Doubly Separable Spacetimes

• Doubly separable spacetimes are defined by the complete variable separation of both the Hamilton–Jacobi equation for geodesic motion and the Klein–Gordon equation in specific coordinate systems.
• Recent analyses use algorithms like the Azreg–Aïnou–Chen–Kocherlakota–Narayan method to derive the KSZ family, highlighting strict conditions and hidden symmetries in stationary, axisymmetric metrics.
• Extensions to doubly foliable and doubly warped metrics reveal practical impacts on black hole shadow modeling, self-gravitating matter constraints, and higher-dimensional separability.

Searching arXiv for papers on doubly separable spacetimes and related terminology. “Doubly separable spacetimes” is not a single universally fixed term in the relativity literature. In its most specific recent use, it denotes stationary, axisymmetric, asymptotically flat circular spacetimes for which both the Hamilton–Jacobi equation for geodesic motion and the scalar Klein–Gordon equation admit complete separation of variables; within that setting, the asymptotically-flat Azreg–Aïnou–Chen–Kocherlakota–Narayan construction yields exactly the Konoplya–Stuchlík–Zhidenko class (Kocherlakota et al., 24 Jul 2025). In other works, the same expression, or closely adjacent terminology, is used for doubly-foliable $2+1+1$ decompositions, doubly twisted or doubly warped product metrics, and higher-dimensional warped products that preserve Klein–Gordon separability (Gergely et al., 2020). This suggests that the term is best understood through its local research context rather than as a single invariant category.

1. Terminology and principal usages

Across the cited literature, the expression appears in several technically distinct settings.

Usage	Defining structure	Representative source
Stationary-axisymmetric separability	Geodesic Hamilton–Jacobi and scalar Klein–Gordon equations both separate	(Kocherlakota et al., 24 Jul 2025)
Doubly-foliable spacetime	Time foliation plus singled-out spatial foliation, leading to a $2+1+1$ split	(Gergely et al., 2020)
Doubly twisted or conformally separable metric	$$ds^2=-b^2(t,q)dt^2+a^2(t,q)g^*_{\mu\nu}(q)dq^\mu dq^\nu$$	(Mantica et al., 2021)
Higher-dimensional warped separability	Warped product of two Klein–Gordon separable spaces	(Kolar et al., 2015)

In the stationary-axisymmetric four-dimensional literature, the formal definition is explicit: a spacetime is “doubly separable” if both the Hamilton–Jacobi equation for geodesic motion and the scalar wave equation admit complete separation of variables (Kocherlakota et al., 24 Jul 2025). In the doubly-foliable literature, “doubly separable” is used in the sense of admitting two independent foliations, one temporal and one spatial, whose intersections are two-dimensional spacelike surfaces supporting a canonical evolution scheme (Gergely et al., 2020). In the geometric classification literature, the same general idea is attached to Yano’s “conformally separable” or “doubly twisted” metrics, later characterized by the existence of doubly torqued vector fields (Mantica et al., 2021).

The coexistence of these usages is not contradictory, but it does mean that separability may refer to distinct structures: separation of dynamical equations, separation by coordinate blocks in the metric, or separation induced by foliation.

2. Stationary, axisymmetric doubly separable metrics

The recent four-dimensional definition is tied to stationary, axisymmetric, asymptotically flat, circular spacetimes written in Boyer–Lindquist coordinates. In that setting, the Azreg–Aïnou–Chen–Kocherlakota–Narayan algorithm produces the metric

$$ds^2 = -\left(1 - \frac{2F}{\Sigma}\right) dt^2 - 2 \frac{2F}{\Sigma} a \sin^2\vartheta \, dt\, d\varphi + \frac{\Pi}{\Sigma} \sin^2\vartheta \, d\varphi^2 + \frac{\Sigma}{\Delta} g \, dr^2 + \Sigma \, d\vartheta^2,$$

with

$$2F(r) = (1-f)R^2,\qquad \Delta(r)=fR^2+a^2,\qquad \Sigma(r,\vartheta)=R^2+a^2\cos^2\vartheta,$$

where $f(r), g(r), R(r)$ are seed functions and $a$ is the spin parameter (Kocherlakota et al., 24 Jul 2025).

A central structural result is that the ACKN metric is identical to the Konoplya–Stuchlík–Zhidenko metric. In the identity map of Boyer–Lindquist coordinates, the free functions are related by

$$R_\Sigma=\frac{R^2}{r^2},\qquad R_B=\sqrt{g},\qquad R_M=\frac{(1-f)R^2}{r}.$$

Accordingly, the asymptotically-flat ACKN algorithm does not generate a broader family than KSZ; it generates the KSZ family exclusively (Kocherlakota et al., 24 Jul 2025).

This result refines an earlier analysis of Newman–Janis–generated rotating metrics. In that framework, the general stationary, axisymmetric metric admits complete Hamilton–Jacobi separation when one metric function is additively separable, while massive Klein–Gordon separability requires $Y\equiv \sqrt{G/F}=1$, equivalently $F=G$, which forces

$2+1+1$0

That condition identifies precisely the KSZ/Carter separable subclass inside the Newman–Janis family (Chen et al., 2019).

The consequence is a sharply delimited metric class: doubly separable stationary spacetimes are not an arbitrary collection of deformed Kerr-like geometries, but a specific separability-preserving family with three radial free functions.

3. Hidden symmetries and the separability structure

Geodesic separability is expressed through a separated principal function. For the ACKN/KSZ class, the momentum 1-form takes the form

$2+1+1$1

with separated radial and polar potentials $2+1+1$2 and $2+1+1$3, and with first-order equations in Mino time derived from these potentials (Kocherlakota et al., 24 Jul 2025). The additional integral of motion is the Carter constant

$2+1+1$4

generated by a rank-2 Killing tensor satisfying

$2+1+1$5

Scalar-wave separability is formulated operatorially. The massive Klein–Gordon equation admits multiplicative separation with

$2+1+1$6

when there is a complete set of mutually commuting operators

$2+1+1$7

built from the same Killing tensor (Kocherlakota et al., 24 Jul 2025).

The separability hierarchy is stricter for spinor fields. A candidate Killing–Yano 2-form can be constructed, but it satisfies the Killing–Yano equation only in the degenerate subclass

$2+1+1$8

Only in this subclass is the Dirac equation expected to be separable in the minimal Killing–Yano sense (Kocherlakota et al., 24 Jul 2025). Earlier Newman–Janis analysis had already isolated the stronger scalar separability condition $2+1+1$9, which places the metric in the KSZ/Carter sector (Chen et al., 2019).

Petrov classification clarifies why geodesic and scalar separability do not automatically imply the full Kerr-like hidden-symmetry package. Generic ACKN/KSZ spacetimes are Petrov Type I; the degenerate subclass $$ds^2=-b^2(t,q)dt^2+a^2(t,q)g^*_{\mu\nu}(q)dq^\mu dq^\nu$$0 is Type D; only Minkowski is Type O (Kocherlakota et al., 24 Jul 2025). A common misconception is therefore ruled out: doubly separable is not synonymous with Type D, nor with Dirac separable.

4. Symmetry constraints on self-gravitating matter

The most restrictive results concern source compatibility. In the rest frame of stationary configurations, the paper considers the diagonal stress–energy forms

$$ds^2=-b^2(t,q)dt^2+a^2(t,q)g^*_{\mu\nu}(q)dq^\mu dq^\nu$$1

These imply corresponding algebraic conditions on the rest-frame Einstein tensor and lead to a series of no-go theorems (Kocherlakota et al., 24 Jul 2025).

Matter model	Rest-frame form	Result in the doubly separable class
Perfect fluid	$$ds^2=-b^2(t,q)dt^2+a^2(t,q)g^*_{\mu\nu}(q)dq^\mu dq^\nu$$2	Only Kerr vacuum satisfies the required equalities
Massless real scalar	$$ds^2=-b^2(t,q)dt^2+a^2(t,q)g^*_{\mu\nu}(q)dq^\mu dq^\nu$$3	No compatible nontrivial doubly separable solution
Electromagnetic field	$$ds^2=-b^2(t,q)dt^2+a^2(t,q)g^*_{\mu\nu}(q)dq^\mu dq^\nu$$4	Only Kerr–Newman occurs

For perfect fluids, the required equality $$ds^2=-b^2(t,q)dt^2+a^2(t,q)g^*_{\mu\nu}(q)dq^\mu dq^\nu$$5 holds only for Kerr, with

$$ds^2=-b^2(t,q)dt^2+a^2(t,q)g^*_{\mu\nu}(q)dq^\mu dq^\nu$$6

Hence nontrivial doubly separable spacetimes cannot be sourced by perfect fluids. For massless real scalar fields, the simultaneous conditions $$ds^2=-b^2(t,q)dt^2+a^2(t,q)g^*_{\mu\nu}(q)dq^\mu dq^\nu$$7 and the meridional/tangential equalities force

$$ds^2=-b^2(t,q)dt^2+a^2(t,q)g^*_{\mu\nu}(q)dq^\mu dq^\nu$$8

and with $$ds^2=-b^2(t,q)dt^2+a^2(t,q)g^*_{\mu\nu}(q)dq^\mu dq^\nu$$9 this becomes $$ds^2 = -\left(1 - \frac{2F}{\Sigma}\right) dt^2 - 2 \frac{2F}{\Sigma} a \sin^2\vartheta \, dt\, d\varphi + \frac{\Pi}{\Sigma} \sin^2\vartheta \, d\varphi^2 + \frac{\Sigma}{\Delta} g \, dr^2 + \Sigma \, d\vartheta^2,$$0, which is electromagnetic rather than scalar. The only electromagnetic members are Kerr–Newman, identified by

$$ds^2 = -\left(1 - \frac{2F}{\Sigma}\right) dt^2 - 2 \frac{2F}{\Sigma} a \sin^2\vartheta \, dt\, d\varphi + \frac{\Pi}{\Sigma} \sin^2\vartheta \, d\varphi^2 + \frac{\Sigma}{\Delta} g \, dr^2 + \Sigma \, d\vartheta^2,$$1

Accordingly, electromagnetic fields lead only to the Kerr–Newman family (Kocherlakota et al., 24 Jul 2025).

A notable consequence is the exclusion of the spinning Janis–Newman–Winicour proposal from this class. Although the static JNW seed solves Einstein–Klein–Gordon in spherical symmetry, its rotating generalization obtained through Azreg–Aïnou methods with the Carter tetrad does not satisfy the required conformal-factor PDE and is not quasi-degenerate, so it is not a doubly separable ACKN/KSZ spacetime (Kocherlakota et al., 24 Jul 2025).

These results establish a substantive cautionary principle: separability-preserving algorithmic metrics can possess strong hidden symmetries while still failing to represent physically admissible Einstein–matter solutions with the intended source.

5. Double foliation, doubly twisted metrics, and warped subclasses

In a different branch of the literature, doubly separability is formulated geometrically through foliations or through blockwise conformal factors. In the $$ds^2 = -\left(1 - \frac{2F}{\Sigma}\right) dt^2 - 2 \frac{2F}{\Sigma} a \sin^2\vartheta \, dt\, d\varphi + \frac{\Pi}{\Sigma} \sin^2\vartheta \, d\varphi^2 + \frac{\Sigma}{\Delta} g \, dr^2 + \Sigma \, d\vartheta^2,$$2 formalism, a spacetime is doubly-foliable if it admits a time foliation with unit timelike normal $$ds^2 = -\left(1 - \frac{2F}{\Sigma}\right) dt^2 - 2 \frac{2F}{\Sigma} a \sin^2\vartheta \, dt\, d\varphi + \frac{\Pi}{\Sigma} \sin^2\vartheta \, d\varphi^2 + \frac{\Sigma}{\Delta} g \, dr^2 + \Sigma \, d\vartheta^2,$$3 and a singled-out spatial foliation with unit spacelike normal $$ds^2 = -\left(1 - \frac{2F}{\Sigma}\right) dt^2 - 2 \frac{2F}{\Sigma} a \sin^2\vartheta \, dt\, d\varphi + \frac{\Pi}{\Sigma} \sin^2\vartheta \, d\varphi^2 + \frac{\Sigma}{\Delta} g \, dr^2 + \Sigma \, d\vartheta^2,$$4, whose intersections $$ds^2 = -\left(1 - \frac{2F}{\Sigma}\right) dt^2 - 2 \frac{2F}{\Sigma} a \sin^2\vartheta \, dt\, d\varphi + \frac{\Pi}{\Sigma} \sin^2\vartheta \, d\varphi^2 + \frac{\Sigma}{\Delta} g \, dr^2 + \Sigma \, d\vartheta^2,$$5 are two-dimensional spacelike surfaces with induced metric

$$ds^2 = -\left(1 - \frac{2F}{\Sigma}\right) dt^2 - 2 \frac{2F}{\Sigma} a \sin^2\vartheta \, dt\, d\varphi + \frac{\Pi}{\Sigma} \sin^2\vartheta \, d\varphi^2 + \frac{\Sigma}{\Delta} g \, dr^2 + \Sigma \, d\vartheta^2,$$6

The line element becomes

$$ds^2 = -\left(1 - \frac{2F}{\Sigma}\right) dt^2 - 2 \frac{2F}{\Sigma} a \sin^2\vartheta \, dt\, d\varphi + \frac{\Pi}{\Sigma} \sin^2\vartheta \, d\varphi^2 + \frac{\Sigma}{\Delta} g \, dr^2 + \Sigma \, d\vartheta^2,$$7

with two lapses, two shifts, and a cross-shift $$ds^2 = -\left(1 - \frac{2F}{\Sigma}\right) dt^2 - 2 \frac{2F}{\Sigma} a \sin^2\vartheta \, dt\, d\varphi + \frac{\Pi}{\Sigma} \sin^2\vartheta \, d\varphi^2 + \frac{\Sigma}{\Delta} g \, dr^2 + \Sigma \, d\vartheta^2,$$8. The Einstein–Hilbert action then yields canonical pairs $$ds^2 = -\left(1 - \frac{2F}{\Sigma}\right) dt^2 - 2 \frac{2F}{\Sigma} a \sin^2\vartheta \, dt\, d\varphi + \frac{\Pi}{\Sigma} \sin^2\vartheta \, d\varphi^2 + \frac{\Sigma}{\Delta} g \, dr^2 + \Sigma \, d\vartheta^2,$$9, $$2F(r) = (1-f)R^2,\qquad \Delta(r)=fR^2+a^2,\qquad \Sigma(r,\vartheta)=R^2+a^2\cos^2\vartheta,$$0, and $$2F(r) = (1-f)R^2,\qquad \Delta(r)=fR^2+a^2,\qquad \Sigma(r,\vartheta)=R^2+a^2\cos^2\vartheta,$$1, together with Hamiltonian, tangential diffeomorphism, and cross-momentum constraints (Gergely et al., 2020).

The same broad theme appears in the classification of doubly twisted metrics,

$$2F(r) = (1-f)R^2,\qquad \Delta(r)=fR^2+a^2,\qquad \Sigma(r,\vartheta)=R^2+a^2\cos^2\vartheta,$$2

which Yano identified with the conformally separable case. Mantica and Molinari characterize these spacetimes by the existence of a timelike doubly torqued vector $$2F(r) = (1-f)R^2,\qquad \Delta(r)=fR^2+a^2,\qquad \Sigma(r,\vartheta)=R^2+a^2\cos^2\vartheta,$$3 satisfying

$$2F(r) = (1-f)R^2,\qquad \Delta(r)=fR^2+a^2,\qquad \Sigma(r,\vartheta)=R^2+a^2\cos^2\vartheta,$$4

The reductions are immediate: doubly warped spacetimes have $$2F(r) = (1-f)R^2,\qquad \Delta(r)=fR^2+a^2,\qquad \Sigma(r,\vartheta)=R^2+a^2\cos^2\vartheta,$$5 and $$2F(r) = (1-f)R^2,\qquad \Delta(r)=fR^2+a^2,\qquad \Sigma(r,\vartheta)=R^2+a^2\cos^2\vartheta,$$6; twisted spacetimes set $$2F(r) = (1-f)R^2,\qquad \Delta(r)=fR^2+a^2,\qquad \Sigma(r,\vartheta)=R^2+a^2\cos^2\vartheta,$$7; generalized Robertson–Walker spacetimes further reduce to $$2F(r) = (1-f)R^2,\qquad \Delta(r)=fR^2+a^2,\qquad \Sigma(r,\vartheta)=R^2+a^2\cos^2\vartheta,$$8 (Mantica et al., 2021). A related classification extends the same doubly torqued framework to Kundt and PP-wave geometries in the null case (Mantica et al., 2021).

Spherical doubly warped spacetimes provide a concrete physical realization: $$2F(r) = (1-f)R^2,\qquad \Delta(r)=fR^2+a^2,\qquad \Sigma(r,\vartheta)=R^2+a^2\cos^2\vartheta,$$9 In this class, the Ricci tensor and electric Weyl tensor are rank-2 objects built from the fluid velocity, acceleration, and radial direction. Einstein’s equations force an imperfect-fluid energy–momentum tensor with anisotropic pressures and heat flux; in radiating collapse, matching to Vaidya yields the boundary condition

$f(r), g(r), R(r)$0

The GRW limit $f(r), g(r), R(r)$1 eliminates heat flux but retains possible anisotropy through the electric Weyl tensor, producing Friedmann-like equations modified by $f(r), g(r), R(r)$2 and $f(r), g(r), R(r)$3 terms (Mantica et al., 2022).

6. Higher-dimensional extensions, spinning particles, and applications

Higher-dimensional separability theory generalizes the notion in two directions. First, canonical $f(r), g(r), R(r)$4 metrics with Stäckel structure admit separated Klein–Gordon solutions and towers of commuting operators built from Killing vectors and rank-2 Killing tensors, often generated by a principal conformal Killing–Yano 2-form. Second, warped products of two Klein–Gordon separable spaces remain separable when the warp factor is chosen as

$f(r), g(r), R(r)$5

Within this ansatz, solving the Einstein equations yields warped Euclidean Kerr–NUT–(A)dS limits and a new Einstein–Kähler branch (Kolar et al., 2015).

A distinct 2026 generalization defines a doubly separable spacetime by two simultaneous requirements: the geodesic Hamilton–Jacobi equation is separable in coordinates $f(r), g(r), R(r)$6, and parallel transport along the corresponding geodesics is torus-reducible and additively separable in the same coordinates. Under these hypotheses, the Hamilton–Jacobi equation for a spinning test particle, to linear order in spin and with the Tulczyjew–Dixon supplementary condition, is also separable. The theorem is illustrated in Kerr–NUT–(A)dS, plane waves, and FLRW cosmologies (Witzany et al., 24 Jun 2026).

The separability structure also has coordinate and observational uses. In stationary spacetimes with separable geodesic action, one can construct geodesic-action slicings with unit lapse, so that the orthogonal congruence consists of freely falling observers; Schwarzschild Painlevé–Gullstrand coordinates, Doran-type Kerr and Kerr–Newman slicings, and a local Gödel construction arise in this way (Bini et al., 2014). In the ACKN/KSZ setting, separability reduces geodesics and scalar waves to ordinary differential equations, which is why the class is explicitly described as useful for black hole shadows, photon rings, photon shells, quasinormal modes, and wave scattering (Kocherlakota et al., 24 Jul 2025).

Another geometric application appears in the doubly warped treatment of the GMGHS interior. There the metric is written as a doubly warped product with base $f(r), g(r), R(r)$7, fibers $f(r), g(r), R(r)$8 and $f(r), g(r), R(r)$9, and two independent scale functions, yielding a Kantowski–Sachs-type anisotropic cosmology after reinterpretation of the base coordinate as cosmological time (Choi, 2014).

The unifying theme across these settings is not a single metric form, but a controlled decomposition of dynamics or geometry into two privileged sectors. In the stationary-axisymmetric case this decomposition is exceptionally rigid: it enforces precise hidden symmetries, admits only a limited Dirac-separable subclass, and strongly constrains the admissible self-gravitating matter content.