---
title: Linearized Einstein–Bianchi System
url: https://www.emergentmind.com/topics/linearized-einstein-bianchi-system
type: topic
---

# Linearized Einstein–Bianchi System

Searching arXiv for recent and foundational papers on the linearized Einstein–Bianchi system.
The linearized Einstein–Bianchi system is the gauge-theoretic and constraint-propagating structure obtained by linearizing Einstein’s equations and their Bianchi identities around a fixed background, most commonly Minkowski spacetime or an Einstein background. In the weak-field flat-background setting, it appears as the linearized Einstein equations together with the divergence-free structure inherited from the contracted Bianchi identities, and it yields a Maxwell-like gravitoelectromagnetic system for the gravitoelectric and gravitomagnetic fields [1710.01593]. On general Einstein backgrounds, the same structure can be formulated invariantly as a differential complex in which infinitesimal diffeomorphisms, metric perturbations, and linearized field equations are linked by natural bundle maps; in that setting the vanishing of the composition of the gauge and field operators encodes the linearized Bianchi identity [2210.17293]. In \(3+1\) and related decompositions, the same phrase also refers to the subsidiary linear first-order system governing the propagation of Einstein constraints, derived from the contracted Bianchi identities and shown to be symmetric hyperbolic under appropriate hypotheses [1406.1016]. A complementary tradition treats the electric and magnetic parts of the Weyl tensor as the primary variables and studies a first-order curl-div system for symmetric trace-free tensors, together with finite element or lattice discretizations designed to preserve the underlying complex and constraints [2103.00088], [1104.1356].

## 1. Linearized geometric framework

The standard weak-field construction starts from a metric decomposition
\[
g_{ik} \simeq \eta_{ik} + h_{ik}, \qquad |h_{ik}|\ll |\eta_{ik}|,
\]
with \(\eta_{ik}\) the Minkowski metric and \(h_{ik}\) the perturbation [1710.01593]. In that setting only terms linear in \(h_{ik}\) are retained, source velocities are taken to satisfy \(|\vec v|\ll c\), and only first-order expressions for the curvature are used. The trace \(h = h^i{}_i\) is introduced, and the trace-reversed perturbation
\[
\Psi^{ik} := h^{ik} - \eta^{ik} h
\]
is used together with the harmonic gauge condition
\[
\partial_i \Psi^{ik} = 0.
\]
Within these assumptions the linearized Riemann tensor is
\[
R_{iklm} \simeq \frac{1}{2}\left( \partial_k\partial_m h_{il} + \partial_i\partial_l h_{km} - \partial_k\partial_l h_{im} - \partial_i\partial_m h_{kl} \right),
\]
while the Ricci tensor and scalar curvature reduce to
\[
R_{ik} \simeq -\frac{1}{2}\Box h_{ik}, \qquad
R \simeq -\frac{1}{2}\Box h,
\]
with
\[
\Box = \nabla^2 - \frac{1}{c^2}\frac{\partial^2}{\partial t^2}
\]
[1710.01593].

On a general Einstein background \((M,g_{ab})\) satisfying
\[
R_{ab} = \Lambda g_{ab},
\]
the perturbation is written
\[
\tilde g_{ab} = g_{ab} + \epsilon h_{ab},
\]
and the linearized geometry is encoded using the Calabi operator \(\mathcal C\) from projective differential geometry [2210.17293]. In that formulation the linearized Riemann tensor is
\[
\delta R_{abcd}[h]
= -\Big[(\mathcal{C}h)_{abcd}
+ R_{abe}{}^{f}h_{cdf} + R_{cde}{}^{f}h_{abf}\Big],
\]
and the linearized Einstein vacuum equation with cosmological constant is expressed through the Ricci contraction
\[
P(h)_{bd} := g^{ac}(\mathcal{C}h)_{abcd}
\]
as
\[
P(h)_{ab} = 0
\]
[2210.17293]. In explicit form,
\[
P(h)_{bd} = \Delta h_{bd}
- 2\nabla^e \nabla_{(b} h_{d)e}
+ \nabla_b \nabla_d h
+ 2R_{(b}{}^e h_{d)e},
\]
and on an Einstein background this becomes
\[
P(h)_{bd} = \Delta h_{bd}
- 2\nabla^e \nabla_{(b} h_{d)e}
+ \nabla_b \nabla_d h
+ 2\Lambda h_{bd}
\]
[2210.17293].

These two formulations are the flat-background and background-covariant versions of the same structure. The first emphasizes harmonic gauge and explicit wave operators; the second packages the system as a differential complex and makes gauge invariance intrinsic. A plausible implication is that the term “linearized Einstein–Bianchi system” has no single canonical realization, but rather a family of equivalent realizations adapted to flat-space perturbation theory, Einstein-background deformation theory, \(3+1\) PDE analysis, or curvature-based formulations.

## 2. Einstein equations, Bianchi identities, and gauge structure

In the weak-field flat-background setting, Einstein’s equations
\[
R_{ik} - \frac{1}{2}g_{ik}R = \frac{8\pi G}{c^4}T_{ik}
\]
reduce, after using the harmonic gauge, to
\[
-\partial_j G^j{}_{ik} \simeq \frac{8\pi G}{c^4}T_{ik},
\]
where
\[
G^i{}_{jk} := \partial_j \Psi^i{}_k - \partial_k \Psi^i{}_j
\]
up to the paper’s normalization [1710.01593]. This is the “Einstein” part of the system. The “Bianchi” part comes from the full identities
\[
R^i{}_{jkl;m} + R^i{}_{jlm;k} + R^i{}_{jmk;l} = 0
\]
and their contraction
\[
R^i{}_{k;i} - \frac{1}{2}R_{;k} = 0
\quad\Longleftrightarrow\quad
\nabla_i G^i{}_k = 0,
\]
which in linearized form becomes
\[
\partial_i G^i{}_k = 0.
\]
Einstein’s equations then imply
\[
\partial_i T^{ik} = 0,
\]
so source conservation is not an independent assumption but a compatibility condition enforced by the Bianchi identities [1710.01593].

On Einstein backgrounds, the same compatibility is encoded by the complex
\[
T^*M \xrightarrow{K} S^2T^*M \xrightarrow{P} S^2T^*M,
\]
where
\[
K(\xi)_{ab} := \nabla_{(a}\xi_{b)}
\]
is the Killing operator and \(P\) is the Einstein deformation operator [2210.17293]. The key structural statement is
\[
P\circ K = 0.
\]
This means that pure-gauge perturbations \(h_{ab}=\nabla_{(a}\xi_{b)}\) automatically solve the linearized equations, and that the field equations satisfy differential identities expressing linearized diffeomorphism invariance [2210.17293]. In this language, gauge-equivalent perturbations differ by elements of \(\operatorname{im}K\), and physical perturbations modulo gauge correspond to
\[
\ker P / \operatorname{im}K.
\]

A recurrent misconception is that the Bianchi identity merely supplies a divergence condition after the field equations have been imposed. The invariant complex formulation shows a stronger statement: the linearized Bianchi identity is built into the operator structure itself through the annihilation of pure gauge modes by the field operator [2210.17293]. In the flat weak-field picture, the same point appears as the preservation of the harmonic gauge and conservation of the linearized sources [1710.01593].

## 3. Maxwell-type and Weyl-curvature formulations

One of the most widely used realizations of the linearized Einstein–Bianchi system is the gravitoelectromagnetic or Weyl-electric/Weyl-magnetic formulation. In the weak-field approximation around Minkowski spacetime, the perturbation components define a gravitoelectric field \(\vec E_g\), a gravitomagnetic field \(\vec B_g\), a vector potential \(\vec A\), a space–time–mass density \(\rho_g\), and a space–time–mass current density \(\vec J_g\), leading to
\[
\nabla\cdot\vec{E}_g \simeq -4\pi G\rho_g,\qquad
\nabla\times\vec{E}_g = -\frac{1}{c}\frac{\partial \vec{B}_g}{\partial t},
\]
\[
\nabla\cdot\vec{B}_g = 0,\qquad
\nabla\times\vec{B}_g \simeq -\frac{4\pi G}{c^2}\vec J_g + \frac{1}{c}\frac{\partial \vec{E}_g}{\partial t}.
\]
These equations, together with the continuity law
\[
\frac{\partial \rho_g}{\partial t} + \nabla\cdot\vec J_g = 0,
\]
are identified as the GEM manifestation of the linearized Einstein–Bianchi system [1710.01593].

In the vacuum case, this reduces to the wave system
\[
\nabla\times\vec{E}_g = -\frac{1}{c}\frac{\partial \vec{B}_g}{\partial t},\qquad
\nabla\times\vec{B}_g = \frac{1}{c}\frac{\partial \vec{E}_g}{\partial t},
\]
and hence
\[
\nabla^2\vec E_g - \frac{1}{c^2}\frac{\partial^2\vec E_g}{\partial t^2}=0,\qquad
\nabla^2\vec B_g - \frac{1}{c^2}\frac{\partial^2\vec B_g}{\partial t^2}=0
\]
[1710.01593]. This is the most direct analogue of Maxwell theory, with the sign difference in Gauss’s law reflecting the attractive nature of gravity.

A more invariant three-dimensional curvature formulation uses the electric and magnetic parts of the Weyl tensor, denoted \(E\) and \(B\), both symmetric and trace-free. In the finite-element literature the first-order system is written
\[
\dot{E}+\curl B=0,\quad \div E=0,\qquad
\dot{B}-\curl E=0,\quad \div B=0,
\]
with curl and divergence acting row-wise on matrix fields [2103.00088]. Introducing
\[
\sigma(t)=\int_0^t \divdiv E(s)\,ds,
\]
the system is recast as
\[
\dot{\sigma}=\divdiv E,\qquad
\dot{E}=-\nabla\nabla\sigma-\sym\curl B,\qquad
\dot{B}=\curl E
\]
[2103.00088]. This version is especially important because it identifies the linearized Einstein–Bianchi system with a Hodge wave equation on a tensor complex.

The 2025 conformal-Hessian formulation sharpens this viewpoint by taking \(E\) and \(B\) to be explicitly symmetric and traceless and by using
\[
\dot{\sigma}=\bm E,\qquad
\dot{\bm E}=-\dev\hess\sigma-\sym\curl \bm B,\qquad
\dot{\bm B}=\sym\curl \bm E,
\]
with \(\bm E,\bm B\in \mathbb S\cap\mathbb T\) [2508.04560]. There the differential constraints are encoded by the exact conformal Hessian complex
\[
P_1^+ \xrightarrow{\subset} H^2(\Omega)
\xrightarrow{\dev\hess} H(\sym\curl,\Omega;\mathbb S\cap\mathbb T)
\xrightarrow{\sym\curl} H(\div,\Omega;\mathbb S\cap\mathbb T)
\xrightarrow{\div} L^2(\Omega)\to0
\]
[2508.04560]. This suggests that the Weyl-electric/Weyl-magnetic formulation is not merely an analogy with electromagnetism, but a precise instance of a Hodge-type evolution system on a geometric complex.

## 4. Constraint propagation and hyperbolic subsidiary systems

In \(1+n\) formulations, the linearized Einstein–Bianchi system frequently means the subsystem governing the propagation of the Hamiltonian and momentum constraints. Let
\[
E_{ab}:=G_{ab}-\mathcal G_{ab}=0,\qquad \nabla^a\mathcal G_{ab}=0,
\]
and define
\[
E^{(H)}:=n^e n^f E_{ef},\qquad
E^{(M)}_b:=n^e h_b{}^f E_{ef}.
\]
These are the Hamiltonian and momentum expressions associated with a foliation by hypersurfaces \(\Sigma_\sigma\) [1406.1016]. Choosing the reduced evolution equations
\[
E^{(\mathrm{EVOL})}_{ab}
=
h_a{}^c h_b{}^d E_{cd}
-\kappa\, h_{ab}E^{(H)},
\]
and setting \(\kappa=1\), one uses the contracted Bianchi identity together with \(\nabla^a\mathcal G_{ab}=0\) to derive a first-order linear homogeneous system for \((E^{(H)},E^{(M)}_a)\) [1406.1016].

In adapted coordinates, the subsidiary system has the form
\[
A^0(x)\,\partial_\sigma v + A^i(x)\,\partial_i v = B(x)\,v,
\qquad
v=(E^{(H)},E^{(M)}_j)^T,
\]
with
\[
A^0 =
\begin{pmatrix}
N & 0\\
0 & h^{ij}
\end{pmatrix}.
\]
When the foliating hypersurfaces are Riemannian, \(h^{ij}\) is positive definite, so \(A^0\) is positive definite and the system is linear first-order symmetric hyperbolic [1406.1016]. The consequence is the standard constraint-propagation result: if the constraints vanish on one leaf and the reduced evolution equations hold, then the constraints vanish throughout the domain of dependence [1406.1016].

This subsidiary hyperbolic viewpoint is analytically distinct from the GEM or Weyl-curl formulations, but it addresses the same compatibility problem. In both cases the Bianchi identities provide transport equations for constraints that are linear and homogeneous in the constraint variables. A plausible implication is that “linearized Einstein–Bianchi system” is best understood as a structural label for the compatibility layer of linearized gravity rather than as a single fixed PDE system.

Related hyperbolic formulations arise in orthonormal-frame and Einstein–Friedrich systems. For a minimally coupled nonlinear scalar field, a frame representation yields a quasi-linear first-order symmetric hyperbolic system for \((\phi,\alpha;E_{ab},B_{ab};e_a{}^\mu;\Gamma^c{}_{ab})\), where \(E_{ab}\) and \(B_{ab}\) are the electric and magnetic parts of the Weyl tensor and the Bianchi equations provide their evolution [1006.3778]. Linearization around expanding FLRW backgrounds then leads to a dissipative linearized Einstein–Bianchi system whose variables decay exponentially under suitable assumptions on the scalar potential [1006.3778]. In the \(1+3\) tetrad formalism, a different structural result shows that for generic timelike or null congruences the Einstein equations themselves can be regarded as integrability conditions for the Jacobi, Ricci, and Bianchi equations [1302.6448]. That result concerns the nonlinear theory, but it clarifies why linearized Einstein–Bianchi formulations can be closed on curvature and connection variables.

## 5. Discrete and computational realizations

Several computational approaches are built directly around the linearized Einstein–Bianchi structure. One line of work develops conforming finite element complexes. In the divdiv-based formulation, the unknowns are a scalar \(\sigma\), a symmetric tensor \(E\in H(\divdiv,\Omega;S)\), and a trace-free tensor \(B\in H(\sym\curl,\Omega;T)\), and the semidiscrete weak system is
\[
(\dot{\sigma},q) = (\divdiv E,q),
\]
\[
(\dot{E},\tau) = -(\sigma,\divdiv\tau) - (\sym\curl B,\tau),
\]
\[
(\dot{B},\chi) = (E,\sym\curl\chi)
\]
for all admissible test functions [2103.00088]. The crucial structure is the exact discrete complex
\[
RT \xrightarrow{\subset} V_{k+2,h}
\xrightarrow{\ddev\nabla} \Lambda_{k+1,h}
\xrightarrow{\sym\curl} \Sigma_{k,h}
\xrightarrow{\divdiv} P_{k-2}(\mathcal T_h)\to0,
\]
which ensures discrete Bianchi identities, compatibility of operators, and inf-sup stability [2103.00088]. With Crank–Nicolson time discretization, the fully discrete scheme achieves \(O(h^{k-1}+\Delta t^2)\) error under the stated regularity assumptions [2103.00088].

The conformal-Hessian approach advances this by enforcing both symmetry and tracelessness strongly at the level of finite element spaces. Its discrete complex is
\[
P_1^+ \xrightarrow{\subset} U_{k+3,h}
\xrightarrow{\dev\hess} \Lambda_{k+1,h}
\xrightarrow{\sym\curl} \Sigma_{k,h}
\xrightarrow{\div} P_{k-2}(\mathcal T_h)\to0,
\]
exact for \(k\ge6\) on contractible domains [2508.04560]. The corresponding semidiscrete and fully discrete Hodge-wave systems preserve symmetry and tracelessness exactly and yield convergence
\[
\|\sigma^j-\sigma_h^j\|_{L^2}
+\|\bm E^j-\bm E_h^j\|_{L^2}
+\|\bm B^j-\bm B_h^j\|_{L^2}
\lesssim
(h^{k-1}+\Delta t^2)\,\mathcal N(\sigma,\bm E,\bm B)
\]
[2508.04560]. This is explicitly presented as a discretization of the linearized Einstein–Bianchi system near Minkowski space.

A very different discretization is the smooth-lattice Einstein–Bianchi system. In the Schwarzschild reduction, the independent curvature variables are
\[
\Rx = \mathcal R_{xyxy},\qquad \Rz = \mathcal R_{xzxz},
\]
subject to the vacuum constraint
\[
\Rx + 2\Rz = 0
\]
[1101.3171]. Together with lattice leg lengths \(\Lxx,\Lzz\) and extrinsic curvatures \(\Kxx,\Kzz\), the evolution equations include
\[
\frac{d\Lxx}{dt} = -N\Kxx\Lxx,\qquad
\frac{d\Lzz}{dt} = -N\Kzz\Lzz,
\]
\[
\frac{d\Rx}{dt} = 2N\Kxx(2\Rx+\Rz),\qquad
\frac{d\Rz}{dt} = 3N\Kxx\Rz + N\Kzz(\Rx+2\Rz),
\]
where the curvature equations are obtained from the second Bianchi identity [1101.3171]. Constraint preservation is explicit:
\[
\frac{d}{dt}(\Rx+2\Rz)
=
2N(2\Kxx+\Kzz)(\Rx+2\Rz),
\]
so the Hamiltonian-like curvature constraint propagates [1101.3171]. The \(3+1\) vacuum generalization evolves all 20 independent Riemann components, uses the full uncontracted Bianchi identities for 14 evolution equations, the vacuum Einstein equations for 6 algebraic relations, and shows that each curvature component satisfies a wave equation, establishing hyperbolic structure and preservation of the associated 10 constraints [1104.1356].

In anisotropic cosmology, the linearized Einstein–Bianchi structure is also realized in Hamiltonian perturbation theory. For Bianchi I backgrounds, Fourier-expanded ADM perturbations are reorganized so that the linear scalar and vector constraints become canonical momenta
\[
\Pi_3=\mathcal S^{(1)},\qquad
\Pi_4,\Pi_5,\Pi_6 \propto \mathcal V_i^{(1)},
\]
while three configuration variables \(T_0,T_1,T_2\) are gauge invariant [2006.03397]. The reduced Hamiltonian is
\[
H_{\text{pert}}
=
N(t)V_0\frac{1}{2a^2(t)}
\sum_{\vec k}
\Big[
\Pi_\mu(\vec k)\Pi_\mu(-\vec k)
+
\sum_{\mu,\mu'=0}^2
\Big(\frac{k^2(t)}{4\kappa a^2(t)}\delta_{\mu\mu'}+U_{\mu\mu'}(t,\vec k)\Big)
T_\mu(\vec k)T_{\mu'}(-\vec k)
\Big],
\]
with \(\mu,\mu'=0,1,2\), and the equations of motion show scalar–tensor mixing through the anisotropic effective potential matrix \(U_{\mu\mu'}\) [2006.03397]. There the linearized Bianchi identities are reflected in the first-class constraint algebra and the automatic propagation of the linear constraints.

## 6. Background dependence, variants, and scope

The linearized Einstein–Bianchi system depends strongly on the chosen background and analytical setting. Around Minkowski space, the weak-field approximation yields the most transparent wave or Maxwell-type expressions [1710.01593]. Around Einstein backgrounds with \(R_{ab}=\Lambda g_{ab}\), the operator \(P\) and the complex
\[
T^*M \xrightarrow{K} S^2T^*M \xrightarrow{P} S^2T^*M
\]
provide the natural invariant formulation [2210.17293]. On compact Riemannian manifolds with boundary and arbitrary interior geometry, the relevant object is the linearized Einstein equation with sources in linear Bianchi gauge,
\[
D Ein_g\,\sigma = T,\qquad
\delta_g B_g\sigma = 0,
\]
supplemented by extended Cauchy data on the boundary [2510.12797]. In that setting global solvability is characterized by a lifted divergence condition
\[
\pzc\delta_g T = 0
\]
together with \(T|_{\partial M}=0\), and uniqueness follows from generalized Hodge theory plus unique continuation [2510.12797]. This is a genuinely elliptic Einstein–Bianchi theory rather than a hyperbolic one.

Cosmological applications supply further variants. In nearly isotropic homogeneous Bianchi cosmologies, the metric perturbation is encoded by a homogeneous trace-free tensor \(\beta_{ab}(t)\), and the linearized Einstein–Bianchi system becomes a finite system of coupled or decoupled ODEs for mode amplitudes, depending on the Bianchi type [1009.3935]. There the momentum constraint
\[
P_{\hat a}
=
e^{-\alpha}\left(
3\dot\beta_{\hat a b}a_b
-
\epsilon_{\hat a b c}\dot\beta_{b d}n_{d c}
\right)
+ \mathcal O(e^{-\alpha}HC\epsilon^2)
\]
plays the role of the linearized Einstein–Bianchi coupling between metric anisotropy and matter tilt [1009.3935]. In maximal–isothermal gauge for axisymmetric vacuum perturbations of Minkowski space, the reduced variables \(v\) and \(B=\beta_\rho\) satisfy a singular hyperbolic–elliptic system,
\[
\ddot v = \Delta v - \frac{1}{\rho}\partial_\rho v + \rho\,\partial_\rho\!\left(\frac{\beta_\rho}{\rho}\right),
\]
\[
\Delta \beta_\rho = \frac{2}{\rho}\left(\Delta v - \frac{1}{\rho}\partial_\rho v\right),
\]
which is solved explicitly by Fourier and Hankel transforms [1005.5347]. The paper does not name it an Einstein–Bianchi system explicitly, but the wave variable and elliptic gauge variable together encode a linear hyperbolic–elliptic realization of the same constraint-preserving structure [1005.5347].

A common misconception is that all linearized Einstein–Bianchi systems are first-order symmetric hyperbolic systems in the same sense. The record is more varied. Some are hyperbolic curvature systems [1104.1356], some are subsidiary symmetric hyperbolic systems for constraints [1406.1016], some are Hodge-wave systems on tensor complexes [2103.00088], [2508.04560], and some are elliptic boundary value problems in Bianchi gauge [2510.12797]. What unifies them is not a unique PDE normal form, but the coexistence of linearized Einstein equations, gauge structure from infinitesimal diffeomorphisms, and compatibility identities inherited from the Bianchi identities.

The subject therefore occupies a central position between linearized general relativity, Weyl-curvature evolution, gauge complexes, constraint propagation, and structure-preserving discretization. In each of these settings, the linearized Einstein–Bianchi system serves as the mechanism by which evolution equations, gauge freedom, and consistency conditions are organized into a closed mathematical structure [1710.01593], [2210.17293], [1406.1016], [2103.00088], [2508.04560].

Source: https://www.emergentmind.com/topics/linearized-einstein-bianchi-system