---
title: Homogeneous Kenmotsu Manifolds
url: https://www.emergentmind.com/topics/homogeneous-kenmotsu-manifold
type: topic
---

# Homogeneous Kenmotsu Manifolds

A homogeneous Kenmotsu manifold is most concretely realized in the Lie-group setting: a Lie group endowed with a left invariant Kenmotsu structure is homogeneous by construction, since left translations act transitively and preserve the metric and the almost contact data. Within the broader theory, Kenmotsu geometry is distinguished by a rigid almost contact metric structure, a local warped-product model \(I\times_f N\) with \(f(t)=ke^t\) or \(f(t)=ce^t\) and \(N\) Kähler, and a strong tendency toward Einstein or hyperbolic-type behavior under additional curvature or soliton hypotheses. The modern literature therefore treats homogeneous Kenmotsu manifolds through two closely related mechanisms: explicit classification of left invariant Kenmotsu structures on Lie groups, and local-homogeneity results for related almost \(\alpha\)-Kenmotsu geometries via canonical connections and \(D\)-homothetic deformation [2407.16506] [1812.03422] [1006.4732].

## 1. Structural framework of Kenmotsu geometry

A classical Kenmotsu manifold is a \((2n+1)\)-dimensional almost contact metric manifold \((M,\phi,\xi,\eta,g)\) satisfying
\[
\phi^2=-I+\eta\otimes\xi,\qquad \eta(\xi)=1,
\]
\[
g(\phi X,\phi Y)=g(X,Y)-\eta(X)\eta(Y),
\]
and the Kenmotsu condition
\[
(\nabla_X\phi)Y=g(\phi X,Y)\xi-\eta(Y)\phi X.
\]
Its characteristic vector field obeys the standard identities
\[
\nabla_X\xi=X-\eta(X)\xi,\qquad R(X,Y)\xi=\eta(X)Y-\eta(Y)X,
\]
together with
\[
Q\xi=-2n\,\xi
\]
or equivalently
\[
S(X,\xi)=-2n\,\eta(X),
\]
and
\[
(\mathcal L_\xi g)(X,Y)=2\{g(X,Y)-\eta(X)\eta(Y)\}.
\]
In particular, \(\xi\) is never Killing on a Kenmotsu manifold [1901.05222] [1812.03422].

The local geometric model is a warped product
\[
I\times_f N^{2n},\qquad f=ke^t \ \text{or}\ f=ce^t,
\]
where \(N\) is Kähler. This warped-product description is not merely a coordinate convenience: it explains why the Reeb direction is geometrically distinguished, why the transverse distribution behaves like a Kähler factor, and why many rigidity arguments reduce to ODEs along the \(t\)-direction [2407.16506] [1812.03422].

For homogeneous questions, these identities have immediate algebraic consequences. The Reeb field is dynamically tied to the metric by \(\nabla_X\xi=X-\eta(X)\xi\), and the curvature already contains a distinguished \(\xi\)-component. This suggests that any transitive symmetry preserving the Kenmotsu structure must interact very tightly with the splitting into the Reeb line and the contact distribution.

## 2. Homogeneity through left invariant Kenmotsu structures

The explicit homogeneous theory begins with Kenmotsu Lie groups. If \(G\) is a Lie group with left invariant metric \(g\), left invariant tensor \(\phi\), and left invariant characteristic vector field \(\xi\), then the entire geometry is determined at the Lie-algebra level. Let \(\mathfrak g\) be the Lie algebra, identified with left invariant vector fields, and let
\[
L:\mathfrak g\times\mathfrak g\to\mathfrak g
\]
be the Levi-Civita product defined by
\[
2\langle L_XY,Z\rangle=\langle [X,Y],Z\rangle+\langle [Z,X],Y\rangle+\langle [Z,Y],X\rangle.
\]
It satisfies
\[
L_X+L_X^{\,t}=0,\qquad L_XY-L_YX=[X,Y].
\]
In this setting the Kenmotsu structure becomes a finite-dimensional algebraic problem on \(\mathfrak g\) [2407.16506].

From the left invariant Kenmotsu identities one obtains
\[
\phi(\xi)=0,\qquad \eta\circ\phi=0,
\]
\[
L_X\xi=X-\eta(X)\xi,
\]
\[
\langle \xi,L_XY\rangle=-\langle X,Y\rangle+\eta(X)\eta(Y),
\]
and, crucially,
\[
d\eta=0.
\]
The condition \(d\eta=0\) forces
\[
\mathfrak h=\ker\eta
\]
to be an ideal of \(\mathfrak g\), so that
\[
\mathfrak g=\mathfrak h\oplus\mathbb R\xi.
\]
Thus homogeneous Kenmotsu geometry acquires a canonical codimension-one ideal \(\mathfrak h\), transverse to the Reeb direction [2407.16506].

The restriction \(J=\phi|_{\mathfrak h}\) defines a complex structure on \(\mathfrak h\), and the induced metric is Hermitian. The resulting triple \((\mathfrak h,\langle\cdot,\cdot\rangle|_{\mathfrak h},J)\) is a Kähler Lie algebra. If
\[
D=\operatorname{ad}_\xi|_{\mathfrak h},
\]
then \(D\) is an invertible derivation satisfying
\[
[D,J]=0,\qquad D+D'=-2\,\mathrm{Id}_{\mathfrak h}.
\]
This is the core algebraic reduction: a homogeneous Kenmotsu structure is encoded by a Kähler ideal together with a derivation commuting with the complex structure and having prescribed symmetric part [2407.16506].

## 3. Classification of connected simply-connected Kenmotsu Lie groups

The classification theorem for connected and simply-connected Kenmotsu Lie groups states that for every such \((G,\phi,\xi,g)\) of dimension \(2n+1\), there exists \((\lambda_1,\dots,\lambda_n)\in\mathbb R^n\) such that \((G,\phi,\xi,g)\) is isomorphic to
\[
(\mathbb R\times \mathbb C^n,\phi_0,\xi_0,g_0),
\]
where the group law is
\[
(t_1,(z_1,\dots,z_n))\cdot (t_2,(w_1,\dots,w_n))
=
\left(t_1+t_2,\left(z_1+w_1e^{-t_1(1+i\lambda_1)},\dots,z_n+w_ne^{-t_1(1+i\lambda_n)}\right)\right),
\]
the characteristic field is
\[
\xi_0=\partial_t,
\]
\(\phi_0\) is the standard complex structure on \(\mathbb C^n\) with \(\phi_0(\partial_t)=0\), and the metric is
\[
g_0=dt^2+e^{2t}\sum_{j=1}^n(dx_j^2+dy_j^2).
\]
This gives a complete normal form for homogeneous Kenmotsu manifolds in the connected simply-connected Lie-group category [2407.16506].

At the Lie-algebra level, the decisive linear algebra lemma asserts that if \(V\) is a real \(2n\)-dimensional inner product space, \(J^2=-\mathrm{Id}\), \([J,D]=0\), and \(D+D'=-2\mathrm{Id}\), then there are real numbers \(\lambda_1,\dots,\lambda_n\) and an orthonormal basis
\[
(e_1,Je_1,\dots,e_n,Je_n)
\]
such that
\[
De_j=-e_j-\lambda_jJe_j,\qquad D(Je_j)=-Je_j+\lambda_j e_j.
\]
Consequently, the nonzero Lie brackets of the model can be written as
\[
[e_0,e_j]=-e_j-\lambda_j f_j,\qquad [e_0,f_j]=-f_j+\lambda_j e_j,
\]
with all brackets among the \(e_j,f_j\) vanishing [2407.16506].

Several structural consequences follow. Every connected simply-connected Kenmotsu Lie group is solvable, completely determined by the parameter \(n\)-tuple \((\lambda_1,\dots,\lambda_n)\), and has Lie algebra of semidirect-product form
\[
\mathfrak g=\mathbb R\xi\ltimes \mathfrak h.
\]
Moreover, no non-abelian Kähler ideal occurs in this classification: \(\mathfrak h\) is abelian. The homogeneous Kenmotsu problem is therefore rigid enough to collapse to a family of explicit solvable Lie groups of exponential warped-product type, with the parameters \(\lambda_j\) encoding the rotational part of the derivation on the complex directions [2407.16506].

## 4. Local homogeneous models beyond the strict Kenmotsu case

A broader local-homogeneity theory appears in almost \(\alpha\)-Kenmotsu geometry. An almost \(\alpha\)-Kenmotsu manifold \((M^{2n+1},\varphi,\xi,\eta,g)\) is defined by
\[
d\eta=0,\qquad d\Phi=2\alpha\,\eta\wedge\Phi,
\]
with \(\alpha\neq 0\), and involves the tensor
\[
h'=-\frac1{2\alpha}(\mathcal L_\xi\varphi)\circ\varphi.
\]
The deformation
\[
\bar\eta=\beta\,\eta,\qquad \bar\xi=\beta^{-1}\xi,\qquad \bar g=\beta\, g+\beta(\beta-1)\,\eta\otimes\eta
\]
transforms an almost \(\alpha\)-Kenmotsu structure into an almost \(\alpha/\beta\)-Kenmotsu one; by choosing \(\beta=\alpha\), one passes to an almost Kenmotsu structure. Under this \(D\)-homothetic deformation, the tensor \(h'\), the condition that \(\varphi\) is \(\mathcal D\)-parallel, the condition \(\nabla_\xi h'=0\), and the canonical connection are preserved [1006.4732].

The pivotal result is the canonical-connection characterization of CR-integrability. A manifold is a CR-integrable almost \(\alpha\)-Kenmotsu manifold if and only if there exists a linear connection \(\widetilde\nabla\) such that \(\varphi,\xi,\eta,g\) are \(\widetilde\nabla\)-parallel and the torsion satisfies
\[
T(X,Y)=0\quad (X,Y\in\mathcal D),\qquad 2T(\xi,X)=\alpha(X+h'X),
\]
with \(T_\xi\) selfadjoint, and
\[
\widetilde\nabla_XY=\nabla_XY+\alpha\, g(X+h'X,Y)\xi-\alpha\,\eta(Y)(X+h'X).
\]
This canonical connection is invariant under \(D\)-homothetic deformations [1006.4732].

If the canonical connection has parallel torsion and curvature, then the local geometry, up to \(D\)-homothetic deformations, is completely determined by \(\dim M\) and the spectrum of \(h'\). If
\[
\operatorname{Spec}(h')=\{0,\lambda_1,-\lambda_1,\dots,\lambda_r,-\lambda_r\},\qquad \lambda_i>0,
\]
then the manifold is locally a warped product
\[
M' \times_{f_0} M_0 \times_{f_1} M_{\lambda_1}\times_{g_1} M_{-\lambda_1}\times\cdots\times_{f_r} M_{\lambda_r}\times_{g_r} M_{-\lambda_r},
\]
with
\[
f_0=c_0 e^{\alpha t},\qquad f_i=c_i e^{\alpha(1+\lambda_i)t},\qquad g_i=c_i' e^{\alpha(1-\lambda_i)t}.
\]
In particular, the manifold is locally equivalent to a solvable non-nilpotent Lie group, which is a subgroup of \(\operatorname{Aff}(2n+1,\mathbb R)\), endowed with a left invariant almost \(\alpha\)-Kenmotsu structure [1006.4732].

This gives a precise local homogeneity statement in a Kenmotsu-type category adjacent to the classical one. A plausible implication is that homogeneous phenomena in classical Kenmotsu geometry are naturally studied together with \(D\)-homothetic deformations, canonical connections, and the spectral data of \(h'\).

## 5. Einstein condition, curvature rigidity, and hyperbolic models

The homogeneous Lie-group classification has a strong curvature conclusion: every connected simply-connected Kenmotsu Lie group is Einstein, with
\[
\operatorname{ric}=-2n\,g.
\]
Thus homogeneous Kenmotsu manifolds in the Lie-group sense automatically carry negative Einstein metrics [2407.16506].

Additional hypotheses can force an even sharper curvature collapse. If a complete Kenmotsu metric satisfies the critical point equation, then the metric is Einstein, the manifold is locally isometric to hyperbolic space \(H^{2n+1}\), and the potential function is locally
\[
\lambda=A\cosh t + B\sinh t,
\]
where \(t\) is the warped-product coordinate. The proof reduces the equation to
\[
\nabla_X D\lambda=\lambda X,
\]
after showing \(QX=-2nX\), and then invokes Tashiro’s theorem [1812.03422].

A comparable rigidity appears for \(\ast\)-Ricci solitons. On a Kenmotsu 3-manifold, if the metric is a \(\ast\)-Ricci soliton, then the Ricci operator becomes
\[
QX=-2X,
\]
and hence
\[
R(X,Y)Z=-\{g(Y,Z)X-g(X,Z)Y\},
\]
so the manifold has constant sectional curvature \(-1\). In the warped-product model
\[
M=\mathbb R\times_f K,\qquad f(t)=ce^t,
\]
this yields local isometry to \(\mathbb H^3(-1)\) [1901.05222].

Riemann-soliton rigidity leads to the same curvature value in dimension \(3\). If \(M\) is a 3-dimensional Kenmotsu manifold and \((g,V)\) is a Riemann soliton with \(\operatorname{div}V\) constant, then \(M\) has constant negative curvature \(-1\). More generally, for a gradient almost Riemann soliton on a Kenmotsu manifold, either \(M\) is Einstein or the soliton vector field is pointwise collinear with \(\xi\) on an open set; in the complete Einstein branch, \(M\) is locally isometric to \(\mathbb H^{2n+1}\) [2003.00786].

These results do not say that every homogeneous Kenmotsu manifold is hyperbolic. Rather, they show that homogeneous Kenmotsu manifolds become hyperbolic or constant-curvature models once one imposes additional variational or soliton constraints. Without those extra assumptions, the Lie-group classification retains the full \((\lambda_1,\dots,\lambda_n)\)-family.

## 6. Solitons, adjacent geometries, and terminological scope

The soliton literature reinforces the general picture that Kenmotsu geometry leaves little room for nontrivial self-similar behavior. For \(\ast\)-Ricci solitons, the soliton constant on a Kenmotsu manifold always satisfies
\[
\lambda=0,
\]
so every such soliton is steady. If the potential vector field is nonzero and pointwise collinear with the characteristic vector field, \(V=a\,\xi\), then the manifold is Einstein and in fact
\[
V=\xi.
\]
For a gradient almost \(\ast\)-Ricci soliton, either the manifold is Einstein or
\[
\operatorname{grad} f=\xi(f)\,\xi
\]
on an open set. These conclusions apply a fortiori to any homogeneous example satisfying the same hypotheses [1901.05222].

A similar rigidity holds for Riemann-type solitons. On an \(\eta\)-Einstein Kenmotsu manifold of dimension \(>3\), a Riemann soliton with \(\operatorname{div}V\) constant forces the Einstein value
\[
S=-2n\,g.
\]
For gradient almost Riemann solitons, one again obtains the Einstein-versus-Reeb-alignment dichotomy, and in the complete Einstein case the local model is hyperbolic [2003.00786].

The almost \(\ast\)-Ricci-Bourguignon theory points in the same direction. If a Kenmotsu manifold admits an almost \(\ast\)-Ricci-Bourguignon soliton, then it is \(\eta\)-Einstein. If it admits a gradient almost \(\ast\)-Ricci-Bourguignon soliton and the Reeb field preserves the scalar curvature, then the manifold is Einstein with
\[
r=-n(2n-1).
\]
The corresponding papers emphasize that they do not explicitly classify homogeneous Kenmotsu manifolds, but the stated rigidity theorems apply unchanged to homogeneous cases [2408.13288].

An important terminological caution concerns the word “homogeneous.” In the \(\ast\)-Ricci-Bourguignon literature, a soliton may be described as a “homogeneous solution” of the flow in the soliton sense; this is not the same as saying that the underlying Kenmotsu manifold is homogeneous in the Lie-group or transitive-symmetry sense. The same paper explicitly states that it does not develop a theory of homogeneous Kenmotsu manifolds as a separate geometric class [2601.15009].

Finally, nearby Kenmotsu-type geometries exhibit analogous homogeneous or rigidity phenomena. In \(LP\)-Kenmotsu geometry, constant scalar curvature implies
\[
r=n(n-1),
\]
hence Einsteinness, and for a Ricci-Yamabe soliton one has
\[
\lambda=\beta\,n(n-1)-\alpha(n-1).
\]
In the gradient case, constant scalar curvature forces the soliton to be trivial. The paper explicitly notes that this is particularly relevant for homogeneous Kenmotsu-type manifolds, because scalar curvature is often constant on homogeneous spaces [2301.02632].

Taken together, these results make the homogeneous Kenmotsu manifold a notably rigid object. In the explicit Lie-group setting it is classified by solvable models on \(\mathbb R\times\mathbb C^n\) and is automatically Einstein; in adjacent almost \(\alpha\)-Kenmotsu settings it arises locally from canonical-connection data and Lie-group models; and under CPE, \(\ast\)-Ricci, Riemann, or Ricci-Bourguignon type equations it frequently collapses to Einstein, \(\eta\)-Einstein, or hyperbolic geometry.

Source: https://www.emergentmind.com/topics/homogeneous-kenmotsu-manifold