Coupled Ricci Flow

• Coupled Ricci flow is a family of geometric evolution systems where the metric evolution interacts with additional dynamical fields, measures, or submanifolds.
• It encompasses variants such as harmonic-Ricci flow, scalar field coupling, modified Ricci flow with weighted measures, and ambient-submanifold flows.
• Analytic tools including entropy functionals, eigenvalue estimates, and rigidity results are used to study singularity formation, stability, and convergence in these flows.

Searching arXiv for recent and foundational papers on coupled Ricci flow and closely related coupled geometric flows. Coupled Ricci flow denotes a family of geometric evolution systems in which Ricci flow is linked to additional dynamical fields, measures, submanifolds, or auxiliary evolution equations rather than acting on a metric alone. In the literature represented here, the coupling takes several distinct forms: Ricci flow coupled with harmonic map flow (Williams, 2010), Ricci flow coupled with a scalar heat equation (Guo et al., 2015), modified Ricci flow for a metric and weighted volume (Colding et al., 2023), Ricci-type ambient flow coupled with mean curvature flow (Koike et al., 2017), and a Hamilton–Jacobi/Ricci-flow formulation intended to provide an external time parameter for canonical general relativity (Alzain, 2022). Across these settings, the common structural feature is that the metric evolution is no longer autonomous: curvature interacts with an additional geometric quantity, and the coupled system is designed so that this interaction preserves or reveals deeper analytic, variational, or rigidity structures.

1. General concept and principal variants

The most basic paradigm in this body of work is harmonic-Ricci flow, defined by

$$\partial_t g = -2\,\mathrm{Rc} + 2c\, \nabla \phi \otimes \nabla \phi, \qquad \partial_t \phi = \tau_{g,h}\phi,$$

where $g(t)$ is a time-dependent metric, $\phi(t)$ is a map into a target manifold, and $c=c(t)\ge 0$ is a non-increasing coupling function (Williams, 2010). The tensor

$$S = \mathrm{Rc} - c\, \nabla \phi \otimes \nabla \phi$$

is introduced so that the metric evolution becomes $\partial_t g = -2S$, with trace $s = \mathrm{scal} - c\,|\nabla \phi|^2$ (Williams, 2010). In this formulation, Ricci flow is modified by the energy density of the evolving map.

A scalar-field variant appears in the coupled Ricci flow

$$\begin{cases} \partial_t g_{ij} = -2\,\mathrm{Ric}_{ij} + 2\,\nabla_i\phi\,\nabla_j\phi,\ \partial_t \phi = \Delta_g \phi, \end{cases}$$

with coupling constant normalized to $1$ (Guo et al., 2015). There the modified tensor

$$Sic_{g,\phi} := \mathrm{Ric}_g - d\phi\otimes d\phi$$

and its trace

$g(t)$0

play the role ordinarily occupied by Ricci curvature and scalar curvature in reduced-geometry arguments (Guo et al., 2015).

Another important class is the modified Ricci flow for a metric and a weight function,

$g(t)$1

where $g(t)$2 is the scalar curvature of $g(t)$3 (Colding et al., 2023). This is a coupled evolution of the metric and weighted volume density $g(t)$4, and the flow preserves the weighted measure in the sense that

$g(t)$5

(Colding et al., 2023). In this setting, a static solution is a gradient shrinking Ricci soliton,

$g(t)$6

so the coupled parabolic system is explicitly tied to the elliptic shrinker equation (Colding et al., 2023).

The phrase also extends beyond metric-plus-field systems. In Ricci-mean curvature flow, the ambient metric evolves by

$g(t)$7

while an immersion evolves by

$g(t)$8

so the ambient Ricci-type flow and mean curvature flow form a coupled system (Koike et al., 2017). In canonical gravity, the coupling can instead be between Ricci flow and the Hamilton–Jacobi equation: the proposal is to evolve a gravitational Hamilton–Jacobi functional $g(t)$9 by scalar curvature,

$\phi(t)$0

thereby introducing an external flow parameter $\phi(t)$1 into canonical general relativity (Alzain, 2022).

These examples show that “coupled Ricci flow” is not a single standard equation but a category of systems in which Ricci evolution interacts with another geometric structure.

2. Harmonic-Ricci flow and extended Ricci-flow systems

The harmonic-Ricci system is one of the foundational coupled models. It is studied both abstractly and as a dimensional reduction of Ricci flow on certain bundle geometries (Williams, 2010). In a flat $\phi(t)$2-vector bundle with volume-preserving flat connection, Ricci flow on the total space reduces to harmonic-Ricci flow on the base with target

$\phi(t)$3

and coupling constant $\phi(t)$4 (Williams, 2010). This places the coupled system within a concrete geometric construction rather than treating it as an ad hoc perturbation.

The same paper defines gradient harmonic-Ricci solitons by

$\phi(t)$5

and harmonic-Einstein pairs by

$\phi(t)$6

(Williams, 2010). A rigidity statement asserts that if $\phi(t)$7 is closed and $\phi(t)$8 is a steady or expanding gradient harmonic-Ricci soliton, then $\phi(t)$9 must actually be harmonic-Einstein (Williams, 2010).

Stability theory for such systems is developed in the framework of extended Ricci flow systems (Williams, 2013). There the normalized harmonic-Ricci-DeTurck flow is written as

$c=c(t)\ge 0$0

with stationary solutions

$c=c(t)\ge 0$1

(Williams, 2013). The linearization at such a fixed point is

$c=c(t)\ge 0$2

so the metric component is governed by the Lichnerowicz Laplacian while the map component is weakly stable because constant variations lie in the kernel (Williams, 2013). Under the assumption that $c=c(t)\ge 0$3 is a strictly linearly stable Einstein metric, nearby solutions exist for all time and converge exponentially fast to a stationary solution $c=c(t)\ge 0$4, where $c=c(t)\ge 0$5 is constant (Williams, 2013).

The same stability framework encompasses more elaborate couplings. For the locally $c=c(t)\ge 0$6-invariant Ricci flow, the reduced system couples a base metric, connection $c=c(t)\ge 0$7, and fiber metric $c=c(t)\ge 0$8: $c=c(t)\ge 0$9

$$S = \mathrm{Rc} - c\, \nabla \phi \otimes \nabla \phi$$0

$$S = \mathrm{Rc} - c\, \nabla \phi \otimes \nabla \phi$$1

(Williams, 2013). Another example is the connection Ricci flow

$$S = \mathrm{Rc} - c\, \nabla \phi \otimes \nabla \phi$$2

where $$S = \mathrm{Rc} - c\, \nabla \phi \otimes \nabla \phi$$3 is the torsion 3-form of a metric-compatible connection (Williams, 2013). These are all instances of Ricci flow coupled with an auxiliary geometric evolution, studied by a common stability methodology based on normalization, the DeTurck trick, linearization, and Simonett’s theorem (Williams, 2013).

3. Analytic structures: weighted measures, heat equations, eigenvalues, and entropy

A recurrent feature of coupled Ricci flows is the emergence of natural weighted measures and modified elliptic or parabolic operators. In modified Ricci flow, the preserved measure

$$S = \mathrm{Rc} - c\, \nabla \phi \otimes \nabla \phi$$4

makes the drift Laplacian

$$S = \mathrm{Rc} - c\, \nabla \phi \otimes \nabla \phi$$5

the natural spectral operator, self-adjoint on $$S = \mathrm{Rc} - c\, \nabla \phi \otimes \nabla \phi$$6 (Colding et al., 2023). The associated eigenvalue problem is

$$S = \mathrm{Rc} - c\, \nabla \phi \otimes \nabla \phi$$7

(Colding et al., 2023). A principal result is the differential inequality

$$S = \mathrm{Rc} - c\, \nabla \phi \otimes \nabla \phi$$8

for every $$S = \mathrm{Rc} - c\, \nabla \phi \otimes \nabla \phi$$9, in the sense of upper forward difference quotients (Colding et al., 2023). When equality persists at the threshold value $\partial_t g = -2S$0, the manifold splits isometrically as

$\partial_t g = -2S$1

with corresponding splitting formulas for $\partial_t g = -2S$2 and $\partial_t g = -2S$3 (Colding et al., 2023). This gives a coupled spectral-rigidity theory paralleling the static shrinker case.

A closely related weighted perspective appears in conformal Ricci flow. On a closed, connected Riemannian manifold of dimension $\partial_t g = -2S$4 with preserved scalar curvature

$\partial_t g = -2S$5

the conformal Ricci flow is

$\partial_t g = -2S$6

where $\partial_t g = -2S$7 is the conformal pressure (Abolarinwa et al., 2023). Using the conjugate heat kernel and weighted measure

$\partial_t g = -2S$8

the drifting Laplacian

$\partial_t g = -2S$9

and Bakry–Émery tensor

$s = \mathrm{scal} - c\,|\nabla \phi|^2$0

are introduced (Abolarinwa et al., 2023). The parabolic frequency functional

$s = \mathrm{scal} - c\,|\nabla \phi|^2$1

is monotone under a Bakry–Émery Ricci curvature bound, with the sign depending on the sign of $s = \mathrm{scal} - c\,|\nabla \phi|^2$2 (Abolarinwa et al., 2023). Equality implies that $s = \mathrm{scal} - c\,|\nabla \phi|^2$3 is an eigenfunction of the drifting Laplacian: $s = \mathrm{scal} - c\,|\nabla \phi|^2$4 (Abolarinwa et al., 2023). The same monotonicity yields backward uniqueness and monotonicity of the first nonzero eigenvalue of $s = \mathrm{scal} - c\,|\nabla \phi|^2$5 (Abolarinwa et al., 2023).

Entropy methods are equally central in harmonic-Ricci flow. The $s = \mathrm{scal} - c\,|\nabla \phi|^2$6-functional

$s = \mathrm{scal} - c\,|\nabla \phi|^2$7

satisfies

$s = \mathrm{scal} - c\,|\nabla \phi|^2$8

so $s = \mathrm{scal} - c\,|\nabla \phi|^2$9 is nondecreasing, with equality characterized by the coupled gradient soliton conditions (Li, 2010). The expanding entropy

$$\begin{cases} \partial_t g_{ij} = -2\,\mathrm{Ric}_{ij} + 2\,\nabla_i\phi\,\nabla_j\phi,\ \partial_t \phi = \Delta_g \phi, \end{cases}$$0

also has a nonnegative first variation, again with equality corresponding to the expanding gradient harmonic-Ricci soliton equations (Li, 2010).

The spectral theory of the ordinary Laplacian is modified as well. If $$\begin{cases} \partial_t g_{ij} = -2\,\mathrm{Ric}_{ij} + 2\,\nabla_i\phi\,\nabla_j\phi,\ \partial_t \phi = \Delta_g \phi, \end{cases}$$1 is an eigenvalue of $$\begin{cases} \partial_t g_{ij} = -2\,\mathrm{Ric}_{ij} + 2\,\nabla_i\phi\,\nabla_j\phi,\ \partial_t \phi = \Delta_g \phi, \end{cases}$$2, then under harmonic-Ricci flow

$$\begin{cases} \partial_t g_{ij} = -2\,\mathrm{Ric}_{ij} + 2\,\nabla_i\phi\,\nabla_j\phi,\ \partial_t \phi = \Delta_g \phi, \end{cases}$$3

(Li, 2010). Here the effective curvature quantity is

$$\begin{cases} \partial_t g_{ij} = -2\,\mathrm{Ric}_{ij} + 2\,\nabla_i\phi\,\nabla_j\phi,\ \partial_t \phi = \Delta_g \phi, \end{cases}$$4

rather than scalar curvature alone (Li, 2010).

4. Singularities, compactness, and pseudo-locality

Coupled Ricci flows admit analogues of central Ricci-flow singularity theory. For the Ricci flow coupled to a scalar heat equation, a version of Perelman’s pseudo-locality theorem is proved (Guo et al., 2015). Under local assumptions

$$\begin{cases} \partial_t g_{ij} = -2\,\mathrm{Ric}_{ij} + 2\,\nabla_i\phi\,\nabla_j\phi,\ \partial_t \phi = \Delta_g \phi, \end{cases}$$5

an almost-Euclidean isoperimetric inequality on subsets of the ball, and an initial bound $$\begin{cases} \partial_t g_{ij} = -2\,\mathrm{Ric}_{ij} + 2\,\nabla_i\phi\,\nabla_j\phi,\ \partial_t \phi = \Delta_g \phi, \end{cases}$$6, one obtains the local curvature estimate

$$\begin{cases} \partial_t g_{ij} = -2\,\mathrm{Ric}_{ij} + 2\,\nabla_i\phi\,\nabla_j\phi,\ \partial_t \phi = \Delta_g \phi, \end{cases}$$7

for points satisfying

$$\begin{cases} \partial_t g_{ij} = -2\,\mathrm{Ric}_{ij} + 2\,\nabla_i\phi\,\nabla_j\phi,\ \partial_t \phi = \Delta_g \phi, \end{cases}$$8

(Guo et al., 2015). The reduced geometry is built from

$$\begin{cases} \partial_t g_{ij} = -2\,\mathrm{Ric}_{ij} + 2\,\nabla_i\phi\,\nabla_j\phi,\ \partial_t \phi = \Delta_g \phi, \end{cases}$$9

the reduced distance $1$0, and reduced volume $1$1, which remains monotone nonincreasing (Guo et al., 2015).

This pseudo-locality theorem is then used in Type I blow-up analysis. If the solution is complete, $1$2-noncollapsed, and has a Type I singularity at time $1$3, then after parabolic rescaling

$1$4

a subsequence converges to a complete ancient limit that is a gradient shrinking soliton for the coupled flow (Guo et al., 2015). If the base point is a Type I singularity point, the limit soliton is nontrivial (Guo et al., 2015). The paper further notes that in the blow-up limit $1$5 becomes constant, so the limiting soliton reduces to an ordinary gradient shrinking Ricci soliton (Guo et al., 2015).

Compactness theory for harmonic-Ricci flow is established in a Hamilton-style form (Williams, 2010). If a sequence of complete pointed solutions has uniformly bounded curvature, a uniform lower bound on injectivity radius at basepoints, compact target manifold, and non-increasing coupling function, then a subsequence converges to a complete pointed harmonic-Ricci flow solution (Williams, 2010). This is extended to pointed étale Riemannian groupoids, providing the collapse-theoretic counterpart of the manifold compactness theorem (Williams, 2010).

A different singularity viewpoint appears in weak super Ricci flow through neckpinch (Lakzian et al., 2020). There a Ricci flow metric measure spacetime is defined as a pseudo-metric measure spacetime of product form, with singular sets retained in the spacetime and the regular set satisfying

$1$6

(Lakzian et al., 2020). Weak super Ricci flow is defined through a dynamic optimal transport contraction property for diffusions on maximal diffusion components: $1$7 for costs $1$8 with $1$9 increasing and convex (Lakzian et al., 2020). The main theorem states that the spacetime is a weak super Ricci flow for such costs if and only if the singularity exhibits the single-point pinching phenomenon (Lakzian et al., 2020). This suggests that, in a metric-measure formulation, coupled contraction of diffusions can detect geometric features of singular set formation.

5. Coupling with submanifold flows and space-time geometry

In several works, Ricci-type ambient evolution is coupled not to a field on the manifold but to a submanifold flow. The Ricci-mean curvature flow consists of the ambient evolution

$$Sic_{g,\phi} := \mathrm{Ric}_g - d\phi\otimes d\phi$$0

together with

$$Sic_{g,\phi} := \mathrm{Ric}_g - d\phi\otimes d\phi$$1

where $$Sic_{g,\phi} := \mathrm{Ric}_g - d\phi\otimes d\phi$$2 is an immersion of fixed codimension (Koike et al., 2017). The natural associated object is the normal Gauss map

$$Sic_{g,\phi} := \mathrm{Ric}_g - d\phi\otimes d\phi$$3

valued in the Grassmann bundle with Sasaki metric (Koike et al., 2017). Under the coupled flow, the Gauss map satisfies

$$Sic_{g,\phi} := \mathrm{Ric}_g - d\phi\otimes d\phi$$4

where $$Sic_{g,\phi} := \mathrm{Ric}_g - d\phi\otimes d\phi$$5 is the vertical part of the tension field and $$Sic_{g,\phi} := \mathrm{Ric}_g - d\phi\otimes d\phi$$6 is a vertical vector field built from ambient curvature (Koike et al., 2017). In codimension one, $$Sic_{g,\phi} := \mathrm{Ric}_g - d\phi\otimes d\phi$$7, so

$$Sic_{g,\phi} := \mathrm{Ric}_g - d\phi\otimes d\phi$$8

the vertically harmonic map heat flow (Koike et al., 2017). This generalizes the Euclidean result of Wang and the classical Ruh–Vilms theorem to evolving ambient manifolds.

A different but related framework is the coupling of totally real submanifold flow to Ricci-type ambient flow (Lotay et al., 2014). The submanifold evolution is the Maslov flow

$$Sic_{g,\phi} := \mathrm{Ric}_g - d\phi\otimes d\phi$$9

while the ambient almost Kähler structure evolves by Streets–Tian symplectic curvature flow (Lotay et al., 2014). The key coupled identity is

$g(t)$00

which follows from the cancellation between the ambient evolution and the contraction of $g(t)$01 with the Maslov-flow vector field (Lotay et al., 2014). In the Kähler setting this recovers preservation of the Lagrangian condition under simultaneous mean curvature flow and Kähler–Ricci flow; more generally it preserves the pullback 2-form for totally real submanifolds (Lotay et al., 2014).

The space-time construction of mean curvature flow in a Ricci flow background gives a further geometric synthesis (Helmensdorfer, 2012). For a hypersurface evolving by

$g(t)$02

inside an ambient Ricci flow, the space-time track

$g(t)$03

inside a canonical Ricci soliton metric becomes an approximate soliton of the coupled flow (Helmensdorfer, 2012). In the expanding case, the ambient canonical Ricci soliton metric is

$g(t)$04

with potential $g(t)$05 (Helmensdorfer, 2012). The second fundamental form of the canonical hypersurface converges to the mean curvature flow Harnack quantity, while the ambient Ricci curvature converges to Hamilton’s Ricci-flow Harnack quantity (Helmensdorfer, 2012). The same geometry matches the boundary term in Lott’s modified $g(t)$06-functional for Ricci flow with boundary (Helmensdorfer, 2012).

An even more elaborate recent variant studies mean curvature flow in an ambient manifold evolving by Ricci flow coupled with harmonic map heat flow (Gomes et al., 27 Oct 2025). The ambient evolution is

$g(t)$07

and the weighted functional

$g(t)$08

is shown to be monotone under a modified flow preserving the weighted measure $g(t)$09 (Gomes et al., 27 Oct 2025). The paper also derives a Huisken-type monotonicity formula for weighted area in self-similar $g(t)$10 backgrounds, with constancy exactly on mean curvature solitons (Gomes et al., 27 Oct 2025).

6. Conceptual reinterpretations and discrete analogues

One line of work uses coupled Ricci flow as a conceptual device for canonical gravity rather than as an analytic evolution on a fixed geometric category. In the ADM $g(t)$11 formulation, the Hamiltonian and momentum constraints are expressed in terms of the three-metric $g(t)$12 and conjugate data, and Peres’ observation yields a Hamilton–Jacobi functional $g(t)$13 satisfying the gravitational Hamilton–Jacobi equation (Alzain, 2022). The proposal is not to evolve the metric directly by Ricci flow, but to evolve the scalar functional $g(t)$14 by scalar curvature: $g(t)$15 Assuming $g(t)$16 also satisfies the gravitational Hamilton–Jacobi equation, one obtains a modified equation with explicit $g(t)$17-dependence (Alzain, 2022). Using the ansatz

$g(t)$18

the equation separates into a Hamilton–Jacobi-type equation for $g(t)$19 and a simple evolution equation for $g(t)$20 (Alzain, 2022). The intended significance is that Ricci flow supplies an external evolution parameter for canonical general relativity, thereby addressing the “problem of time” in quantum gravity (Alzain, 2022). A direct Einstein-Ricci parabolic tensor system is explicitly rejected in the same work because it leads to a backward heat equation for scalar curvature and therefore lacks short-time solutions (Alzain, 2022).

A distinct but related development appears in coupled Kähler-Einstein geometry through the coupled Ricci iteration (Takahashi, 2019). Given a decomposition

$g(t)$21

a $g(t)$22-tuple $g(t)$23 is a coupled Kähler-Einstein metric if

$g(t)$24

(Takahashi, 2019). The discrete dynamical system is

$g(t)$25

so each step solves a twisted Kähler-Einstein equation (Takahashi, 2019). The iteration is interpreted as a Ricci-flow-inspired discrete analogue for the coupled Einstein problem, with the Ding functional serving as a Lyapunov functional (Takahashi, 2019). In the negative first Chern class case it converges smoothly to the unique coupled Kähler-Einstein metric; in the positive case, existence is characterized by $g(t)$26-coercivity of the Ding functional (Takahashi, 2019). Although this is not a continuous coupled Ricci flow, it occupies a closely allied conceptual position.

A possible misconception is that “coupled Ricci flow” always refers to the harmonic-Ricci system or to a single canonical PDE. The literature represented here suggests otherwise. The term covers continuous metric-map systems (Williams, 2010), metric-scalar systems (Guo et al., 2015), metric-weight systems (Colding et al., 2023), ambient-submanifold systems (Koike et al., 2017), variational or canonical reformulations in gravity (Alzain, 2022), and discrete Ricci-inspired iterations in Kähler geometry (Takahashi, 2019). A plausible implication is that the unifying content is structural rather than terminological: Ricci evolution is coupled whenever the curvature-driven deformation law is closed only after adjoining an additional geometric variable.

7. Significance, recurring themes, and scope

Several themes recur across these disparate formulations. One is the replacement of ordinary curvature by an effective curvature tensor incorporating the coupled field, such as

$g(t)$27

in Ricci-harmonic map flow (Băileşteanu, 2013), or

$g(t)$28

in harmonic-Ricci flow (Li, 2010). Another is the appearance of weighted measures preserved by the flow, as in modified Ricci flow (Colding et al., 2023) and in weighted functionals for Ricci–harmonic map backgrounds with boundary (Gomes et al., 27 Oct 2025). A third is the extension of Ricci-flow tools—compactness, reduced volume, entropy monotonicity, Harnack quantities, eigenvalue control, pseudo-locality, and splitting theorems—to the coupled setting [(Guo et al., 2015); (Williams, 2010); (Li, 2010); (Colding et al., 2023)].

Coupled systems also arise naturally from geometric reduction. Harmonic-Ricci flow appears as a reduction of Ricci flow on locally $g(t)$29-invariant bundles (Williams, 2010), while multiply warped product Ricci flow becomes a coupled system for the base metric and warping functions (Williams, 2013). Conversely, Ricci-type ambient flow can organize the dynamics of submanifolds, Gauss maps, and canonical space-time tracks [(Koike et al., 2017); (Helmensdorfer, 2012); (Lotay et al., 2014)].

At the same time, the coupling can sharpen rigidity. In modified Ricci flow, persistence of the critical eigenvalue $g(t)$30 forces isometric splitting and Gaussian weight on Euclidean directions (Colding et al., 2023). In harmonic-Ricci flow, monotonicity of entropy and eigenvalues rules out nontrivial compact steady and expanding breathers except in soliton or Einstein-constant-map cases (Li, 2010). In coupled Ricci flow with a scalar field, Type I blow-ups still converge to ordinary gradient shrinking Ricci solitons because the scalar field becomes asymptotically constant (Guo et al., 2015).

The scope of the subject therefore extends from PDE and geometric analysis to singularity theory, calibrated and symplectic geometry, submanifold geometry, optimal transport, and mathematical relativity. “Coupled Ricci flow” names not a single equation but a research program: the systematic study of curvature evolution when the geometry interacts with an additional field, measure, or flow in a way that is analytically nontrivial and geometrically natural [(Williams, 2010); (Guo et al., 2015); (Colding et al., 2023); (Koike et al., 2017); (Alzain, 2022)].