu_2}}{\ell^2}\right)\left(R^{\mu_3\mu_4}_{\nu_3\nu_4}+\frac{\delta^{\mu_3\mu_4}_{\nu_3\nu_4}}{\ell^2}\right).

Using the identity 1

δν1ν2ν3ν4μ1μ2μ3μ4W(E)μ1μ2ν1ν2W(E)μ3μ4ν3ν4=16ℓ2R+96ℓ4+4G,(13)\delta^{\mu_1\mu_2\mu_3\mu_4}_{\nu_1\nu_2\nu_3\nu_4}W^{\nu_1\nu_2}_{(E)\mu_1\mu_2}W^{\nu_3\nu_4}_{(E)\mu_3\mu_4}=\frac{16}{\ell^2}R+\frac{96}{\ell^4}+4\mathcal{G}, \tag*{(13)}

with

G=R2−4RμνRμν+RμναβRμναβ(14)\mathcal{G}=R^2-4R_{\mu\nu}R^{\mu\nu}+R^{\mu\nu\alpha\beta}R_{\mu\nu\alpha\beta} \tag*{(14)}

the MacDowell-Mansouri action becomes the Einstein-AdS gravity action augmented by the Gauss-Bonnet invariant with a fixed coupling,

IMM=Itop=−116πG∫M4d4xg(R+6ℓ2+ℓ24G)(15)I_{MM}=I_{\mathrm{top}}=-\frac{1}{16\pi G}\int_{M_4}d^4x\sqrt{g}\left(R+\frac{6}{\ell^2}+\frac{\ell^2}{4}\mathcal{G}\right) \tag*{(15)}

In the Einstein branch, R=−12/ℓ2R=-12/\ell^2, and G=W(E)2+24/ℓ4\mathcal{G}=W_{(E)}^2+24/\ell^4, so the constant terms cancel leading to

Itop=ℓ264πG∫M4d4xg W(E)2(16)I_{\mathrm{top}}=\frac{\ell^2}{64\pi G}\int_{M_4}d^4x\sqrt{g}\,W_{(E)}^2 \tag*{(16)}

The addition of the Gauss-Bonnet term does not affect the equations of motion. The coupling ℓ2/4\ell^2/4 is singled out by a finite action for asymptotically AdS Einstein spaces [9]. In this way, the consistent inclusion of a topological term of the Euler class renormalises the gravitational action in a fully covariant form. This defines the concept of Topological Renormalisation [10, 11]. In point of fact, the Fefferman-Graham coordinate frame (6)–(7), when plugged into the on-shell variation of the MacDowell-Mansouri action into

δItop=ℓ264πG∫M4d3x−h nν1δμ1⋯μ4ν1⋯ν4δgμ2λΓλν2μ1(Rν3ν4μ3μ4+1ℓ2δν3ν4μ3μ4)(17)\delta I_{\mathrm{top}}=\frac{\ell^2}{64\pi G}\int_{M_4}d^3x\sqrt{-h}\,n_{\nu_1}\delta^{\nu_1\cdots\nu_4}_{\mu_1\cdots\mu_4}\delta g^{\mu_2\lambda}\Gamma^{\mu_1}_{\lambda\nu_2}\left(R^{\mu_3\mu_4}_{\nu_3\nu_4}+\frac{1}{\ell^2}\delta^{\mu_3\mu_4}_{\nu_3\nu_4}\right) \tag*{(17)}

reproduces the holographic content of four-dimensional AdS gravity, that is,

δItop=12∫∂M4d3x−g(0)(316πGℓg(3)ijδg(0)ij)(18)\delta I_{\mathrm{top}}=\frac{1}{2}\int_{\partial M_4}d^3x\sqrt{-g_{(0)}}\left(\frac{3}{16\pi G\ell}g_{(3)}^{ij}\delta g_{(0)ij}\right) \tag*{(18)}

In other words, the total variation of the action is free from infrared divergences and duly expressed in terms of the holographic stress tensor Tij=3g(3)ij/(16πGℓ)T^{ij}=3g_{(3)}^{ij}/(16\pi G\ell) and the variation of the holographic source g(0)ijg_{(0)ij} [12, 13].

A finite term proportional to the Euler characteristic drops out of this variation at fixed topology and couplings. This is more evident when comparing thermodynamic quantities. For smooth manifolds, the topological action produces a constant shift in the entropy, as illustrated in [10] for the Kerr–AdS black hole. In the presence of conical deficits, the finite difference remains after applying the Euler normalisation appropriate to smooth fillings [14]. The slowly accelerating AdS C-metric provides a setting in which to trace this residual difference to the curvature supported on the conical axes. Unequal string tensions drive the acceleration, while the absence of an acceleration horizon allows the Euclidean time period to be fixed by regularity at the black hole horizon. Even in this regime, the thermodynamic description requires a choice of time normalisation and a specification of which deficit variations are allowed. The first law was formulated with fixed deficits [15], and later extended to include varying string tensions and their role as conjugate thermodynamic lengths [16]. See also [17] for the analysis of the charged C-metric first law using boost time. Holographic calculations related the mass and free energy for a suitable time normalisation [18, 19, 20]. In covariant phase space, Kim et al. [21] showed how the usual time normalisation is recovered at fixed total deficit and how holographic counterterms induce a corner contribution to the charges where the strings meet conformal infinity.

The role of the defect is more evident in three dimensions. The accelerating BTZ black hole [22, 23] illustrates how the defect can enter the renormalisation of the Euclidean action itself. The domain wall—the codimension-one analogue to the cosmic string—reaches the conformal boundary, and the usual asymptotic counterterms leave a subleading divergence. The wall contribution, including the internal Gibbons–Hawking–York terms, cancels this divergence and takes the Nambu–Goto form on shell by the Israel junction conditions [24, 25]. In the generalised Fefferman–Graham gauge, the divergence is cancelled by a corner counterterm where the wall meets the conformal boundary [26]. Both calculations recover the quantum statistical relation.

These results motivate a closer look at how the two renormalisation prescriptions account for the conical axes in four dimensions. Hale et al. [14] found a finite difference between the holographically and topologically renormalised Euclidean actions of charged accelerating black holes after applying Euler normalisation for smooth filling. We examine this mismatch for the neutral AdS C-metric. By separating the regular curvature from the contributions on the axes, we relate the residual difference to the tension-weighted Euler terms of the renormalised worldsheet areas and identify it with the Gauss–Bonnet entropy contribution at the polar conical points.

This paper is organised as follows: § 2 deals with black hole thermodynamics of regular horizons by directly comparing the Gauss–Bonnet term with holographic counterterms. § 3 introduces the C-metric solution to the Einstein equations, and highlights its main features. It also shows the results for the different prescriptions of Euclidean action evaluated in the C-metric. § 4 defines the notion of Renormalised Area, as a consequence of the addition of a topological term in the bulk. § 5 studies the contribution from conical defects to the black hole thermodynamics. The last section is devoted to some final remarks.

Counterterms in AdS gravity and Regular Thermodynamics

Within a more physical framework, the relation to a renormalised action based on the addition of counterterms [12, 27] is possible only if one breaks general covariance of the spacetime by introducing a boundary. This is explicitly carried out, e.g., in Gaussian coordinates

ds2=N2(z)dz2+hij(x,z)dxidxj.(19)ds^2=N^2(z)dz^2+h_{ij}(x,z)dx^i dx^j . \tag*{(19)}

In asymptotically AdS spaces, the lapse function NN is given by N=ℓ/zN=\ell/z, while the induced metric takes the form hij(x,z)=ℓ2gˉij/z2h_{ij}(x,z)=\ell^2\bar{g}_{ij}/z^2.

This comes from insisting on a Dirichlet problem for the boundary metric hijh_{ij} which, in principle, is ill-defined for asymptotically AdS spacetimes [28]. The inclusion of the Gibbons–Hawking term plus boundary counterterms, only when properly resummed– can be expressed as bulk-covariant quantities. The explicit proof goes as follows: consider a smooth Euclidean four-manifold with boundary. In our conventions, the Chern-Gauss-Bonnet theorem takes the following form

∫M4d4xg G=32π2χ(M4)−∫∂M4B3.(20)\int_{M_4} d^4x\sqrt{g}\,\mathcal{G}=32\pi^2\chi(M_4)-\int_{\partial M_4} B_3 . \tag*{(20)}

where B3B_3 denotes the density of the boundary Chern form,

B3=4h δj1j2j3i1i2i3Ki1j1(12Ri2i3j2j3−13Ki2j2Ki3j3).(21)B_3=4\sqrt{h}\,\delta^{i_1i_2i_3}_{j_1j_2j_3}K^{j_1}_{i_1}\left(\frac{1}{2}\mathcal{R}^{j_2j_3}_{i_2i_3}-\frac{1}{3}K^{j_2}_{i_2}K^{j_3}_{i_3}\right). \tag*{(21)}

The Chern density is a given polynomial of the extrinsic curvature, defined by

Kij=hi μhj ν∇μnν,(22)K_{ij}=h_i^{\ \mu}h_j^{\ \nu}\nabla_\mu n_\nu, \tag*{(22)}

where nμ=−Nδμzn_\mu=-N\delta^z_\mu is the outward-pointing unit normal to the boundary. The intrinsic Riemann tensor is computed from the boundary metric, Rklij=Rklij(h)\mathcal{R}^{ij}_{kl}=\mathcal{R}^{ij}_{kl}(h).

The Einstein-AdS action, renormalised with a topological term, can be equivalently written as

Itop+πℓ22Gχ(M4)=IEH+ℓ264πG∫∂M4d3x B3.(23)I_{\mathrm{top}}+\frac{\pi\ell^{2}}{2G}\chi(M_{4})=I_{\mathrm{EH}}+\frac{\ell^{2}}{64\pi G}\int_{\partial M_{4}}d^{3}x\,\mathcal{B}_{3}. \tag*{(23)}

Adding and subtracting the Euclidean Gibbons-Hawking term, the above action takes the form

Itop+πℓ22Gχ(M4)=IEH−18πG∫∂M4d3x−hK+∫∂M4d3x Lct.(24)I_{\mathrm{top}}+\frac{\pi\ell^{2}}{2G}\chi(M_{4})=I_{\mathrm{EH}}-\frac{1}{8\pi G}\int_{\partial M_{4}}d^{3}x\sqrt{-h}K+\int_{\partial M_{4}}d^{3}x\,\mathcal{L}_{\mathrm{ct}}. \tag*{(24)}

where Lct\mathcal{L}_{\mathrm{ct}} is the resummation of local counterterms in terms of the extrinsic curvature

Lct=ℓ216πG−hδj1j2j3i1i2i3Ki1j1(12Ri2i3j2j3−13Ki2j2Ki3j3+1ℓ2δi2j2δi3j3).(25)\mathcal{L}_{\mathrm{ct}}=\frac{\ell^{2}}{16\pi G}\sqrt{-h}\delta^{i_{1}i_{2}i_{3}}_{j_{1}j_{2}j_{3}}K^{j_{1}}_{i_{1}}\left(\frac{1}{2}\mathcal{R}^{j_{2}j_{3}}_{i_{2}i_{3}}-\frac{1}{3}K^{j_{2}}_{i_{2}}K^{j_{3}}_{i_{3}}+\frac{1}{\ell^{2}}\delta^{j_{2}}_{i_{2}}\delta^{j_{3}}_{i_{3}}\right). \tag*{(25)}

This becomes manifest once one considers the asymptotic expansion of KjiK^{i}_{j} as

Kji=1ℓδji+ℓSji(h)+O(z3),(26)K^{i}_{j}=\frac{1}{\ell}\delta^{i}_{j}+\ell S^{i}_{j}(h)+\mathcal{O}(z^{3}), \tag*{(26)}

where Sji(h)S^{i}_{j}(h) is the Schouten tensor of the boundary metric.

In point of fact, by expanding and further collecting the first two leading orders in the boundary curvature, this expression results in [12, 27]

Lct=h8πG(2ℓ+ℓ2R)+⋯ ,(27)\mathcal{L}_{\mathrm{ct}}=\frac{\sqrt{h}}{8\pi G}\left(\frac{2}{\ell}+\frac{\ell}{2}\mathcal{R}\right)+\cdots, \tag*{(27)}

which are the standard regulating counterterms in AdS4_4 gravity, such that the r.h.s. of Eq. (24) is the renormalised AdS action IrenI_{\mathrm{ren}} in the context of gauge/gravity duality.

Hence, on a smooth manifold Itop+πℓ22Gχ(M4)=IrenI_{\mathrm{top}}+\frac{\pi\ell^{2}}{2G}\chi(M_{4})=I_{\mathrm{ren}}, with

Iren=IEH−18πG∫∂M4d3xhK+18πG∫∂M4d3xh(2ℓ+ℓ2R).(28)I_{\mathrm{ren}}=I_{\mathrm{EH}}-\frac{1}{8\pi G}\int_{\partial M_{4}}d^{3}x\sqrt{h}K+\frac{1}{8\pi G}\int_{\partial M_{4}}d^{3}x\sqrt{h}\left(\frac{2}{\ell}+\frac{\ell}{2}\mathcal{R}\right). \tag*{(28)}

For an Einstein manifold, the on-shell bulk density is −(R+6/ℓ2)=6/ℓ2-(R+6/\ell^{2})=6/\ell^{2}. The above expression allows us to express the renormalised action in terms of the Renormalised Volume

Iren=38πGℓ2Vren(E)(29)I_{\mathrm{ren}}=\frac{3}{8\pi G\ell^{2}}V_{\mathrm{ren}}^{(E)} \tag*{(29)}
= Itop+πℓ22Gχ(M4).(30)=\ I_{\mathrm{top}}+\frac{\pi\ell^{2}}{2G}\chi(M_{4}). \tag*{(30)}

As it is clear from this expression, on a smooth filling, holographically and topologically renormalised actions differ by the Euler term. At fixed topology, ℓ\ell and GG, this gives a constant shift to Wald entropy. More generally, conical defects may extend through the bulk to the conformal boundary. Each action then receives contributions supported on the codimension-two surfaces, which must be separated from the regular contribution [29]. For the AdS C-metric, we will show how this decomposition relates the finite action mismatch to the renormalised areas of the string worldsheets and their Euler characteristics weighted by the conical deficits. We refer to the inclusion of such contributions as singular. These singular contributions vary with the string tensions, even when the topology remains fixed.

The AdS4_4 C-metric

The AdS C-metric belongs to the Plebański–Demiański family of solutions and describes a pair of uniformly accelerating black holes in an asymptotically AdS spacetime [30, 31, 32, 33, 34]. The acceleration is driven by conical deficits along the polar axis, which may be interpreted as cosmic strings or struts. We restrict our attention to the neutral, non-rotating solution, parametrised by the mass parameter mm, the acceleration AA, the AdS radius ℓ\ell, and the conical parameter K\mathcal{K}. In Schwarzschild-like coordinates, the line element is [15, 35]

ds2=1Ω2[−f(r)α2dt2+1f(r)dr2+r2(dθ2g(θ)+g(θ)sin⁡2θdϕ2K2)].(31)ds^2 = \frac{1}{\Omega^2}\left[-\frac{f(r)}{\alpha^2}dt^2 + \frac{1}{f(r)}dr^2 + r^2\left(\frac{d\theta^2}{g(\theta)} + g(\theta)\sin^2\theta\frac{d\phi^2}{\mathcal{K}^2}\right)\right]. \tag*{(31)}

where

Ω=1+Arcos⁡θ,f(r)=(1−A2r2)(1−2mr)+r2ℓ2,g(θ)=1+2mAcos⁡θ,α=1−A2ℓ2.(32)\begin{aligned} \Omega&= 1 + Ar\cos\theta,\\ f(r) &= (1-A^2r^2)\left(1-\frac{2m}{r}\right)+\frac{r^2}{\ell^2},\\ g(\theta) &= 1+2mA\cos\theta,\qquad\alpha=\sqrt{1-A^2\ell^2}. \tag*{(32)} \end{aligned}

The azimuthal coordinate has periodicity ϕ∼ϕ+2π\phi\sim\phi+2\pi. The parameters mm and AA determine the local geometry, while K\mathcal{K} fixes the global identification of the azimuthal coordinate, and hence, the conical deficits. The conformal factor Ω\Omega diverges at r=−1/Acos⁡θr=-1/A\cos\theta where the conformal boundary is located. The factor α\alpha in (31) fixes the normalisation of the timelike Killing vector at conformal infinity [15, 18] and can be obtained by mapping the m=0m=0 solution to global AdS.

In principle, there are two horizons depending on the interplay between mass, acceleration, and cosmological constant. We restrict ourselves to the slow acceleration approximation where Aℓ<1A\ell<1, in which the exterior region of the black hole contains no acceleration horizon. In this case, the black hole horizon equation determines the mass parameter as

m=r+2[1+r+2ℓ2(1−A2r+2)].(33)m=\frac{r_+}{2}\left[1+\frac{r_+^2}{\ell^2(1-A^2r_+^2)}\right]. \tag*{(33)}

After a Wick rotation, regularity at the Euclidean horizon fixes the inverse temperature,

β=4παf′(r+).(34)\beta=\frac{4\pi\alpha}{f'(r_+)}. \tag*{(34)}

The physical mass may be computed either from the Ashtekar-Magnon-Das formula or from the holographic stress tensor, see for instance [14, 21, 36, 37, 38]. With the asymptotic time normalization used in (31), one obtains

M=αmGK.(35)M=\frac{\alpha m}{\mathcal{G}\mathcal{K}}. \tag*{(35)}

The Schwarzschild-AdS result is recovered in the limit A→0A\to0 and K→1\mathcal{K}\to1. The parameter K\mathcal{K} controls the conical geometry of the polar axes. At the north and south poles,

θ+=0,θ−=π,(36)\theta_+=0,\qquad\theta_-=\pi, \tag*{(36)}

the angular function takes the values

g±:=g(θ±)=1±2mA.(37)g_\pm:=g(\theta_\pm)=1\pm2mA. \tag*{(37)}

The ratio between the proper circumference and proper radius near each pole is 2πg±/K2\pi g_\pm/\mathcal{K}. It is therefore useful to introduce the two cone-opening parameters

a±=g±K=1±2mAK.(38)a_\pm=\frac{g_\pm}{\mathcal{K}}=\frac{1\pm2mA}{\mathcal{K}}. \tag*{(38)}

The corresponding deficit angles and cosmic string tensions are

δ±=2π(1−a±),μ±=δ±8πG=14G(1−g(θ±)K)=14(1−1±2mAK).(39)\delta_{\pm}=2\pi(1-a_{\pm}),\qquad\mu_{\pm}=\frac{\delta_{\pm}}{8\pi G}=\frac{1}{4G}\left(1-\frac{g(\theta_{\pm})}{\mathcal{K}}\right)=\frac{1}{4}\left(1-\frac{1\pm2mA}{\mathcal{K}}\right). \tag*{(39)}

It is also useful to introduce the sum and difference of tensions

μ++μ−=12G(1−1K),μ+−μ−=mAGK.(40)\mu_{+}+\mu_{-}=\frac{1}{2G}\left(1-\frac{1}{\mathcal{K}}\right),\qquad\mu_{+}-\mu_{-}=\frac{mA}{G\mathcal{K}}. \tag*{(40)}

Thus, K\mathcal{K} fixes the total conical deficit, while mAmA measures the imbalance between the two tensions that produce the acceleration. One may remove either conical singularity by an appropriate choice of K\mathcal{K}. For example,

K=g+=1+2mA,(41)\mathcal{K}=g_{+}=1+2mA, \tag*{(41)}

making the north axis regular and leaving a cosmic string on the south axis with tension

μ−=mAGK.(42)\mu_{-}=\frac{mA}{G\mathcal{K}}. \tag*{(42)}

In the present work, we shall not impose this regularity condition. Both axes will instead be retained as codimension-two conical loci, with opening parameters a+a_{+} and a−a_{-}.

At fixed tt and rr, the angular part of the metric is topologically a two-sphere with conical singularities at its north and south poles. For generic values of a±a_{\pm}, it should therefore be regarded as a two-dimensional cone-manifold, or cone-sphere. For the special choice

a±=1n±,n±∈N,(43)a_{\pm}=\frac{1}{n_{\pm}},\qquad n_{\pm}\in\mathbb{N}, \tag*{(43)}

the two cone angles are orbifold angles, and the surface becomes the weighted projective space Σ=WCP[n−,n+]1\Sigma=\mathrm{WCP}^{1}_{[n_{-},n_{+}]} [20]. The generic C-metric considered here is not assumed to obey the quantisation condition (43); the spindle provides a useful orbifold specialisation to more general cone-manifolds. We will come back to this point later.

On-shell Euclidean Actions

We now review the two different prescriptions that render the Euclidean action finite. The first is the standard holographic renormalisation procedure based on intrinsic counterterms at the conformal boundary. The second is the topological prescription, in which the Einstein-AdS action is supplemented by the four-dimensional Gauss-Bonnet invariant.

Holographic renormalisation

For the C-metric (31), the conformal boundary is not located at constant rr. Therefore, it is convenient to perform a suitable change of coordinates to facilitate the evaluation of Euclidean action from the horizon to the conformal boundary.

We choose the transformation

x=cos⁡θ,1ρ=1r+Ax,such thatr=ρ1−Aρx.(44)x=\cos\theta,\qquad\frac{1}{\rho}=\frac{1}{r}+Ax,\qquad\text{such that}\quad r=\frac{\rho}{1-A\rho x}. \tag*{(44)}

Using the above, the line element (31) in Euclidean signature becomes

dsE2=dρ2H(ρ,x)+2Aρ2H(ρ,x)dρdx+H(ρ,x)α2dτ2+ρ2((A2ρ2H(ρ,x)+1P(x))dx2+P(x)K2dϕ2).(45)ds_{E}^{2}=\frac{d\rho^{2}}{H(\rho,x)}+\frac{2A\rho^{2}}{H(\rho,x)}d\rho dx+\frac{H(\rho,x)}{\alpha^{2}}d\tau^{2}+\rho^{2}\left(\left(\frac{A^{2}\rho^{2}}{H(\rho,x)}+\frac{1}{P(x)}\right)dx^{2}+\frac{P(x)}{\mathcal{K}^{2}}d\phi^{2}\right). \tag*{(45)}

where

H(ρ,x)=ρ2[(1ρ2−2Axρ−(1−x2)A2)(1−2m(1ρ−Ax))+1ℓ2](46)H(\rho,x)=\rho^2\left[\left(\frac{1}{\rho^2}-\frac{2Ax}{\rho}-(1-x^2)A^2\right)\left(1-2m\left(\frac{1}{\rho}-Ax\right)\right)+\frac{1}{\ell^2}\right] \tag*{(46)}

and

P(x)=(1−x2)(1+2mAx).(47)P(x)=(1-x^2)(1+2mAx). \tag*{(47)}

In these coordinates −1≤x≤1-1 \le x \le1 and ρh(x)≤ρ≤∞\rho_h(x) \le\rho\le\infty, where ρh(x)\rho_h(x) is evaluated at the horizon as

ρh(x)=r+1+Ar+x,(48)\rho_h(x)=\frac{r_+}{1+Ar_+x}, \tag*{(48)}

while 0≤τ≤β0 \le\tau\le\beta and 0≤ϕ≤2π0 \le\phi\le2\pi. These coordinates cover the complete axial disks. In particular, the north axis continues through 1/r=01/r=0 before reaching the conformal infinity, see for instance [39] or, in the case of three dimensions, [24].

Evaluating the Einstein action, the Gibbons-Hawking term, and the two intrinsic contributions of (28), we get [18, 20]

IrenE=β2αGK(m(1−2A2ℓ2)−r+3ℓ2(1−A2r+2)2).(49)I_{\mathrm{ren}}^{E}=\frac{\beta}{2\alpha G\mathcal{K}}\left(m(1-2A^2\ell^2)-\frac{r_+^3}{\ell^2(1-A^2r_+^2)^2}\right). \tag*{(49)}

Here β\beta is the period of the normalised Euclidean time (34). The standard volume and curvature counterterms suffice for this on-shell evaluation.

Topological renormalisation

The C-metric can be directly evaluated into the bulk action (15), without the need for the FG expansion for the metric, as it is a fully covariant expression. On a regular Einstein geometry, we have R=−12/ℓ2R=-12/\ell^2 and G=W2+24/ℓ2\mathcal{G}=W^2+24/\ell^2. For the C-metric, W2=48m2/ρ6W^2=48m^2/\rho^6, and g=ρ2/(αK)\sqrt{g}=\rho^2/(\alpha\mathcal{K}). Evaluating the integral in (16), we obtain

ItopE=−βm2ℓ2αKG(1r+3+A2r+).(50)I_{\mathrm{top}}^{E}=-\frac{\beta m^2\ell^2}{\alpha\mathcal{K}G}\left(\frac{1}{r_+^3}+\frac{A^2}{r_+}\right). \tag*{(50)}

Comparison of the two actions

Now, let us compare the results of both renormalisation prescriptions explicitly. Let us define

D+:=1−A2r+2.(51)D_+:=1-A^2r_+^2. \tag*{(51)}

The horizon equation and the radial derivative of the metric function at the horizon give

m=r+2(1+r+2ℓ2D+),f′(r+)=2m(1r+2+A2)+2r+(1ℓ2−A2).(52)m=\frac{r_+}{2}\left(1+\frac{r_+^2}{\ell^2D_+}\right),\qquad f'(r_+)=2m\left(\frac{1}{r_+^2}+A^2\right)+2r_+\left(\frac{1}{\ell^2}-A^2\right). \tag*{(52)}

Using the above expression and the regularity of the Euclidean time, we can see that the actions (15) and (28) differ by a constant

Itop−Iren=−βℓ2f′(r+)4GαK=−πℓ2GK.(53)I_{\mathrm{top}}-I_{\mathrm{ren}}=-\frac{\beta\ell^2f'(r_+)}{4G\alpha\mathcal{K}}=-\frac{\pi\ell^2}{G\mathcal{K}}. \tag*{(53)}

This is the mismatch emphasised in [14], where the remaining term is a finite Euler contribution. The conformally compactified Euclidean black hole exterior has the underlying topology

M4≃D2×S2,χ(M4)=2.(54)M_4\simeq D^2\times S^2,\qquad\chi(M_4)=2. \tag*{(54)}

It is useful to normalise the topological action as in the smooth Schwarzschild-AdS case, that is

IT:=Itop+πℓ22Gχ(M4)=Itop+πℓ2G.(55)I_T := I_{\mathrm{top}} + \frac{\pi\ell^2}{2G}\chi(M_4) = I_{\mathrm{top}} + \frac{\pi\ell^2}{G}. \tag*{(55)}

Then, the difference between the actions is

IT−Iren=πℓ2G(1−1K)=2πℓ2μ,μ:=μ++μ−.(56)I_T-I_{\mathrm{ren}}=\frac{\pi\ell^2}{G}\left(1-\frac{1}{\mathcal{K}}\right)=2\pi\ell^2\mu,\qquad\mu:=\mu_++\mu_-. \tag*{(56)}

It is easy to see that the distinction persists even when A→0A\to0 with K>1\mathcal{K}>1. Then, there is an equal tension at the two poles. The azimuthal integral of every regular curvature density is multiplied by 1/K1/\mathcal{K}, while the ordinary Euler characteristic remains two.

Renormalised Area

A conical defect separates a curvature integral into a regular part and a contribution supported on a codimension-two surface Σ\Sigma. The surface carries the induced metric γab\gamma_{ab}, with γ=det⁡(γab)\gamma=\det(\gamma_{ab}). We denote the opening fraction as aa, so that the cone angle is 2πa2\pi a and the deficit angle is 2π(1−a)2\pi(1-a). The corresponding physical tension is μ=(1−a)/4G\mu=(1-a)/4G, as discussed in § 3. To regulate the noncompact integrals, we introduce a common outer boundary at ρ=ρc\rho=\rho_c. Let M4,c(a)M_{4,c}^{(a)} denote the truncated conical geometry and set

Σc=Σ∩M4,c(a),Cc=Σ∩{ρ=ρc}=∂Σc.(57)\Sigma_c=\Sigma\cap M_{4,c}^{(a)},\qquad C_c=\Sigma\cap\{\rho=\rho_c\}=\partial\Sigma_c. \tag*{(57)}

For the axial disks below, the Euclidean time circle closes smoothly at the horizon at the Hawking period, so CcC_c is their only boundary. We write Mreg,cM_{\mathrm{reg},c} for the regular part of the same conical geometry, retaining its angular range. A smooth regulator of the cone is introduced at fixed ρc\rho_c, and the conical limit is taken before removing the outcutoff. At finite ρc\rho_c, the scalar curvature satisfies

∫M4,c(a)gR(a)=∫Mreg,cd4xgR+4π(1−a)A[Σc],(58)\int_{M_{4,c}^{(a)}}\sqrt{g}R^{(a)}=\int_{M_{\mathrm{reg},c}}d^4x\sqrt{g}R+4\pi(1-a)\mathcal{A}[\Sigma_c], \tag*{(58)}

where the singular contribution is given by the area functional of a codimension-two surface Σc\Sigma_c

A[Σc]=∫Σcd2yγ.(59)\mathcal{A}[\Sigma_c]=\int_{\Sigma_c}d^2y\sqrt{\gamma}. \tag*{(59)}

Let eaμ=∂Xμ/∂yae^\mu_a=\partial X^\mu/\partial y^a be tangent vectors to Σ\Sigma, so that γab=gμνeaμebν\gamma_{ab}=g_{\mu\nu}e^\mu_a e^\nu_b. For orthonormal normal vectors nAμn^\mu_A, satisfying gμνnAμnBν=δABg_{\mu\nu}n^\mu_A n^\nu_B=\delta_{AB}, and gμνnAμeaν=0g_{\mu\nu}n^\mu_A e^\nu_a=0. The extrinsic curvature and their traces are

kab(A)=eaμebν∇μnν(A),k(A)=γabkab(A).(60)k^{(A)}_{ab}=e^\mu_a e^\nu_b\nabla_\mu n^{(A)}_\nu,\qquad k^{(A)}=\gamma^{ab}k^{(A)}_{ab}. \tag*{(60)}

These quantities are evaluated in a smooth regularisation of the defect, with ∇\nabla the ambient Levi-Civita connection. Tangential indices {a,b}\{a,b\} are raised with γab\gamma^{ab}, while normal indices {A,B}\{A,B\} with δAB\delta^{AB}. We can evaluate the quadratic invariants up to linear terms in the deficit [29]

∫M4,c(a)d4xg(Rie(a))2=∫Mreg,cd4xg Rie2+8π(1−a)∫Σd2yγ(R  ABAB−kab(A)k(A)ab)+⋯(61)\int_{M_{4,c}^{(a)}}d^4x\sqrt{g}\left(\mathrm{Rie}^{(a)}\right)^2 = \int_{M_{\mathrm{reg},c}}d^4x\sqrt{g}\,\mathrm{Rie}^2 +8\pi(1-a)\int_{\Sigma}d^2y\sqrt{\gamma}\left(R^{AB}_{\ \ AB}-k^{(A)}_{ab}k_{(A)}^{ab}\right)+\cdots \tag*{(61)}
∫M4,c(a)d4xg(Ric(a))2=∫Mreg,cd4xg Ric2+4π(1−a)∫Σd2yγ(RAA−12k(A)k(A))+⋯\int_{M_{4,c}^{(a)}}d^4x\sqrt{g}\left(\mathrm{Ric}^{(a)}\right)^2 = \int_{M_{\mathrm{reg},c}}d^4x\sqrt{g}\,\mathrm{Ric}^2 +4\pi(1-a)\int_{\Sigma}d^2y\sqrt{\gamma}\left(R_{AA}-\frac{1}{2}k^{(A)}k_{(A)}\right)+\cdots
∫M4,c(a)d4xg(R(a))2=∫Mreg,cd4xg R2+8π(1−a)∫Σγ R+⋯\int_{M_{4,c}^{(a)}}d^4x\sqrt{g}\left(R^{(a)}\right)^2 = \int_{M_{\mathrm{reg},c}}d^4x\sqrt{g}\,R^2 +8\pi(1-a)\int_{\Sigma}\sqrt{\gamma}\,R+\cdots

The contractions RAA=RμνnAμnAνR_{AA}=R_{\mu\nu}n_A^\mu n_A^\nu and RABAB=RμνρσnAμnBνnAρnBσR_{ABAB}=R_{\mu\nu\rho\sigma}n_A^\mu n_B^\nu n_A^\rho n_B^\sigma use the regular ambient curvature. The dots denote regulator-dependent terms of order (1−a)2(1-a)^2 and higher. The Gauss relation is

R=R+2RAA−RABAB−k(A)k(A)+kab(A)k(A)ab.(62)R=\mathcal{R}+2R_{AA}-R_{ABAB}-k^{(A)}k_{(A)}+k_{ab}^{(A)}k_{(A)}^{ab}. \tag*{(62)}

where we denoted intrinsic scalar curvature of γab\gamma_{ab} by R\mathcal{R}. Consequently, their Gauss-Bonnet combination has the intrinsic defect term

∫M4,cad4xg G(a)=∫M4,cregd4xg G+8π(1−a)∫Σd2yγR.(63)\int_{M^a_{4,c}}d^4x\sqrt{g}\,\mathcal{G}^{(a)} =\int_{M^{\mathrm{reg}}_{4,c}}d^4x\sqrt{g}\,\mathcal{G} +8\pi(1-a)\int_{\Sigma}d^2y\sqrt{\gamma}\mathcal{R}. \tag*{(63)}

We now insert these relations into the action. The deficit fraction combines with Newton’s constant into the physical tension μ\mu, and combining (58) with (63), the topological action, at finite cutoff, gives

Itop[M4,ca]=Itopreg−μ(A[Σc]+ℓ22∫ΣcγR).(64)I_{\mathrm{top}}[M^a_{4,c}]=I_{\mathrm{top}}^{\mathrm{reg}}-\mu\left(\mathcal{A}[\Sigma_c]+\frac{\ell^2}{2}\int_{\Sigma_c}\sqrt{\gamma}\mathcal{R}\right). \tag*{(64)}

Here Itopreg(ρc)I_{\mathrm{top}}^{\mathrm{reg}}(\rho_c) is the action evaluated on the regular curvature densities at the cutoff surface. The Einstein contribution is −μA[Σc]-\mu\mathcal{A}[\Sigma_c], while the Gauss-Bonnet coupling supplies the intrinsic curvature term. To identify the finite area in (64), we apply the two-dimensional Gauss-Bonnet theorem to the compact surface Σc\Sigma_c, that is

∫ΣcγR+2∫Cckg dl=4πχ(Σc)=4πχ(Σ).(65)\int_{\Sigma_c}\sqrt{\gamma}\mathcal{R}+2\int_{C_c}k_g\,dl=4\pi\chi(\Sigma_c)=4\pi\chi(\Sigma). \tag*{(65)}

The boundary geodesic curvature kgk_g is defined with the sign that is positive on a flat disk. Multiplying by ℓ2\ell^2, adding A[Σc]\mathcal{A}[\Sigma_c] on both sides, and rearranging the identity leads to

A[Σc]+ℓ22∫ΣcγR=A[Σc]−ℓ2∫Cckg dl+2πℓ2χ(Σ).(66)\mathcal{A}[\Sigma_c]+\frac{\ell^2}{2}\int_{\Sigma_c}\sqrt{\gamma}\mathcal{R} =\mathcal{A}[\Sigma_c]-\ell^2\int_{C_c}k_g\,dl+2\pi\ell^2\chi(\Sigma). \tag*{(66)}

The cutoff is removed by taking ρc→∞\rho_c\to\infty, so that CcC_c approaches the conformal boundary and Σc\Sigma_c exhaust the complete worldsheet. The area and intrinsic curvature integrals diverge separately, but their combination has the finite limit

lim⁡ρc→∞(A[Σc]−ℓ2∫Cckg dl)=Aren[Σ]+2πℓ2χ(Σ).(67)\lim_{\rho_c\to\infty}\left(\mathcal{A}[\Sigma_c]-\ell^2\int_{C_c}k_g\,dl\right)=\mathcal{A}_{\mathrm{ren}}[\Sigma]+2\pi\ell^2\chi(\Sigma). \tag*{(67)}

Subtracting this finite Euler term gives the renormalised area formula [40, 41, 42]

Aren[Σ]=ℓ24∫Σd2y γcdab(Rabcd[γ]+1ℓ2δabcd)−2πℓ2χ(Σ).(68)\mathcal{A}_{\mathrm{ren}}[\Sigma]=\frac{\ell^2}{4}\int_{\Sigma}d^2y\,\gamma^{ab}_{cd}\left(\mathcal{R}^{cd}_{ab}[\gamma]+\frac{1}{\ell^2}\delta^{cd}_{ab}\right)-2\pi\ell^2\chi(\Sigma). \tag*{(68)}

This is the codimension-two counterpart of the curvature combination that renormalises the bulk volume. The cancellation removes the area divergence, while the Euler term fixes the finite normalisation. Thus, after removing the cutoff, the defect action (64) becomes

Itop[M4(μ)]=Itopreg−μ(Aren[Σ]+2πℓ2χ(Σ)).(69)I_{\mathrm{top}}[M_4^{(\mu)}]=I_{\mathrm{top}}^{\mathrm{reg}}-\mu\left(\mathcal{A}_{\mathrm{ren}}[\Sigma]+2\pi\ell^2\chi(\Sigma)\right). \tag*{(69)}

The above is valid for any manifold with conical deficits.

Renormalised worldsheet areas in the AdS C-metric

We now compute (68) for the C-metric under consideration. First we separate the two poles of the C-metric, writing s=+1s=+1 at the north axis and s=−1s=-1 at the south axis. Then, the worldsheet Σs\Sigma_s has induced metric γijs\gamma^s_{ij} which corresponds to the ρ−τ\rho-\tau plane,

dsΣs2=Hs(ρ)α2dτ2+dρ2Hs(ρ)(70)\mathrm{d}s^2_{\Sigma_s}=\frac{H_s(\rho)}{\alpha^2}\mathrm{d}\tau^2+\frac{\mathrm{d}\rho^2}{H_s(\rho)} \tag*{(70)}

with

Hs(ρ)=ρ2ℓ2−2As(1+2mAs)ρ+1+6smA−2mρ.(71)H_s(\rho)=\frac{\rho^2}{\ell^2}-2As(1+2mAs)\rho+1+6smA-\frac{2m}{\rho}. \tag*{(71)}

The induced metric has a horizon at

ρh,s=r+1+sAr+(72)\rho_{h,s}=\frac{r_+}{1+sAr_+} \tag*{(72)}

and the truncated disk is Σs,c={ρh,s≤ρ≤ρc,0≤τ≤β}\Sigma_{s,c}=\{\rho_{h,s}\leq\rho\leq\rho_c,0\leq\tau\leq\beta\}, with boundary Cs,cC_{s,c} at ρc\rho_c. Since γ=1/α\sqrt{\gamma}=1/\alpha, its regulated area and intrinsic curvature are

A[Σs]=βα(ρc−ρh,s),Rs=−Hs′′=−2ℓ2+4mρ3(73)\mathcal{A}[\Sigma_s]=\frac{\beta}{\alpha}(\rho_c-\rho_{h,s}),\qquad\mathcal{R}_s=-H_s^{\prime\prime}=-\frac{2}{\ell^2}+\frac{4m}{\rho^3} \tag*{(73)}

The above expression shows that at the conformal boundary ρc→∞\rho_c\to\infty the first term in (73) diverges. This divergence is precisely corrected by the singular contribution of the Gauss-Bonnet term that enters in the definition of the renormalised area. Removing the cutoff and evaluating the renormalised area formula gives

Aren[Σs]=βmℓ2αr+2(1+sAr+)2−2πℓ2χ(Σs).(74)\mathcal{A}_{\mathrm{ren}}[\Sigma_s]=\frac{\beta m\ell^2}{\alpha r_+^2}(1+sAr_+)^2-2\pi\ell^2\chi(\Sigma_s). \tag*{(74)}

Since each worldsheet is a disk, χ(Σs)=1\chi(\Sigma_s)=1. Therefore, the Euler part of each renormalised area is

[Aren(Σ)]Euler=−2πℓ2.(75)[\mathcal{A}_{\mathrm{ren}}(\Sigma)]_{\mathrm{Euler}}=-2\pi\ell^2. \tag*{(75)}

Multiplying by the corresponding tensions and summing over both axes gives

∑s=±μs[Aren(Σs)]Euler=−2πℓ2(μ++μ−)=−2πℓ2μ.(76)\sum_{s=\pm}\mu_s[\mathcal{A}_{\mathrm{ren}}(\Sigma_s)]_{\mathrm{Euler}}=-2\pi\ell^2(\mu_++\mu_-)=-2\pi\ell^2\mu. \tag*{(76)}

Consequently, the finite action mismatch (56) satisfies

IT−Iren=−∑s=±μs[Aren(Σs)]Euler.(77)I_T-I_{\mathrm{ren}}=-\sum_{s=\pm}\mu_s[\mathcal{A}_{\mathrm{ren}}(\Sigma_s)]_{\mathrm{Euler}}. \tag*{(77)}

The same Euler contribution can be identified directly at the horizon. Let H\mathcal{H} denote its compact two-dimensional cross-section, with induced metric σij\sigma_{ij} and scalar curvature RH\mathcal{R}_{\mathcal{H}}. Each string worldsheet meets H\mathcal{H} at a conical point with deficit fraction δs/2π=1−as=4Gμs\delta_s/2\pi=1-a_s=4G\mu_s. The two-dimensional Gauss-Bonnet theorem gives

∫Hregd2xσRH+4π∑s=±δs2π=4πχ(H)=8π,(78)\int_{\mathcal{H}_{\mathrm{reg}}}\mathrm{d}^2x\sqrt{\sigma}\mathcal{R}_{\mathcal{H}}+4\pi\sum_{s=\pm}\frac{\delta_s}{2\pi}=4\pi\chi(\mathcal{H})=8\pi, \tag*{(78)}

where Hreg\mathcal{H}_{\mathrm{reg}} is the horizon with the two conical points excluded. Using a++a−=2/Ka_++a_-=2/\mathcal{K}, the integral over Hreg\mathcal{H}_{\mathrm{reg}} gives 8π/K8\pi/\mathcal{K}, and the two conical points contribute 8π(1−K−1)8\pi(1-\mathcal{K}^{-1}). Their sum is 8π as required by χ(H)=2\chi(\mathcal{H}) = 2. Using the Gauss-Bonnet coupling ℓ2/4\ell^2/4, the entropy contribution is [29]

SGB=ℓ28G∫Hd2xσRH=SGBreg+SGBtips(79)S_{GB} = \frac{\ell^2}{8G}\int_{\mathcal{H}} d^2x\sqrt{\sigma}\mathcal{R}_{\mathcal{H}} = S_{GB}^{\mathrm{reg}} + S_{GB}^{\mathrm{tips}} \tag*{(79)}

where the integral over H\mathcal{H} includes the curvature at the poles in the limit of a smooth regularisation. The regular and tip contributions are

SGBreg=πℓ2GK,SGBtips=2πℓ2μ.(80)S_{GB}^{\mathrm{reg}} = \frac{\pi\ell^2}{G\mathcal{K}}, \qquad S_{GB}^{\mathrm{tips}} = 2\pi\ell^2\mu. \tag*{(80)}

The tip contribution equals the residual mismatch between the regular action evaluations after Euler normalisation. In terms of the renormalised worldsheet areas, this gives

SGBtips=IT−Iren=−∑s=±μs[Aren(Σs)]Euler(81)S_{GB}^{\mathrm{tips}} = I_T-I_{\mathrm{ren}} = -\sum_{s=\pm} \mu_s[\mathcal{A}_{\mathrm{ren}}(\Sigma_s)]_{\mathrm{Euler}} \tag*{(81)}

Singular thermodynamics: conical deficits

The thermodynamics of accelerating black AdS black holes with fixed conical deficits was studied in [15] and subsequently extended to include varying deficits in both accelerating and static backgrounds [16]. When the string tensions are allowed to vary, the first law acquires additional work terms whose conjugate quantities are known as thermodynamic lengths. In this framework, we give a geometric interpretation of the relative thermodynamic length in terms of the difference of the renormalised worldsheet areas per unit Euclidean time. At fixed total string tension, this difference determines the work associated with redistributing tension between the axes. The Euclidean renormalised on-shell action satisfies the quantum statistical relation

Iren=βM−S,S=AH4G=πr+2GKD+.(82)I_{\mathrm{ren}} = \beta M-S,\qquad S=\frac{A_{\mathcal{H}}}{4G}=\frac{\pi r_+^2}{G\mathcal{K}D_+}. \tag*{(82)}

At fixed GG, variations within the slowly accelerating family satisfy [16]

dM=TdS+VdP−∑s=±1λsdμs,P=38πGℓ2,T=β−1,(83)dM = T dS + V dP-\sum_{s=\pm1}\lambda_s d\mu_s,\qquad P=\frac{3}{8\pi G\ell^2},\qquad T=\beta^{-1}, \tag*{(83)}

where

λs=1α(r+1+sAr+−m−sAℓ2),V=4π3αK(r+3D+2+mA2ℓ4).(84)\lambda_s=\frac{1}{\alpha}\left(\frac{r_+}{1+sAr_+}-m-sA\ell^2\right),\qquad V=\frac{4\pi}{3\alpha\mathcal{K}}\left(\frac{r_+^3}{D_+^2}+mA^2\ell^4\right). \tag*{(84)}

Here, VV is the thermodynamic three-volume, which must not be confused with Vren(E)V_{\mathrm{ren}}^{(E)}, and λs\lambda_s are the corresponding thermodynamic lengths. For the Euler normalised topological action, IT=Iren+2πℓ2μI_T=I_{\mathrm{ren}}+2\pi\ell^2\mu. At fixed tensions and pressure, the black hole entropy from the topological action gives

ST=(β∂β−1)IT=AH4G−2πℓ2μ.(85)S_T=(\beta\partial_\beta-1)I_T=\frac{A_{\mathcal{H}}}{4G}-2\pi\ell^2\mu. \tag*{(85)}

This shift is constant on each fixed-tension. When the total tension varies, the shift also must be varied. At fixed pressure, its contribution shifts the conjugate lengths to λs(T)=λs−2πℓ2β−1\lambda_s^{(T)}=\lambda_s-2\pi\ell^2\beta^{-1}.

The complete renormalised worldsheet areas contain further geometric information. The renormalised area for the C-metric (74) with (84) gives

−Aren[Σs]β=λs+mα(1−2A2ℓ2).(86)-\frac{\mathcal{A}_{\mathrm{ren}}[\Sigma_s]}{\beta}=\lambda_s+\frac{m}{\alpha}(1-2A^2\ell^2). \tag*{(86)}

Multiplying by the total tension on both sides and taking its variation gives

−∑s=±1λsdμs=1β∑s=±1Aren[Σs]dμs+mα(1−2A2ℓ2)dμ.(87)-\sum_{s=\pm1}\lambda_s d\mu_s = \frac{1}{\beta}\sum_{s=\pm1}\mathcal{A}_{\mathrm{ren}}[\Sigma_s]d\mu_s + \frac{m}{\alpha}(1-2A^2\ell^2)d\mu. \tag*{(87)}

For slowly accelerating black holes, we can redistribute the tension between the two axes while keeping the total tension fixed, that is

dμ=dμ++dμ−=0⇒dμ−=−dμ+.(88)d\mu= d\mu_{+} + d\mu_{-} = 0 \quad\Rightarrow\quad d\mu_{-} = -d\mu_{+} . \tag*{(88)}

With this, the total tension term at the right-hand side of (87) vanishes. Substituting back into the first law, we get

dM=T4GdAH+VdP+Aren[Σ+]−Aren[Σ−]βdμ+.(89)dM = \frac{T}{4G}dA_{\mathrm{H}} + VdP + \frac{\mathcal{A}_{\mathrm{ren}}[\Sigma_{+}] - \mathcal{A}_{\mathrm{ren}}[\Sigma_{-}]}{\beta}d\mu_{+} . \tag*{(89)}

The difference between the renormalised areas of the string worldsheets determines the relative tension work exactly. The individual thermodynamic lengths still differ from the value of the renormalised area. At fixed pressure, the volume term drops out, and the Euler mismatch 2πℓ2μ2\pi\ell^2\mu remains constant even as tension is transferred between the axes.

Conclusions

In this work, we have investigated the relation between holographic and topological renormalisation of four-dimensional AdS gravity in the presence of codimension-two conical defects. Using the slowly accelerating AdS C-metric, we have shown that the finite mismatch between the two renormalised Euclidean actions, after the standard Euler normalisation, is entirely accounted for by the tension-weighted Euler contributions of the renormalised worldsheet areas (81). These contributions coincide with the Gauss–Bonnet entropy supported at the conical points of the horizon.

In this sense, singular thermodynamics retains the finite contributions supported on the conical defects that are omitted from evaluations restricted to the regular or smooth fillings. The defect Euler contribution is constant at fixed pressure and total string tension, but varies across the broader thermodynamic family.

We have also obtained explicit expressions for the renormalised areas of the cosmic string worldsheets. At fixed total string tension, their difference per unit Euclidean time determines the thermodynamic work associated with redistributing tension between the two axes (89). This provides a geometric interpretation of the difference of thermodynamic lengths appearing in the first law of accelerating black hole thermodynamics.

More generally, our results illustrate how topological renormalisation naturally incorporates singular contributions that are not manifest in the standard holographic counterterm prescription.

In a more physical context, Renormalised Volume is linked to a renormalised version of Einstein gravity action with negative cosmological constant in even dimensions, which is rendered finite by the addition of spacetime covariant terms. Indeed, a partial renormalisation is achieved with the addition of the Euler (topological) term to the Einstein-AdS action, which, for a metric g(0)g_{(0)} that is conformally flat, is equivalent to the counterterm method [6, 11].

However, in even dimensions equal to or higher than six, the difference between Topological and Holographic Renormalisation for a generic conformal data g(0)g_{(0)} starts with a Weyl-squared term at the boundary, plus higher-derivative terms constructed with conformal tensors of the boundary. Conformal Renormalisation is a way of eliminating the above drawback by the conformal completion of the Einstein-AdS action plus a topological term. This means that the regulated Einstein action in even dimensions is the second-derivative sector of a higher-derivative gravity which is conformally invariant. A nontrivial example of renormalisation of the Einstein-AdS action was given in six dimensions in [43]. Indeed, Holographic Renormalisation appears as encoded in a precise combination of conformal invariants in the bulk. Conformal Renormalisation in eight dimensions relies on the classification of conformal invariants made in [44]. In addition, the existence of a generic Einstein sector in 8D Conformal Gravity was proven for a fine-tuned combination of these invariants [45]. In [46], the exact agreement between the Einstein sector of this particular 8D Conformal Gravity and Holographic Renormalisation was found. This argument provides further support to the idea of a deep connection between conformal invariance and finiteness of the AdS gravity action. In particular, as there is always a topological term involved in the Renormalised Volume for Einstein spaces, it is expected that one could readily capture codimension-two defects of gravitational solutions (and singular thermodynamic contributions) in this way.

Note added. While this work was being completed, a manuscript with conceptual overlap regarding the interpretation of finite topological contributions to the Euclidean action and black hole entropy appeared on arXiv [47].

Acknowledgments

We thank Felipe Diaz for his collaboration during the initial stage of this work. We also thank Alexey Koshelev and Anna Tokareva for their valuable comments and discussions on related topics. The work of GAH was supported by the National Natural Science Foundation of China (NSFC) grant No. 12505091. This work was supported by the Agencia Nacional de Investigación y Desarrollo (ANID) through FONDECYT Regular grants No. 1230492, 1231779, 1240043, 1240955, 1250672, and 1261016.

References

  1. [1]C. R. Graham, Volume and area renormalizations for conformally compact Einstein metrics, Rend. Circ. Mat. Palermo S 63 (2000) 31–42, [math/9909042].
  2. [2]M. Anderson, L² curvature and volume renormalization of AHE metrics on 4-manifolds, Mathematical Research Letters 8 (2000) 171–188.
  3. [3]A. Chang, J. Qing, and P. Yang, On the renormalized volumes for conformally compact Einstein manifolds, math/0512376.
  4. [4]C. Fefferman and C. R. Graham, Conformal invariants, in Élie Cartan et les mathématiques d’aujourd’hui - Lyon, 25-29 juin 1984, no. S131 in Astérisque, pp. 95–116. Société mathématique de France, 1985.
  5. [5]J. Maldacena, Einstein Gravity from Conformal Gravity, arXiv:1105.5632.
  6. [6]G. Anastasiou, O. Miskovic, R. Olea, and I. Papadimitriou, Counterterms, Kounterterms, and the variational problem in AdS gravity, JHEP 08 (2020) 061, [arXiv:2003.06425].DOI
  7. [7]S. W. MacDowell and F. Mansouri, Unified Geometric Theory of Gravity and Supergravity, Phys. Rev. Lett. 38 (1977) 739. [Erratum: Phys.Rev.Lett. 38, 1376 (1977)].
  8. [8]K. S. Stelle and P. C. West, Spontaneously Broken De Sitter Symmetry and the Gravitational Holonomy Group, Phys. Rev. D 21 (1980) 1466.DOI
  9. [9]R. Aros, M. Contreras, R. Olea, R. Troncoso, and J. Zanelli, Conserved charges for gravity with locally AdS asymptotics, Phys. Rev. Lett. 84 (2000) 1647–1650, [gr-qc/9909015].DOI
  10. [10]R. Olea, Mass, angular momentum and thermodynamics in four-dimensional Kerr-AdS black holes, JHEP 06 (2005) 023, [hep-th/0504233].DOI
  11. [11]O. Miskovic and R. Olea, Topological regularization and self-duality in four-dimensional anti-de Sitter gravity, Phys. Rev. D 79 (2009) 124020, [arXiv:0902.2082].DOI
  12. [12]V. Balasubramanian and P. Kraus, A Stress tensor for Anti-de Sitter gravity, Commun. Math. Phys. 208 (1999) 413–428, [hep-th/9902121].DOI
  13. [13]S. de Haro, S. N. Solodukhin, and K. Skenderis, Holographic reconstruction of space-time and renormalization in the AdS / CFT correspondence, Commun. Math. Phys. 217 (2001) 595–622, [hep-th/0002230].
  14. [14]T. Hale, D. Kubizňák, J. Menšíková, R. B. Mann, and J. Yang, Thermodynamics of charged and accelerating black holes, Phys. Rev. D 111 (2025), no. 10 104004, [arXiv:2501.13679].DOI
  15. [15]M. Appels, R. Gregory, and D. Kubiznak, Thermodynamics of Accelerating Black Holes, Phys. Rev. Lett. 117 (2016), no. 13 131303, [arXiv:1604.08812].DOI
  16. [16]M. Appels, R. Gregory, and D. Kubiznak, Black Hole Thermodynamics with Conical Defects, JHEP 05 (2017) 116, [arXiv:1702.00490].DOI
  17. [17]A. Ball and N. Miller, Accelerating black hole thermodynamics with boost time, Class. Quant. Grav. 38 (2021), no. 14 145031, [arXiv:2008.03682].DOI
  18. [18]A. Anabalón, M. Appels, R. Gregory, D. Kubizňák, R. B. Mann, and A. Ovgun, Holographic Thermodynamics of Accelerating Black Holes, Phys. Rev. D 98 (2018), no. 10 104038, [arXiv:1805.02687].DOI
  19. [19]A. Anabalón, F. Gray, R. Gregory, D. Kubizňák, and R. B. Mann, Thermodynamics of Charged, Rotating, and Accelerating Black Holes, JHEP 04 (2019) 096, [arXiv:1811.04936].DOI
  20. [20]D. Cassani, J. P. Gauntlett, D. Martelli, and J. Sparks, Thermodynamics of accelerating and supersymmetric AdS₄ black holes, Phys. Rev. D 104 (2021), no. 8 086005, [arXiv:2106.05571].DOI
  21. [21]H. Kim, N. Kim, Y. Lee, and A. Poole, Thermodynamics of accelerating AdS₄ black holes from the covariant phase space, Eur. Phys. J. C 83 (2023), no. 12 1095, [arXiv:2306.16187].DOI
  22. [22]M. Astorino, Accelerating black hole in 2+1 dimensions and 3+1 black (st)ring, JHEP 01 (2011) 114, [arXiv:1101.2616].DOI
  23. [23]G. Arenas-Henriquez, R. Gregory, and A. Scoins, On Acceleration in Three Dimensions, arXiv:2202.08823.
  24. [24]G. Arenas-Henriquez, A. Cisterna, F. Diaz, and R. Gregory, Accelerating Black Holes in 2 + 1 dimensions: Holography revisited, JHEP 09 (2023) 122, [arXiv:2308.00613].
  25. [25]A. Cisterna, F. Diaz, R. B. Mann, and J. Oliva, Exploring accelerating hairy black holes in 2+1 dimensions: the asymptotically locally anti-de Sitter class and its holography, JHEP 11 (2023) 073, [arXiv:2309.05559].DOI
  26. [26]G. Arenas-Henriquez, F. Diaz, and D. Rivera-Betancour, Generalized Fefferman-Graham gauge and boundary Weyl structures, JHEP 02 (2025) 007, [arXiv:2411.12513].
  27. [27]R. Emparan, C. V. Johnson, and R. C. Myers, Surface terms as counterterms in the AdS / CFT correspondence, Phys. Rev. D 60 (1999) 104001, [hep-th/9903238].DOI
  28. [28]I. Papadimitriou and K. Skenderis, Thermodynamics of asymptotically locally AdS spacetimes, JHEP 08 (2005) 004, [hep-th/0505190].
  29. [29]D. V. Fursaev, A. Patrushev, and S. N. Solodukhin, Distributional Geometry of Squashed Cones, Phys. Rev. D 88 (2013), no. 4 044054, [arXiv:1306.4000].DOI
  30. [30]J. F. Plebanski and M. Demianski, Rotating, charged, and uniformly accelerating mass in general relativity, Annals Phys. 98 (1976) 98–127.DOI
  31. [31]O. J. Dias and J. P. Lemos, Pair of accelerated black holes in anti-de Sitter background: AdS C metric, Phys. Rev. D 67 (2003) 064001, [hep-th/0210065].
  32. [32]J. Podolsky and J. B. Griffiths, Accelerating Kerr-Newman black holes in (anti-)de Sitter space-time, Phys. Rev. D 73 (2006) 044018, [gr-qc/0601130].DOI
  33. [33]J. Griffiths and J. Podolsky, A New look at the Plebanski-Demianski family of solutions, Int. J. Mod. Phys. D 15 (2006) 335–370, [gr-qc/0511091].
  34. [34]P. Krtous, Accelerated black holes in an anti-de Sitter universe, Phys. Rev. D 72 (2005) 124019, [gr-qc/0510101].DOI
  35. [35]K. Hong and E. Teo, A New form of the C metric, Class. Quant. Grav. 20 (2003) 3269–3277, [gr-qc/0305089].DOI
  36. [36]A. Ashtekar and A. Magnon, Asymptotically anti-de sitter space-times, Classical and Quantum Gravity 1 (jul, 1984) L39–L44.
  37. [37]A. Ashtekar and S. Das, Asymptotically Anti-de Sitter space-times: Conserved quantities, Class. Quant. Grav. 17 (2000) L17–L30, [hep-th/9911230].
  38. [38]D. P. Jatkar, G. Kofinas, O. Miskovic, and R. Olea, Conformal Mass in AdS gravity, Phys. Rev. D 89 (2014), no. 12 124010, [arXiv:1404.1411].DOI
  39. [39]V. E. Hubeny, D. Marolf, and M. Rangamani, Black funnels and droplets from the AdS C-metrics, Class. Quant. Grav. 27 (2010) 025001, [arXiv:0909.0005].DOI
  40. [40]S. Alexakis and R. Mazzeo, Renormalized area and properly embedded minimal surfaces in hyperbolic 3-manifolds, Commun. Math. Phys. 297 (2010) 621–651.arxiv.org/abs/0802.2250
  41. [41]G. Anastasiou, I. J. Araya, C. Arias, and R. Olea, Einstein-AdS action, renormalized volume/area and holographic Rényi entropies, JHEP 08 (2018) 136, [arXiv:1806.10708].DOI
  42. [42]G. Anastasiou, I. J. Araya, and R. Olea, Energy functionals from Conformal Gravity, JHEP 10 (2022) 123, [arXiv:2209.02006].DOI
  43. [43]G. Anastasiou, I. J. Araya, and R. Olea, Einstein Gravity from Conformal Gravity in 6D, JHEP 01 (2021) 134, [arXiv:2010.15146].DOI
  44. [44]N. Boulanger and J. Erdmenger, A Classification of local Weyl invariants in D = 8, Class. Quant. Grav. 21 (2004), no. 18 4305–4316, [hep-th/0405228]. [Erratum: Class.Quant.Grav. 39, 039501 (2022)].
  45. [45]N. Boulanger and D. Rovere, 8D conformal gravity with Einstein sector, and its relation to the Q-curvature, JHEP 02 (2026) 101, [arXiv:2511.01368].arxiv.org/abs/2511.01368
  46. [46]G. Anastasiou, I. J. Araya, N. Boulanger, R. Olea, and D. Rovere, Conformal Renormalisation of 8D Einstein Gravity, arXiv:2607.03679.arxiv.org/abs/2607.03679
  47. [47]M. Liška and D. Kubiznak, On Thermodynamic Role of the Gauss-Bonnet Topological Term, arXiv:2611.12698.arxiv.org/abs/2611.12698

Paper details

vlnc:W_rzaf8yxa
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 0

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Physics and Astronomy

Not recorded for this paper

View statement
Looking for related papers…
Contents