Lovelock-type Brane Gravity

Updated 4 July 2026

Lovelock-type Brane Gravity (LBG) is a geometric framework that uses antisymmetrized extrinsic curvature invariants to yield second-order field equations.
It extends the Dirac–Nambu–Goto paradigm by introducing a finite series of Lovelock-type invariants, ensuring higher-curvature terms avoid higher derivatives in the dynamics.
LBG connects with various formulations like Born–Infeld actions, holographic decompositions, and FRW brane cosmologies, offering insights into braneworld and higher-curvature gravity.

Searching arXiv for Lovelock-type brane gravity and closely related papers. Lovelock-type Brane Gravity (LBG) is a geometric theory of relativistic extended objects in which the dynamical variables are the embedding functions of a brane worldvolume and the action is built from a finite tower of special antisymmetrized scalars constructed from the extrinsic curvature. In the codimension-one formulation most commonly associated with the term, a $p$ -brane sweeps a $(p+1)$ -dimensional timelike worldvolume embedded in a flat $(p+2)$ -dimensional Minkowski spacetime, and the Lovelock-type brane invariants are arranged so that the equations of motion remain second order in the embedding variables despite the presence of higher-curvature structures (Cruz et al., 2012). In this sense, LBG is a higher-dimensional generalization of Dirac–Nambu–Goto (DNG) theory and includes, as special sectors or closely related limits, the Regge–Teitelboim or geodetic brane gravity model, Born–Infeld-type brane actions, holographic bulk–surface decompositions, and FRW brane cosmologies with GHY- and GHYM-type terms (Rojas et al., 2024).

1. Geometric setting and kinematics

The basic configuration is a $p$ -dimensional extended object $\Sigma$ whose history is a $(p+1)$ -dimensional timelike worldvolume $m$ , embedded in flat Minkowski spacetime $\mathcal M$ of dimension $N=p+2$ . The embedding is described by

$y^\mu=X^\mu(x^a), \qquad a=0,1,\dots,p,$

with tangent vectors

$(p+1)$ 0

and induced metric

$(p+1)$ 1

For codimension one there is a single unit spacelike normal $(p+1)$ 2, and the extrinsic curvature is

$(p+1)$ 3

The same geometry may also be described through the first, second, and third fundamental forms,

$(p+1)$ 4

with

$(p+1)$ 5

by the Gauss–Weingarten equations (Bagatella-Flores et al., 2016).

Because the ambient spacetime is flat, the Gauss–Codazzi relations express intrinsic worldvolume curvature in terms of $(p+1)$ 6. In particular,

$(p+1)$ 7

This permits an intrinsic rewriting of several apparently extrinsic invariants and is the mechanism behind the parallel between LBG and ordinary Lovelock gravity (Cruz et al., 2012).

The DNG model is recovered as the lowest-order sector. Its action,

$(p+1)$ 8

yields the minimal-surface condition

$(p+1)$ 9

LBG preserves the same embedding-based kinematics but replaces the single volume term by a finite sequence of higher geometric scalars. This suggests a direct generalization of minimal-surface dynamics rather than an arbitrary higher-derivative deformation (Rojas et al., 2024).

2. Lovelock brane invariants and action functionals

The standard LBG action is written as

$(p+2)$ 0

with

$(p+2)$ 1

Here $(p+2)$ 2 is the generalized Kronecker delta. The $(p+2)$ 3 are the Lovelock-type brane invariants, they vanish for $(p+2)$ 4, and the $(p+2)$ 5 term is topological and does not affect the field equations (Bagatella-Flores et al., 2016).

The first few invariants make the analogy with Lovelock gravity explicit: $(p+2)$ 6

$(p+2)$ 7

$(p+2)$ 8

and

$(p+2)$ 9

Even orders are Gauss–Bonnet-type invariants on the brane, odd orders are Gibbons–Hawking–York–Myers-like boundary terms, $p$ 0 is the DNG action, and $p$ 1 yields the Regge–Teitelboim model (Rojas, 2019).

A distinctive feature of the brane theory is its counting. In ordinary Lovelock gravity the metric is the field variable and the number of nontrivial densities is limited by the bulk dimension. In LBG the embedding functions are the fundamental variables, and the independent brane invariants are tied to the worldvolume dimension, producing a finite tower $p$ 2 (Cruz et al., 2012).

A geometrically economical reformulation arises from a parallel-surface construction. If one shifts the original worldvolume by a constant proper distance $p$ 3 along the normal,

$p$ 4

then

$p$ 5

and the induced metric on the displaced worldvolume becomes

$p$ 6

Its determinant factorizes as

$p$ 7

Writing the DNG action for the displaced worldvolume therefore gives

$p$ 8

where

$p$ 9

In this formulation LBG emerges from the volume of a parallel brane, and the DNG theory is recovered in the limit $\Sigma$ 0 (Rojas et al., 2024).

A closely related Born–Infeld form packages the same finite series into a determinant: $\Sigma$ 1 with expansion

$\Sigma$ 2

This determinant is a geometric BI volume element built from intrinsic and extrinsic worldvolume geometry (Cruz et al., 2012).

3. Conserved tensors and second-order dynamics

The mechanical content of LBG is encoded in a sequence of symmetric conserved tensors,

$\Sigma$ 3

which satisfy

$\Sigma$ 4

by the Codazzi–Mainardi relation. They obey the recursion

$\Sigma$ 5

and the contraction identity

$\Sigma$ 6

These objects are the brane analogues of Lovelock tensors in bulk gravity (Bagatella-Flores et al., 2016).

For normal deformations the first variation of the action yields

$\Sigma$ 7

hence the Euler–Lagrange equation is

$\Sigma$ 8

Using $\Sigma$ 9, the equation of motion may equivalently be written as a finite sum of Lovelock-type geometric invariants (Bagatella-Flores et al., 2016).

The second-order character of the theory is fundamental. Although $(p+1)$ 0 contains second derivatives of the embedding, the antisymmetrized structure of the $(p+1)$ 1 together with the flat-space Codazzi identity ensures that no higher derivatives survive in the field equations. In the classification developed for general reparametrization-invariant theories $(p+1)$ 2, the subset with second-order equations is singled out by choosing the Lagrangians to be the discriminants of the extrinsic-curvature matrix,

$(p+1)$ 3

The resulting general LBG action is

$(p+1)$ 4

and the corresponding conserved current becomes purely tangential (Rojas, 26 Feb 2026).

This framework contains geodetic brane gravity as a particular case. Since

$(p+1)$ 5

the $(p+1)$ 6 sector reproduces the Regge–Teitelboim or geodetic brane gravity equation, so GBG is a special case of LBG (Rojas, 26 Feb 2026). In the cosmological formulation based on GHY- and GHYM-type terms, the combined tensor

$(p+1)$ 7

gives the compact equation

$(p+1)$ 8

or, with matter,

$(p+1)$ 9

This is again second order in the embedding functions (Arroyo et al., 7 Sep 2025).

4. Perturbations, Jacobi equation, and stability

LBG admits a fully covariant perturbation theory for small deformations of the worldvolume. Only normal deformations are physical, so the perturbation is taken to be

$m$ 0

with $m$ 1 a scalar field on the worldvolume. The induced variations are

$m$ 2

This isolates a single propagating degree of freedom, the transverse “breathing mode” $m$ 3 (Bagatella-Flores et al., 2016).

The second variation leads to a quadratic action

$m$ 4

and the corresponding Jacobi equation is

$m$ 5

Using $m$ 6, this can be written as a wave-type equation for $m$ 7 (Bagatella-Flores et al., 2016).

The DNG limit is immediate: because $m$ 8, one recovers

$m$ 9

Thus the DNG model is the lowest-order special case inside the LBG perturbation hierarchy (Bagatella-Flores et al., 2016).

For a de Sitter worldvolume,

$\mathcal M$ 0

with

$\mathcal M$ 1

the conserved tensors and invariants reduce to

$\mathcal M$ 2

Then the entire Jacobi equation collapses, provided the prefactor does not vanish, to

$\mathcal M$ 3

In Klein–Gordon form this corresponds to

$\mathcal M$ 4

so the mass-squared is tachyonic. The analysis of spherical harmonics on the de Sitter slices shows that the de Sitter membrane inherits the instability structure already known in the DNG case (Bagatella-Flores et al., 2016).

5. Reformulations: holography, disformal geometry, and mimetic embedding gravity

A central structural result is that LBG is naturally holographic in the sense that the action splits into bulk and surface terms, with the surface term completely determined by the bulk term. Using the Gauss–Codazzi rewriting of the brane invariants in terms of the intrinsic Riemann tensor $\mathcal M$ 5, one can express even and odd sectors as

$\mathcal M$ 6

where $\mathcal M$ 7 and $\mathcal M$ 8 have the algebraic symmetries of the Riemann tensor and satisfy the requisite conservation laws (Rojas, 2019).

For Lagrangians of the form

$\mathcal M$ 9

the action decomposes as

$N=p+2$ 0

with the explicit holographic identity

$N=p+2$ 1

or equivalently

$N=p+2$ 2

This is the brane analogue of the $N=p+2$ 3-plus-boundary decomposition in Einstein–Hilbert and Lanczos–Lovelock gravity (Rojas, 2019).

The parallel-surface construction also has a variable-distance extension,

$N=p+2$ 4

which yields the induced metric

$N=p+2$ 5

This is interpreted as a disformal transformation between the geometries of the original and displaced worldvolumes, and the construction opens a connection with scalar-tensor theories, including Horndeski- and Galileon-type structures, on the brane trajectory (Rojas et al., 2024).

A further reformulation presents LBG as a mimetic embedding gravity. In that construction the geometric current is augmented by a conserved “dark” current $N=p+2$ 6 with tangential decomposition

$N=p+2$ 7

Consistency of the second-order brane dynamics forces the current to be purely tangential,

$N=p+2$ 8

and the modified equation of motion becomes

$N=p+2$ 9

In the GBG limit this extra tangential source behaves as an embedding-matter contribution. The paper interprets $y^\mu=X^\mu(x^a), \qquad a=0,1,\dots,p,$ 0 through elasticity theory as an internal stress current and notes that the associated fictional energy-momentum tensor is conserved, tangential, and fluid-like (Rojas, 26 Feb 2026).

A cosmological realization of LBG is constructed for a $y^\mu=X^\mu(x^a), \qquad a=0,1,\dots,p,$ 1-dimensional FRW brane evolving in a $y^\mu=X^\mu(x^a), \qquad a=0,1,\dots,p,$ 2-dimensional Minkowski background. The action is

$y^\mu=X^\mu(x^a), \qquad a=0,1,\dots,p,$ 3

Here $y^\mu=X^\mu(x^a), \qquad a=0,1,\dots,p,$ 4 is GHY-type, $y^\mu=X^\mu(x^a), \qquad a=0,1,\dots,p,$ 5 is the intrinsic Einstein–Hilbert term on the brane, and the cubic term is GHYM-type. The embedding

$y^\mu=X^\mu(x^a), \qquad a=0,1,\dots,p,$ 6

into

$y^\mu=X^\mu(x^a), \qquad a=0,1,\dots,p,$ 7

induces the FRW metric

$y^\mu=X^\mu(x^a), \qquad a=0,1,\dots,p,$ 8

Reparametrization invariance gives a conserved quantity $y^\mu=X^\mu(x^a), \qquad a=0,1,\dots,p,$ 9, and in cosmic gauge the master equation becomes

$(p+1)$ 00

The integration constant $(p+1)$ 01 is interpreted as a dark-radiation-like or extra-dimensional energy contribution. The model exhibits self-accelerating and non-self-accelerating branches; when $(p+1)$ 02 it reduces to DGP-type cosmology, and Einstein cosmology is recovered when the radiation-like contribution and the odd extrinsic-curvature polynomials vanish (Arroyo et al., 7 Sep 2025).

The phrase “Lovelock branes” also appears in a distinct but related bulk-gravity literature. For warped-product metrics of the form

$(p+1)$ 03

generic Lovelock couplings require the base metric to be a Lovelock space, meaning that its intrinsic Lovelock tensors are proportional to the metric. For unique-vacuum Lovelock theories the constraints are much weaker, and black strings or branes can be constructed from lower-dimensional Lovelock solutions (Kastor et al., 2017). This is not the embedding-based LBG action, but it belongs to the broader Lovelock-brane landscape.

A second neighboring framework concerns codimension-even conical defects in topological AdS gravity. There the defect contributes a delta-function term to the Lovelock scalar,

$(p+1)$ 04

and the Lovelock–Chern–Simons action localizes on the defect as a lower-dimensional Lovelock action. In Euclidean signature these codimension-even defects appear as brane solutions, and the logarithmic divergence of the defect partition function matches the Euclidean brane on-shell action (Kastikainen, 2020).

Further related constructions include horizonless magnetic brane geometries in Lovelock gravity coupled to nonlinear electrodynamics, where the spacetime has no curvature singularity and no horizon but does have a conic singularity with deficit angle, and the main conclusion is that the deficit angle is independent of the Lovelock coefficients $(p+1)$ 05 (Hendi et al., 2015). In holography, third-order Lovelock-Maxwell black branes admit a finite-cutoff fluid dual obeying forced incompressible Navier–Stokes equations, with $(p+1)$ 06 independent of the cutoff surface and unaffected by the third-order Lovelock term, while the kinematic viscosity receives higher-curvature corrections (Zou et al., 2013). In black-hole thermodynamics, scalar-hairy Lovelock gravity preserves the zeroth law: for a general Killing horizon, the surface gravity is constant provided the matter sector satisfies the dominant energy condition (Fang et al., 2022).

Taken together, these developments place LBG at the intersection of embedded-surface mechanics, higher-curvature gravity, holography, and braneworld cosmology. The common structural theme is the use of Lovelock-type combinations to preserve second-order dynamics while extending the geometric content well beyond the DNG model.