1. Introduction
Ricci solitons play a crucial role in the study of Ricci flow, acting as self-similar solutions within this flow. Consequently, when investigating a geometric flow, it is essential to consider the solitons associated with that flow. Specifically, a family of metrics
$g(t)$
on a Riemannian manifold
$(M^n,g)$
is said to evolve according to the Ricci-Bourguignon flow (RB flow) if
$g(t)$
satisfies the following evolution equation:
In this context, Ric represents the Ricci tensor corresponding to the metric,
$r$
stands for the scalar curvature, and
$\rho \in \mathbb{R}$
. The flow equation in (1) was first introduced by Jean-Pierre Bourguignon [Reference Bourguignon1], building on earlier work by Lichnerowicz and a publication by Aubin [Reference Aubin2]. Notably, when
$\rho = 0$
, equation (1) reduces to the Ricci flow. The right-hand side of the evolution equation in (1) is significant for different values of
$\rho$
. The parabolic theory of the RB flow was systematically investigated in [Reference Catino, Cremaschi, Djadli, Mantegazza and Mazzieri3], which marked a significant advancement in the field. Among their key findings, the authors demonstrated the short-term existence of the flow (1) on a closed n-dimensional manifold, starting with any initial metric
$ g_0$
, for
$ \rho \lt \frac {1}{2(n-1)}$
.
The concept of almost Ricci solitons was introduced in [Reference Pigola, Rigoli, Rimoldi and Setti4], where the traditional notion of a Ricci soliton was redefined by allowing the parameter
$f$
, which was originally constant, to vary as a function. See [Reference Güler5–Reference Barros and Ribeiro7] for a more detailed understanding of almost Ricci solitons. These references offer comprehensive insights into the properties, applications, and recent advancements in the study of almost Ricci solitons, providing valuable resources for further research. Drawing inspiration from the Ricci flow framework, Dwivedi [Reference Dwivedi8] proceeded with the following considerations.
Definition 1. [Reference Dwivedi8] An almost Ricci-Bourguignon soliton (almost RB soliton for short) is a Riemannian manifold
$(M^n, g)$
endowed with a vector field
$V$
on
$M$
and a soliton function
$f\;:\; M\to \mathbb{R}$
that satisfies
where
$\unicode{x00A3}_{V} g$
denotes the Lie derivative of the metric
$g$
with respect to the vector field
$V$
,
$r$
denotes the scalar curvature of the almost RB soliton and
$\rho \in \mathbb{R}$
is a constant.
Again, if the vector field
$V$
is a gradient of some smooth function
$s$
on
$M$
, then it is called a gradient almost RB soliton. In this case, the foregoing equation reduces to
where
$\nabla ^{2} s$
stands for the Hessian of
$s$
. Note that if
$s$
is constant, the soliton becomes trivial (Einstein). Perelman [Reference Perelman9] proved that the Ricci solitons on a compact manifold are always gradient Ricci solitons, and, hence, the potential vector field can be expressed as the sum of the gradient of a function and a Killing vector field. Such a result is also true for the almost RB solitons, if the scalar curvature is constant (see [Reference Aqib, Shah and Patra10]). In the following discussion, we will refer to
$(M^n, g, V, f, \rho )$
as an almost RB-soliton, with
$V$
being the potential vector field. If
$f$
is constant, the almost RB soliton becomes an RB soliton. When the potential vector field
$V$
is given by the gradient of a smooth function
$s$
on
$M$
, the almost RB soliton is termed a gradient almost RB soliton. Depending on whether
$f$
is negative, zero, or positive, the soliton is classified as expanding, steady, or shrinking, respectively. If
$V$
is a Killing vector field, the almost RB soliton reduces to an RB soliton with
$f$
being constant, referred to as a trivial RB soliton. For further details on almost RB solitons; see [Reference Dwivedi8] and [Reference Ghosh11].
Thus, while the direct relation of almost RB solitons to the long-time behavior of the Ricci-Bourguignon flow remains an open question, their study on almost
$\alpha$
-cosymplectic manifolds is highly productive: it reveals classification patterns, clarifies the interplay between contact-like structures and curvature, and provides a systematic way to explore the geometric consequences of the parameter
$\rho$
.
In this article, we investigate RB solitons on almost
$\alpha$
-cosymplectic manifolds. RB solitons generalize quasi-Einstein metrics arising in string theory, while
$\alpha$
-cosymplectic manifolds extend symplectic geometry to odd dimensions. Let
$M$
be a
$(2n+1)$
-dimensional Riemannian manifold. An almost contact structure on
$M^{2n+1}$
is a triple
$(\varphi, \xi, \eta )$
, where
$\varphi$
is a
$(1,1)$
-tensor,
$\xi$
a global vector field, and
$\eta$
a 1-form such that
The vector field
$\xi$
is called the Reeb vector field. Here we use the metric to identify the Reeb vector field
$\xi$
with its dual
$1$
-form
$\eta$
that is
$\eta (.)=g(.,\xi )$
. If there exists a Riemannian metric
$g$
such that
for any
$X, Y \in \mathfrak{X}(M)$
, the metric
$g$
is said to be compatible with an almost contact structure, and
$(\varphi, \xi, \eta, g)$
is said to be an almost contact metric structure. An almost contact metric structure is said to be normal, if the almost complex structure
$J$
on
$M^{2 n+1} \times \mathbb{R}$
is integrable, where
$J$
is defined by
$J\left (X, f \frac {d}{d t}\right )=\left (\varphi X-f \xi, \eta (X) \frac {d}{d t}\right )$
for
$X \in \mathfrak{X}(M), t$
is the coordinate of
$\mathbb{R}$
and
$f$
a smooth function on
$M \times \mathbb{R}$
. Or equivalently, an almost contact metric structure is said to be normal if the Nijenhuis tensor of
$\varphi$
satisfies
$[\varphi, \varphi ]=-2 d \eta \otimes \xi$
, where the Nijenhuis tensor of
$\varphi$
is defined by
for any
$X, Y \in \mathfrak{X}(M)$
.
We denote a fundamental 2-form
$\Phi$
on an almost contact metric manifold
$M^{2 n+1}$
by
$\Phi (X, Y)=g(X, \varphi Y)$
for any
$X, Y \in \mathfrak{X}(M)$
. In general, the almost contact metric structure is called almost
$\alpha$
-cosymplectic, if
$d \eta =0$
and
$d \Phi =2 \alpha \eta \wedge \xi$
for any constant
$\alpha$
. Almost
$\alpha$
-cosymplectic structure is said to be
$\alpha$
- cosymplectic, if the almost contact metric structure is normal. In particular, if
$\alpha$
is non-zero, the almost contact metric structure is said to be almost
$\alpha$
-Kenmotsu. If
$\alpha =0$
, the almost contact metric structure is called almost cosymplectic. If
$\alpha =1$
, the almost contact metric structure is called almost Kenmotsu.
On an almost contact metric manifold, one defines a tensor field
$h$
by
$h=\frac {1}{2} \mathfrak{L}_{\xi } \varphi$
. On almost
$\alpha$
-cosymplectic
$(2 \textrm{n}+1)$
manifolds, we have the following properties from [Reference Perrone12] for the tensor
$h$
and
$\varphi$
:
and
In particular, an almost
$\alpha$
-cosymplectic three-manifold is
$\alpha$
-cosymplectic if and only if
$\mathfrak{L}_{\xi } h=0$
(see [Reference Perrone12]).
For an almost
$\alpha$
-cosymplectic three-manifold, according to the properties of tensor
$h$
, there exists a
$\varphi$
-basis
$\{\xi, e, \varphi e\}$
such that
where
$\lambda$
is a local smooth function. The Levi-Civita connection is given by (cf. [Reference Perrone12])
\begin{equation} \begin{cases} \nabla _{e} e=-\alpha \xi -a \varphi e, \quad \nabla _{e} \varphi e=\lambda \xi +a e, \quad \nabla _{e} \xi =\alpha e-\lambda \varphi e \\ \nabla _{\varphi e} e=\lambda \xi -b \varphi e, \quad \nabla _{\varphi e} \varphi e=-\alpha \xi +b e, \quad \nabla _{\varphi e} \xi =-\lambda e+\alpha \varphi e \\ \nabla _{\xi } e=\mu \varphi e, \quad \nabla _{\xi } \varphi e=-\mu e, \quad \nabla _{\xi } \xi =0, \end{cases} \end{equation}
where
$\lambda, \mu, a, b$
are local smooth functions. Then, the Lie brackets are given by
Then, using (9) in the Jacobi identity, we get
\begin{equation} \begin{cases} a\alpha -e(\lambda )+e(\mu )+\xi (a)+b\lambda -b\mu =0,\\ b\alpha -\varphi e(\lambda )+\varphi e(\mu )+\xi (b)+a\lambda +a\mu =0. \end{cases} \end{equation}
According to (8) and (9), the Ricci operator
$Q$
is given by (cf. [Reference Perrone12]):
\begin{equation} \begin{cases} Q e=-\left (2 \alpha ^{2}+a^{2}+b^{2}+2 \lambda \mu +(\varphi e)(a)-e(b)\right ) e+(2 \alpha \lambda +\xi (\lambda )) \varphi e-(2 a \lambda +(\varphi e)(\lambda )) \xi, \\ Q \varphi e=(2 \alpha \lambda +\xi (\lambda )) e-\left (2 \alpha ^{2}+a^{2}+b^{2}-2 \lambda \mu +(\varphi e)(a)-e(b)\right ) \varphi e+(2 b \lambda -e(\lambda )) \xi, \\ Q \xi =-(2 a \lambda +(\varphi e)(\lambda )) e+(2 b \lambda -e(\lambda )) \varphi e-\left (2 \alpha ^{2}+\operatorname {tr }h^{2}\right ) \xi . \end{cases} \end{equation}
Then, the scalar curvature
$r$
is expressed by
In what follows, we will use the important Remark
$3.3$
of [Reference Perrone12] for an almost
$\alpha$
-cosymplectic three-manifold: If
$\lambda$
is constant then
$\nabla _\xi h=2\mu \varphi h$
and the condition
$\nabla _\xi h=0$
is related to an interesting geometric property: it is equivalent to the property that the sectional curvatures of the planes
$P_1 = span(\xi, e)$
and
$P_2 = span(\xi, \varphi e)$
, at a given point, are equal.
On an almost
$\alpha$
-cosymplectic three-manifold
$(M, \varphi, \xi, \eta, g)$
, the canonical foliation
$\mathcal{F}$
of codimension one on
$M$
is defined by the contact distribution
$\operatorname {ker} \eta$
. We note that the extrinsic curvature
$K_{e}$
and Gaussian curvature
$K_{G}$
of the canonical foliation
$\mathcal{F}$
are given ([Reference Perrone12], p.73) by
Now we construct an
$\alpha$
-cosymplectic manifold which admits an almost RB soliton as follows.
Example 1.
We consider the
$5$
-dimensional manifold
$M=\mathbb{R}^{5}$
. Coordinatize
$\mathbb{R}^{5}$
by
$\{\left (x_{1}, x_{2}, y_{1}, y_{2}, z\right) \in \mathbb{R}^{5}\}$
, where
$\left (x_{1}, x_{2}, y_{1}, y_{2}, z\right)$
are the standard coordinates in
$\mathbb{R}^{5}$
. Let
$e_{1}, e_{2}, e_{3}, e_{4}$
and
$e_{5}$
be the vector fields on
$M$
given by
where
$\alpha$
is some constant. Let
$g$
be the Riemannian metric defined by
Let
$\eta$
be the 1-form on
$M$
defined by
$\eta (X)=g\left (X, e_{5}\right )=g(X, \xi )$
for all
$X \in \chi (M)$
. Let
$\varphi$
be the (1, 1) tensor field on
$M$
defined by
By applying the linearity of
$\varphi$
and
$g$
, we have
for all
$X, Y \in \chi (M)$
. Thus,
$(\mathbb{R}^5,g,\varphi, \xi, \eta )$
defines an almost contact structure. Then, we have
We can easily calculate,
\begin{equation*} \begin{gathered} \nabla _{e_{1}} e_{1}=-\alpha e_{5}, \quad \nabla _{e_{1}} e_{5}=\alpha e_{1},\quad \nabla _{e_{4}} e_{4}=-\alpha e_{5}, \\ \nabla _{e_{2}} e_{2}=-\alpha e_{5}, \quad \quad \nabla _{e_{2}} e_{5}=\alpha e_{2},\quad \nabla _{e_{4}} e_{5}=\alpha e_{4}, \\ \nabla _{e_{3}} e_{3}=-\alpha e_{5}, \quad \nabla _{e_{3}} e_{5}=\alpha e_{3}. \end{gathered}\end{equation*}
All the others are zero. It can be easily verified that the manifold
$M$
is an
$\alpha$
-cosymplectic manifold. Ricci tensor components are given by
Therefore, the scalar curvature
$r=-20\alpha ^2$
. Let us choose a potential vector field.
where
$f(z),g(z),h(z),k(z), m(z)$
are arbitrary smooth functions. By the soliton equation
(2)
, we get the soliton function
$\lambda$
,
Thus,
$(\mathbb{R}^5,V,\lambda, \rho, g)$
defines almost RB soliton.
Remark 1.
In the term
$e^{\alpha z}$
dependence of
$\lambda$
suggests a geometric soliton “wave” propagating along
$\xi$
.
This article is divided into
$3$
sections. In Section 2, we classify
$3$
-dimensional almost
$\alpha$
-cosymplectic manifold admitting an almost RB soliton. For this purpose, we use the classification of
$3$
-dimensional almost
$\alpha$
-cosymplectic Lie groups as in [Reference Perrone12]. In Section 3, we explore Ricci semisymmetric manifolds admitting RB solitons. In the sequel,
$(\tilde M, g)$
denotes the simply connected covering space of
$(M,g)$
. First, we describe the motivation behind our present exploration.
Motivation. The study of almost Ricci-Bourguignon solitons on almost
$\alpha$
-cosymplectic manifolds is motivated by several interrelated geometric goals:
-
1. Unification and generalization. The Ricci-Bourguignon soliton equation
provides a single framework that interpolates between Ricci, Einstein, and traceless Ricci solitons as
\begin{equation*} \textrm{Ric} + \frac 12\unicode{x00A3}_{V} g = (f + \rho r)g \end{equation*}
$\rho$
varies. Investigating this general structure on almost
$\alpha$
-cosymplectic manifolds (includes almost
$\alpha$
-Kenmotsu manifolds) yields classification results applicable to a whole family (depending on each
$\alpha$
) of geometric flows simultaneously.
-
2. Geometric rigidity. Almost contact metric structures are rich but complex. Almost contact metric manifolds are analogues of almost Hermitian manifolds in odd dimensions. The presence of a soliton imposes a strong differential constraint that often forces the underlying geometry to become more regular. In our present research, we show that an almost RB soliton can, under some appropriate natural conditions, force an almost
$\alpha$
-cosymplectic manifold to become
$\alpha$
-cosymplectic, or even reduce to an almost cosymplectic manifold (
$\alpha =0$
). This soliton thus acts as an instrument of geometric reduction in the family. -
3. Link to curvature and symmetry. In dimension three, the soliton equation simplifies the complicated curvature expressions of almost
$\alpha$
-cosymplectic manifolds, allowing a complete classification in terms of Lie group structures (Theorem2.1). This reveals precisely how the parameters
$(\lambda, a, b, \alpha, \mu )$
interact with the soliton function and the scalar curvature. -
4. Geometry behind the condition being
$\boldsymbol{\eta}$
-parallel in contact geometry: The contact and complex geometries interact very extensively. In almost Hermitian manifolds, the complex structure
$J$
being parallel, viz.,
$\nabla J = 0$
, corresponds to the vanishing of the Nijenhuis tensor being zero. This implies that
$J$
is integrable. Analogously, in an almost contact manifold, an associated CR-structure is integrable if and only if the structure tensor
$\phi$
is
$\eta$
parallel. In classical contact geometry,
$\eta$
-parallel conditions on tensors such as
$h$
or
$\tau$
have been shown to impose strong curvature restrictions (e.g., forcing
$(k,\mu )$
-structures or constant curvature). By studying almost RB solitons on almost
$\alpha$
-cosymplectic manifolds with
$\eta$
-parallel
$\varphi h$
, we extend this notion to a broader setting and obtain new rigidity theorems (Section 3).
2. Classification of almost
$\alpha$
-cosymplectic
$3$
D almost RB solitons
In this section, we classify
$3$
D almost RB transversal solitons on almost
$\alpha$
-cosymplectic manifolds having
$\xi$
as an eigenvector for the Ricci operator, for gradient almost RB transversal solitons, when the Reeb vector field is conformal, and also when the soliton is compact. The main tool used for this purpose is classification of almost
$\alpha$
-cosymplectic
$3$
D Lie groups by [Reference Perrone12].
An almost RB soliton is called transversal if its potential vector field
$V$
satisfies
$\eta (V)=0$
. The condition
$\nabla s = f_1(e + \varphi e)$
is a strong geometric constraint that implies the gradient of the potential function lies in a specific
$2$
-dimensional subbundle of the contact distribution and has a symmetric form with respect to the basis
$e, \varphi e$
. While it may appear synthetic, it is a natural ansatz that leads to a clear dichotomy in Theorem2.2: either the structure is trivial (Einstein) or the Reeb flow is degenerate (
$\mu$
= 0). This helps classify gradient solitons by ruling out non-trivial possibilities in certain geometric settings.
Theorem 2.1.
Let
$(M^3,\varphi, \xi, \eta, V,f,\rho, g)$
be a complete almost
$\alpha$
-cosymplectic almost RB transversal soliton. Suppose that
$\xi$
is an eigenvector of the Ricci operator with constant eigenvalue. Then
$({M}^3,\varphi, \xi, \eta, V,f,\rho, g)$
is an RB transversal soliton. Furthermore,
$\tilde {M^3}$
is a Lie group
$\mathcal{G}$
with a canonical left-invariant metric of type
$\mathbb{H}^3(\!-\alpha ^2)$
or
$\mathbb{H}^2(\!-a^2)\times \mathbb{R}$
or
$\mathcal{G}_{\alpha \pm \alpha \pm \alpha }$
if
$\alpha \neq 0$
. If
$\alpha =0$
, then
$\tilde M^3$
is almost cosymplectic and in this case
$\mathcal{G}=\tilde {E}^2$
with flat left-invariant metric, or
$\mathcal{G}=Sol_3$
,
$\mathcal{G}=\mathcal{G}_{0a0}$
or
$\mathcal{G}$
is
$Nil_3$
.
Proof.
We proceed in three steps: (i) derive constraints on
$f$
from the soliton equation, (ii) analyze the potential vector field
$V$
, and (iii) classify via Lie group structures.
From the definition of almost RB solitons, we have
If
$X=Y=\xi$
in (13) yields that,
From the above equation and (12) we have,
We assume that the potential vector field
$V$
is orthogonal to the Reeb vector field
$\xi$
, then
$V$
can be denoted by
$V=f_{1} e+f_{2} \varphi e$
. If
$X=$
$Y=e$
in (13) and using (8), we affirm
\begin{align} 0 = &\,g\left (\nabla _{e} f_{1} e+f_{2} \varphi e, e\right )+g(e, Q e)-(f+\rho r) g(e, e), \nonumber \\ = &\,g\left (e\left (f_{1}\right ) e+f_{1} \nabla _{e} e+e\left (f_{2}\right ) \varphi e+f_{2} \nabla _{e} \varphi e, e\right )-\left (2 \alpha ^{2}+a^{2}+b^{2}+2 \lambda \mu +\varphi e(a)-e(b)\right )\!,\nonumber \\ &-(f+\rho r),\nonumber \\ = &\,e\left (f_{1}\right )+2 \lambda ^{2}-2 \lambda \mu +a f_{2}-a^{2}-b^{2}-\varphi e(a)+e(b). \end{align}
In case,
$X=Y=\varphi e$
in (13), we affirm,
\begin{align} 0 = &\,g\left (\nabla _{\varphi e} f_{1} e+f_{2} \varphi e, \varphi e\right )+g(\varphi e, Q \varphi e)-(f+\rho r) g(\varphi e, \varphi e),\nonumber \\ = &\,g\left (\varphi e\left (f_{1}\right ) e+f_{1} \nabla _{\varphi e} e+\varphi e\left (f_{2}\right ) \varphi e+f_{2} \nabla _{\varphi e} \varphi e, \varphi e\right )-\left (2 \alpha ^{2}+a^{2}+b^{2}-2 \lambda \mu +\varphi e(a)-e(b)\right )\!,\nonumber \\ &-(f+\rho r)\nonumber \\ = &\,\varphi e\left (f_{2}\right )+2 \lambda ^{2}+2 \lambda \mu -b f_{1}-a^{2}-b^{2}-\varphi e(a)+e(b) . \end{align}
If
$X=\xi, Y=e$
in (13), we have
\begin{align} 0= &\, g\left (\nabla _{\xi } f_{1} e+f_{2} \varphi e, e\right )+g\left (\xi, \nabla _{e} f_{1} e+f_{2} \varphi e\right )+2 g(Q \xi, e)\nonumber \\ = &\, g\left (\xi \left (f_{1}\right ) e+f_{1} \nabla _{\xi } e+\xi \left (f_{2}\right ) \varphi e+f_{2} \nabla _{\xi } \varphi e, e\right )\nonumber \\ & +g\left (e\left (f_{1}\right ) e+f_{1} \nabla _{e} e+e\left (f_{2}\right ) \varphi e+f_{2} \nabla _{e} \varphi e, \xi \right )-4 a \lambda -2\varphi e(\lambda )\nonumber \\ = &\, \xi \left (f_{1}\right )+(\lambda -\mu ) f_{2}-\alpha f_{1}-4 a \lambda -2\varphi e(\lambda ). \end{align}
Again if
$X=\xi, Y=\varphi e$
in (13) implies
\begin{align} 0= &\, g\left (\nabla _{\xi } f_{1} e+f_{2} \varphi e, \varphi e\right )+g\left (\xi, \nabla _{\varphi e} f_{1} e+f_{2} \varphi e\right )+2 g(\xi, Q \varphi e)\nonumber \\ = &\, g\left (\xi \left (f_{1}\right ) e+f_{1} \nabla _{\xi } e+\xi \left (f_{2}\right ) \varphi e+f_{2} \nabla _{\xi } \varphi e, \varphi e\right )\nonumber \\ & +g\left (\varphi e\left (f_{1}\right ) e+f_{1} \nabla _{\varphi e} e+\varphi e\left (f_{2}\right ) \varphi e+f_{2} \nabla _{\varphi e} \varphi e, \xi \right )+4 b \lambda -2e(\lambda )\nonumber \\ = &\, \xi \left (f_{2}\right )+(\lambda +\mu ) f_{1}-\alpha f_{2}+4 b \lambda -2e(\lambda ). \end{align}
If
$X=e, Y=\varphi e$
in (13) gives,
\begin{align} 0= &\, g\left (\nabla _{e} f_{1} e+f_{2} \varphi e, \varphi e\right )+g\left (e, \nabla _{\varphi e} f_{1} e+f_{2} \varphi e\right )+2 g(e, Q \varphi e)\nonumber \\ = &\, g\left (e\left (f_{1}\right ) e+f_{1} \nabla _{e} e+e\left (f_{2}\right ) \varphi e+f_{2} \nabla _{e} \varphi e, \varphi e\right ) \nonumber \\ & +g\left (\varphi e\left (f_{1}\right ) e+f_{1} \nabla _{\varphi e} e+\varphi e\left (f_{2}\right ) \varphi e+f_{2} \nabla _{\varphi e} \varphi e, e\right )+4 \alpha \lambda +2\xi (\lambda )\nonumber \\ = &\, \varphi e\left (f_{1}\right )+e\left (f_{2}\right )-a f_{1}+b f_{2}+4 \alpha \lambda +2\xi (\lambda ) \end{align}
By hypothesis, if
$\xi$
is an eigenvector field of the Ricci operator
$Q\xi =\sigma \xi$
, where
$\sigma$
is constant, then we get that
$\lambda$
is constant and
$a \lambda =$
$b \lambda =0$
. If
$\lambda =0$
, manifold will become
$\alpha$
-cosymplectic. Now suppose
$\lambda \neq 0$
, then we have that
$a=b=0$
. Then the Lie brackets (9) reduce to
Operating
$f_{2}$
by
$[\xi, e]$
, we obtain
Applying the equation (19), we derive
Again operating
$f_{1}$
on
$[\xi, \varphi e]$
, we confirm
Applying the equation (18), we derive
The action of
$\xi$
on (20) yields
Combining (21) with (22), we conclude that
If
$\mu \neq 0$
, then we have
$\varphi e\left (f_{2}\right )-e\left (f_{1}\right )=0$
. Taking (16) and (17) into it, we find that
$\lambda$
must be zero, which is a contradiction. Observe that by (16) and (17)
$e(f_1)$
and
$\varphi e(f_2)$
are constant and equal to
$-2\lambda ^2$
. Now using the second identity of (9), we establish that
$\varphi e(f_1)=e(f_2)=-2\alpha \lambda$
. (If we assume that
$\lambda =a=b=0$
then
$\varphi e(f_1)=e(f_2)=0$
). Now, the proof follows from [Reference Perrone12], and we obtain the classification as given in the table below.

The gradient solitons exhibit special geometric properties. Here, we prove that under a transversal symmetry assumption
$(\nabla s = f_1(e + \varphi e))$
, the soliton collapses to a trivial Einstein structure unless the Reeb flow is degenerate (
$\mu$
= 0).
Theorem 2.2.
Let
$(M^3,\varphi, \xi, \eta, \nabla s,f,\rho, g)$
be a complete almost
$\alpha$
-cosymplectic almost gradient RB transversal soliton of type
$\nabla s = f_1(e + \varphi e)$
, for some
$f_1 \in C^{\infty }(M)$
. Then if
$\mu \neq 0$
, then the soliton is trivial (zero), and in this case
$\tilde {M}=\mathbb{R}^3$
or
$\tilde {E^2}$
.
Proof. We can write a gradient almost RB soliton as
If
$X=\xi$
with
$\nabla s=f_1e+f_2\varphi e$
, then employing (8), (11) yields,
Taking the inner product of (24) with
$\xi$
,
$e$
, and
$\varphi e$
respectively, we obtain
Subsituting
$X=e$
in (23), we affirm
Taking the inner product of (28) with
$\xi$
,
$e$
, and
$\varphi e$
, respectively, we get
Substituting
$X=\varphi e$
in (23), we confirm
Taking the inner product of (32) with
$\xi$
,
$e$
, and
$\varphi e$
, respectively, we establish
Next, we have if
$f_1=f_2$
, then from (26), (27), (29), and (33), we obtain
Next, if
$\mu \neq 0$
then
$f_1=0$
and the soliton is trivial, and for
$n\geq 3$
, we observe
$f+\rho r$
is constant; thus, the manifold is Einstein. Then by (25)
$\lambda$
is constant and eventually
$0$
. If
$\alpha =0$
, then
$M$
is flat and
$\tilde {M}=\mathbb{R}^3$
or
$\tilde {E^2}$
and
$\alpha$
nonzero is not possible [cf. table of Theorem2.1].
Proposition 2.3.
If
$(M^{2n+1},g,\phi, \xi, \eta )$
is an almost
$\alpha$
-cosymplectic manifold that admits a conformal Reeb vector field, then it must be Killing; consequently,
$\alpha =0$
and
$ h=0$
.
Proof. Recall the standard identity obtained from (7)
Assume
$\mathcal{L}_\xi g=2\sigma g$
for some
$\sigma \in C^\infty (M)$
. Setting
$X=\xi$
in the last equation and using
$\eta (\xi )=1$
,
$\phi \xi =0$
and
$h\xi =0$
, we obtain
hence
$\sigma =0$
and therefore
$\mathcal{L}_\xi g=0$
(so
$\xi$
is Killing). With
$\mathcal{L}_\xi g=0$
, equation reduces to
If
$X,Y\in \ker \eta$
, then
so as endomorphisms of the horizontal distribution
$\ker \eta$
,
Apply
$\phi$
to both sides, using
$\phi ^2|_{\ker \eta }=-\textrm{Id}$
we get,
thus,
The left-hand side is symmetric with respect to
$g$
while the right-hand side is skew-symmetric. Hence,
$\alpha =0$
and
$h|_{\ker \eta }=0$
. Since
$h\xi =0$
as well, we conclude
$h=0$
on
$M$
. This completes the proof.
Theorem 2.4.
An almost
$\alpha$
-Kenmotsu manifold can’t admit constant Reeb sectional curvatures.
Proof.
The curvature operator
$R$
is defined by
On an almost
$\alpha$
-Kenmotsu manifold
$M$
, from (7) we affirm
for any vector fields
$X$
,
$Y$
, where
$h^{'}=h\circ \varphi .$
Suppose that an almost
$\alpha$
-Kenmotsu manifold is of constant Reeb sectional curvatures, say
$\kappa \in \mathbb{R}$
. That is for any
$X$
,
Equivalently, for any
$X$
Setting
$Y=\xi$
in (36) yields,
Comparing above equation with (37) we infer,
As
$\nabla _{\xi } \varphi =0$
on an almost
$\alpha$
-Kenmotsu manifold, we have
$\nabla _{\xi } h^{\prime }=\nabla _{\xi } h \circ \varphi$
. Back substitution implies that
$\nabla _{\xi } h \circ \varphi =(\kappa +\alpha ) \varphi ^{2}-2 h^{\prime }-h^{2}$
. Thus,
As
$\nabla _\xi h$
is self-adjoint, we conclude
Back substitution of (38) in the previous equation yields,
Covariantly differentiating (38) along the Reeb vector field
$\xi$
, using the fact that
$\nabla _{\xi } \varphi =0$
and
$\kappa \in \mathbb{R}$
, we obtain
$\nabla _{\xi } h \circ h+h \circ \nabla _{\xi } h=0$
. Now, substituting
$\nabla _{\xi } h=-2 h$
into the previous equality gives
$h^{2}=0$
. Since
$h$
is self-adjoint, we have
$h=0$
. Hence, from (6) we have that
$\xi$
is parallel vector field. Hence, from the above Proposition
$\alpha = 0$
which is a contradiction.
Remark 2.
Note that in an almost
$\alpha$
-Kenmotsu manifold,
$\xi$
cannot be a conformal vector field as it forces
$\alpha =0$
from Proposition 2.3, which is a contradiction.
When the Reeb field is conformal, the behavior of the soliton function mirrors that of the scalar curvature. The result below establishes a decisive link: constancy of the soliton function is equivalent to constant scalar curvature, provided
$\rho \neq 0$
.
Theorem 2.5.
Suppose a three-dimensional almost cosymplectic manifold with a conformal Reeb vector field admits an almost gradient RB soliton with
$\rho \neq 0$
. Then the soliton function is constant if and only if the scalar curvature is constant.
Proof. For a gradient almost RB soliton, we have
Differentiating the above equation covariantly in the direction of
$X$
yields,
Thus, we affirm that
As
$\xi$
is conformal, then
$h=0$
, thus
$\xi$
is Killing, and the manifold is cosymplectic. We have Ricci curvature from [Reference Perrone12] as,
Differentiating the aforementioned equation covariantly in the direction of
$X$
, we confirm
Here, we have used [Reference Goldberg and Yano13]
$\nabla \xi =0$
if and only if
$\nabla \eta =0$
. In view of (42) and the above equation, we get
This implies that,
Note that we used
$\xi r=0$
because
$\xi$
is a Killing vector field. Also, we have
Consequently, we get from (45) that for any vector field
$X$
on
$M$
,
Contracting
$X$
in previous equation we get
$\xi f=0$
and hence from (47) for
$\rho \neq 0$
, we obtain
$f+\rho r= C$
, where
$C$
is some constant. This completes the proof.
Compactness imposes drastic restrictions. For 3D almost cosymplectic manifolds with conformal Reeb fields, we show that RB solitons must either be gradient Einstein or flat, reducing the classification to
$\mathcal{G}=Sol_3 \operatorname { or } Nil_3, \mathbb{R}^3, SL(2,\mathbb{R})\operatorname { or } \mathbb{S}^2(\frac {-c}{\rho })\times \mathbb{R}.$
Where
$c=-\frac {\rho r}{2}$
.
Theorem 2.6.
In a three-dimensional compact orientable almost cosymplectic RB soliton
$M^{3}$
without boundary, with the conformal Reeb vector field, has soliton function is constant if and only if the scalar curvature is constant, provided
$\xi f=0$
and
$\rho \neq 0$
. In this case,
$M^3$
can be
$\mathcal{G}=Sol_3\ or\ Nil_3, \mathbb{R}^3, SL(2,\mathbb{R})\ or\ \mathbb{S}^2(\frac {-c}{\rho })\times \mathbb{R}.$
Proof.
Suppose that (
$M^{3}, \varphi, \xi, \eta, g$
) is a three-dimensional almost cosymplectic manifold. As the Reeb vector field is conformal, so
$h=0$
,;thus, the Ricci operator is given by (43). It can be easily shown that the Ricci tensor for
$M^3$
is given by,
By our hypothesis (
$M^{3}, \varphi, \xi, \eta, g$
) admits an almost RB soliton. Therefore, (2) becomes
Covariant differentiation of
$\unicode{x00A3}_{V} g$
in the direction of
$X$
yields for any vector fields
$Y, Z$
on
$M$
,
Following Yano [Reference Yano14], the following formula holds true.
As
$g$
is parallel for the Levi-Civita connection
$\nabla$
, the above relation becomes
Since
$\unicode{x00A3}_{V} \nabla$
is a symmetric tensor of type (1,2), i.e.,
$\left (\unicode{x00A3}_{V} \nabla \right )(X, Y)=\left (\unicode{x00A3}_{V} \nabla \right )(Y, X)$
, it follows that
Employing (49) in (51) we affirm,
\begin{align} 2 g\left (\left (\unicode{x00A3}_{V} \nabla \right )(X, Y), Z\right )= &\, [2(X f)+(\rho -1)(X r)] g(Y, Z)+(X r) \eta (Y) \eta (Z)\nonumber \\ & +[2(Y f)+(\rho -1)(Y r)] g(X, Z)+(Y r) \eta (X) \eta (Z)\nonumber \\ & -[2(Z f)+(\rho -1)(Z r)] g(X,Y)-(Z r) \eta (X) \eta (Y). \end{align}
From the above equation, we have
\begin{align} 2\left (\unicode{x00A3}_{V} \nabla \right )(X, Y)= &\, [2(X f)+(\rho -1)(X r)]Y+(X r) \eta (Y)\xi \nonumber \\ & +[2(Y f)+(\rho -1)(Y r)]X+(Y r) \eta (X)\xi \nonumber \\ & -[2(D f)+(\rho -1)(D r)] g(X,Y)-(D r) \eta (X) \eta (Y). \end{align}
Covariantly differentiating
$\unicode{x00A3}_{V} \nabla$
in direction of
$X$
, we confirm
\begin{align} 2\left (\nabla _{X} \unicode{x00A3}_{V} \nabla \right )(Y, Z)= &\, \left [2 g\left (\nabla _{X}(D f), Y\right )+(\rho -1)g\left (\nabla _{X}(D r), Y\right )\right ] Z\nonumber \\ &+g\left (\nabla _{X}(D r), Y\right ) \eta (Z) \xi \nonumber \\ & +\left [2 g\left (\nabla _{X}(D f), Z\right )+(\rho -1)g\left (\nabla _{X}(D r), Z\right )\right ] Y \nonumber \\ & +g\left (\nabla _{X}(D r), Z\right ) \eta (Y) \xi \nonumber \\ & -g(Y, Z)\left [2 \nabla _{X} D f+(\rho -1)\nabla _{X} D r\right ]-\eta (Y) \eta (Z) \nabla _{X} D r. \end{align}
Using the foregoing equation in the following formula from [Reference Yano14]
We establish that
\begin{align} 2\left (\unicode{x00A3}_{V} R\right )(X, Y) Z= &\, \left [2 g\left (\nabla _{X} D f, Z\right )+(\rho -1)g\left (\nabla _{X} D r, Z\right )\right ] Y+g\left (\nabla _{X} D r, Z\right ) \eta (Y) \xi \nonumber \\ & -g(Y, Z)\left [2 \nabla _{X} D f+(\rho -1)\nabla _{X} D r\right ]-\eta (Y) \eta (Z) \nabla _{X} D r\nonumber \\ & -\left [2 g\left (\nabla _{Y} D f, Z\right )+(\rho -1)g\left (\nabla _{Y} D r, Z\right )\right ] X-g\left (\nabla _{Y} D r, Z\right ) \eta (X) \xi \nonumber \\ & +g(X, Z)\left [2 \nabla _{Y} D f+(\rho -1)\nabla _{Y} D r\right ]+\eta (X) \eta (Z) \nabla _{Y} D r . \end{align}
Contracting
$X$
in the previous equation, we infer
where
$\Delta$
denotes the Laplacian. Moreover, from (48) it follows directly that
\begin{align} \left (\unicode{x00A3}_{V} \operatorname {Ric}\right )(Y, Z)= &\, \frac {(V r)}{2} g(Y, Z)-\frac {(V r)}{2} \eta (Y) \eta (Z)+\frac {r}{2}\left [g\left (\nabla _{Y} V, Z\right )+g\left (Y, \nabla _{Z} V\right )\right ] \nonumber \\ & -\frac {r}{2} \eta \left (\nabla _{Y} V\right ) \eta (Z)-\frac {r}{2} \eta \left (\nabla _{Z} V\right ) \eta (Y). \end{align}
Employing (57) and (58) we confirm that,
\begin{align*} &-2 g\left (\nabla _{Y} D f, Z\right )-[2 \Delta f+(\rho -1)\Delta r] g(Y, Z)\\ &+g\left (\nabla _{\xi } D r, Z\right ) \eta (Y)-\Delta r \eta (Y) \eta (Z) -\rho g(\nabla _Y Dr,Z) \\ = &\, (V r) g(Y, Z)-(V r) \eta (Y) \eta (Z)+r\left [g\left (\nabla _{Y} V, Z\right )+g\left (Y, \nabla _{Z} V\right )\right ] \\ & -r \eta \left (\nabla _{Y} V\right ) \eta (Z)-r \eta \left (\nabla _{Z} V\right ) \eta (Y). \end{align*}
Then, substituting
$Y=\xi$
and
$Z=\xi$
in the foregoing equation, we obtain,
Now we assume that
$\xi f=0$
. Then (59) implies that
Consequently,
where
$c$
is some constant. This completes the proof. Furthermore, by (14), we have
$f+\rho r=0$
, thus
Hence,
$r$
is constant and if
$c=0$
, then
$r=0$
implies
$M^3=\mathbb{R}^3 /\Gamma$
is flat or
$M^3=SL(2,\mathbb{R})/\Gamma$
, where
$\Gamma$
is a discrete subgroup of rotations in
$M$
. Next, we assume if
$c\gt 0$
, then
$K(\varphi e,\xi )=K( e,\xi )=0$
and
$K(\varphi e,e)=\frac {r}{2}=\frac {-c}{\rho }=K_e$
. Then if
$\rho \gt 0$
$\mathcal{G}=Sol_3/\Gamma \operatorname { or } Nil_3/\Gamma$
and if
$\rho \lt 0$
, then
$\tilde {M}=\mathbb{S}^2(\frac {-c}{\rho })\times \mathbb{R}$
by [Reference Perrone12].
In [Reference Aqib, Shah and Patra10], it was proved that a compact non-trivial almost RB soliton with constant scalar curvature is isometric to a Euclidean sphere
$\mathbb{S}^{n}$
and is also gradient. Thus, we conclude that:
Corollary 2.1.
If
$M^3$
is a compact orientable almost cosymplectic manifold having a conformal Reeb vector field, admitting RB solitons, then it is isometric to a Euclidean sphere
$\mathbb{S}^{n}$
, and is a gradient.
Corollary 2.2.
Let
$M^3$
be a compact orientable almost cosymplectic manifold having a conformal Reeb vector field admitting RB solitons. Then an RB soliton reduces to a Ricci soliton.
Steady solitons induce geometric splitting. We prove that in the steady case, the manifold decomposes into a product of a line and a coKähler surface, with explicit descriptions of the universal cover.
Proposition 2.7.
Suppose that (
$M^{3}, \varphi, \xi, \eta, g$
) is a three-dimensional almost
$\alpha$
cosymplectic manifold admitting a steady transversal almost RB soliton. Then,
$M^3$
is cosymplectic and splits as
$(a, b) \times N^2,$
where
$N$
is a coKähler surface. In fact, in this case
$\tilde M$
is
$\mathbb{R}^3$
or
$\tilde {E}^2$
or
$\mathcal{G}_{0a0}$
.
Proof.
Let
$V=f_1 e +f_2 \varphi e$
. Now by (14), we have
Suppose that almost RB soliton is steady, then
Suppose that
$\rho r=0$
, then
$\alpha =\lambda =0$
, i.e., the manifold is cosymplectic. That implies
$\xi$
is a Killing vector field. Hence,
$M^3$
can be written as a product of the real line and a coKähler manifold. If
$r = 0$
, then because
Thus, in this case
$\tilde M$
is
$\mathbb{R}^3$
or
${\tilde {E}}^2$
. If
$\rho = 0$
, then
$\tilde M$
is
$\mathbb{R}^3$
or
$\tilde {E}^2$
or
$\mathcal{G}_{0a0}$
.
Suppose
$\rho \neq 0$
and
$\alpha \neq 0$
, then
$M^3$
is almost
$\alpha$
-Kenmotsu manifold in this case using (36),
Using these relations and the Ricci operator
$Q \xi$
(11), it follows that
$\lambda$
is constant. And thus by Remark 3.3 of [Reference Perrone12],
$\lambda$
is zero. Again, after some similar computations, it follows that
$\alpha = 1$
, that is
$M^3$
is almost Kenmotsu, scalar curvature is constant
$2$
and
$\rho = -1$
. But by the classification of
$3D$
Lie groups by [Reference Perrone12], it follows that this case can’t occur.
Do non-trivial almost RB solitons exist for
$\alpha \neq 0$
when
$\rho \gt \frac {1}{2(n-1)}?$
3. Ricci semi-symmetric almost
$\alpha$
-cosymplectic manifold
In this section, we prove that if a Ricci semi-symmetric almost
$\alpha$
-cosymplectic manifold, admitting an almost Ricci Bourguignon soliton and having
$\phi h$
$\eta$
-parallel tensor, is an almost cosymplectic manifold (i.e.,
$\alpha =0$
). We begin with the definition of
$\eta$
-parallel tensor field on a Riemannian manifold
$(M,g)$
. In the introduction geometry behind the condition being
$\eta$
-parallel in contact geometry is discussed. We have described the analogues of complex and contact geometry. It turns out that
$\eta$
-parallel tensor field corresponds to being
$J$
parallel in complex geometry. See Motivation in the introduction for more details.
For any vector field
$X$
on
$(M,g)$
decompose
$X = X^T+\eta (X)\xi$
, where
$X^T$
is tangential part of
$X$
and
$\eta (X) \xi$
is the normal part of
$X$
. The Jacobi operator associated with the Reeb vector field
$\xi$
is denoted by
$l$
and
$l=R(.,\xi )\xi$
.
Definition 2. [Reference Aktan, Yildirim and Murathan15] We say that any symmetric
$(1, 1)$
-type tensor field
$B$
on a Riemannian manifold
$(M,g)$
is said to be a
$\eta$
-parallel tensor, if it satisfies the equation
for all tangent vectors
$X^T, Y^T, Z^T$
orthogonal to
$\xi .$
The notion of an
$\eta$
-parallel tensor arises naturally in contact Riemannian geometry as a weakened parallelity condition adapted to the contact structure. In a contact metric manifold
$(M, g, \phi, \xi, \eta )$
, the tangent space splits as
$T_p M = D_p \oplus \mathbb{R} \xi$
, where
$D = \ker \eta$
is the contact distribution. Thus, in this case, a tensor field
$B$
is
$\eta$
- parallel if it is parallel when restricted to
$D$
.
It turns out that almost RB solitons with
$\eta$
-parallel tensor field on three-dimensional almost
$\alpha$
cosymplectic manifold have a simpler Ricci tensor.
Proposition 3.1.
Suppose that (
$M^{3}, \varphi, \xi, \eta, g$
) is a three-dimensional almost
$\alpha$
-cosymplectic manifold that admits an almost RB soliton with
$\eta$
-parallel tensor field
$\varphi h$
. Then the Ricci operator
$Q$
is given by
Proof.
On an almost
$\alpha$
-cosymplectic manifold
$M$
, the following relation holds [Reference Öztürk and Öztürk16].
If the tensor field
$\varphi h$
is
$\eta$
-parallel, then we have [Reference Aktan, Yildirim and Murathan15],
Using (60) and (62), we obtain
From (7), we have
This implies from (2) and (64) that
The aforementioned equation implies that
This yields
This gives the desired statement.
Riemannian manifolds admitting semi-parallel Ricci operators have been explored widely. See, for example, [Reference Singh and Lalmalsawma17] and the references therein.
Definition 3. [Reference Singh and Lalmalsawma17] A Riemannian manifold
$(M,g)$
is said to be Ricci semi-symmetric if
This is equivalent to
Now we prove the main result of this section. Ricci semi-symmetry triggers collapse to cosymplectic geometry. Our final theorem shows that semi-symmetry combined with
$\eta$
-parallel
$\varphi h$
-annihilates the
$\alpha$
-term, reducing the structure to the
$\alpha =0$
case.
Theorem 3.2.
A Ricci semi-symmetric almost
$\alpha$
-cosymplectic manifold with
$\varphi h$
is
$\ eta$
-parallel,
$ n\geq 2$
, admitting an almost Ricci Bourguignon soliton is an almost cosymplectic manifold.
Proof.
Consider an almost
$\alpha$
-cosymplectic manifold which is Ricci semi-symmetric. Then we have, by definition,
In particular, if
$X = \xi$
, then
From (68), it follows that
Now, using (63), (65) and (61) in the aforementioned identity, we get
Contracting it over
$X$
and
$Y$
we get
Therefore,
$\alpha =0$
. This completes the proof.
Acknowledgements
The authors would like to thank the referee for carefully reading the manuscript.
Funding statement
Not applicable.
Competing interests
The authors have no conflict of interest and no financial interests in this article.
Ethics approval
The submitted work is original and not submitted to more than one journal for simultaneous consideration.
Consent to participate
Not applicable.
Consent for publication
Not applicable.
Availability of data and materials
This manuscript has no associated data.
Code availability
Not applicable.
Author contributions
Conceptualization, methodology, investigation, validation, writing-original draft, review, editing, and heading have been performed by all the paper’s authors.
α
3
α




