Finite Volume Einstein Finsler Warped Product Manifolds of Non-positive or Non-negative Scalar Curvature Mohammad Aqib, Hemangi Madhusudan Shah, Pankaj Kumar, Anjali Shriwastawa Abstract We partially answer a question by A. L. Besse in the Finslerian settings by investigating the obstructions to the existence of Einstein warped product manifold, when the base is compact or finite volume complete Riemannian manifold and the fiber is a weakly Berwald Finsler manifold. In case of compact base, we establish that, this Finslerian warped must have dimension greater than 3 3 , if the Ricci scalar is non-positive or non-negative. We relate various conditions on the scalar curvatures of the base, fiber, and the warped product manifold which imply that the resulting warped product is trivial. We also show that, when Ricci scalar of the warped product is positive, then Ricci scalar of the fiber must be positive, in order that the warped product is non-trivial. Moreover, we prove that, if the base manifold is a complete, Riemannian manifold of finite volume with a bounded warping function, and the Ricci scalar of the warped product is non-positive, then the warped product is trivial. Our findings extend the results presented in Kim and Kim paper entitled, ”Compact Einstein warped product spaces with non-positive scalar curvature,” published in (2003), and Dan Dumitru’s results on ”On compact Einstein warped products,” published in (2011), in the Finslerian settings. keywords Finsler warped product, Finite volume Riemannian manifold, Weakly Berwald Finsler manifold, Omori-Yau maximum principle. † † equal-contributors: These authors contributed equally to this work. † † equal-contributors: These authors contributed equally to this work. † † equal-contributors: These authors contributed equally to this work. AMS classification 2020: 53B40, 52B60. 1 Introduction The notion of warped product plays an important role in Riemannian geometry, moreover in geodesic metric spaces. The warped product was first introduced by Bishop and O’Neill to study Riemannian manifolds of negative curvature (see lota ). Warped products have been mainly used to construct new examples of Riemannian manifolds with prescribed curvature conditions. This construction can be extended for Finslerian metrics with some minor restrictions. This is motivated by Asanov’s papers gsa ; gsa1 , where some models of relativity theory are described through the warped product of Finsler metrics. These metrics are in the form of (\alpha,\beta) -metrics, which are the generalization of the Randers metrics; which are being asymmetric Finsler metrics in four-dimensional space-time. The product was later extended to the warped product case of Finsler manifolds by the work of Kozma, Peter and Varga (see LPC ). This article is devoted to the study of Einstein Finslerian warped product manifolds, with base a Riemannian manifold and fiber a weakly Berwald manifold. It turns out that, in the case of Finsler manifolds of dimensions greater than 2 , in general, Schur’s Lemma does not hold true. Hence, in this case, the Ricci curvature turns out to be a scalar function on the manifold (see Definition 7 and description after it). This is how Finsler geometry differs from Riemannian one drastically. In his inquiry, A.L. Besse ( bas , p. 265 ) posed the question: ”Does there exist a compact Einstein warped product with nonconstant warping function?” This was answered partially and negatively by Kim and Kim who proved in kim : Theorem 1 ( Theorem 1 ( kim , p. 2573 ) ) . If (M,g) is an Einstein Riemannian warped product manifold with non-positive scalar curvature and compact base, then M is simply a Riemannian product space. Also towards the aforementioned question by Besse, Dan Dumitru dam showed that the fiber of a compact Einstein warped product should have strictly positive scalar curvature, in order that the warped product is non-trivial. And furthermore, he explored the new obstructions to the existence of such spaces (see Theorem 2.5 of dam for more details). In particular, it is also observed that a compact Einstein warped product must have a dimension greater than 5 dam . We can extend the Besse question in the Finslerian settings as follows: Question by A.L. Besse in the Finslerian settings: Does there exist a finite volume Einstein Finsler warped product space with a nonconstant warping function? In this article, we further provide partial negative answers to Besse’s question in the Finslerian settings by demonstrating that, if M is an Einstein Finslerian warped product of constant non-positive Ricci (non-positive Ricci scalar), having base a finite volume complete Riemannian manifold (compact Riemannian manifold) and fiber a weakly Berwald Finsler manifold, then the resulting manifold M is a weakly Berwald Finslerian Riemannian product manifold. In particular, we also obtain the Finslerian generalization of Theorem 1 . See our Theorem 10 and Theorem 11 proved in § 5 . We also answer Besse’s question partially negatively, by exploring obstructions in terms of scalar curvatures of base (a compact manifold) and fiber manifold, to the existence of an Einstein Finslerian warped product of positive Ricci scalar function. Thus we also extend Theorem 2.5 of dam . See Theorem 13 of § 6 . We also observe that, the dimension of the Einstein Finslerian warped product space with compact base, in case when Ricci scalar function is positive or negative, must be at least 4 (Proposition 4 of § 5 ). It is well known that, the Finsler manifolds have several non-equivalent volume measures. To study Einstein Finslerian warped product of finite volume, we first establish their admissible class of warped product metrics, by affirming that the warped product of Finsler manifolds of finite Holmes-Thompson volume, maximum volume or minimum volume, must be of finite volume in the same category (Theorem 5 Theorem 6 of § 3 ). The article is divided into 6 sections. In Section 2 , we discuss the basic notions required to prove our main results. In Section 3 , we discuss the formulation of Einstein Finsler warped product manifolds and the admissible class of such manifolds of finite well-known volumes (as mentioned above). In Section 4 , we show that the warped product of two weakly Berwald Finsler manifolds is again weakly Berwald, if the warping function is constant (Corollary 4.2 ). Section 5 and Section 6 , respectively, deal with the investigation of Einstein Finslerian warped product of non-positive and non-negative scalar curvatures, respectively. 2 Preliminaries In this section, we provide some basic notions required to prove our main result. Let M be an n -dimensional smooth manifold and let TM:={\sqcup_{x\in M}}T_{x}M denote the tangent bundle of M . Let TM^{0}:=TM\backslash\{0\} be the slit tangent bundle of M and correspondingly we also express T_{x}M^{0}:=T_{x}M\backslash\{0\} as slit tangent space at x\in M . We first begin with the definition of a Finsler metric. Definition 1 (Finsler metric ( SSZ , p. 12 ) ) . A Finsler metric on M is a continuous function F:TM\to[0,\infty) satisfying the following conditions: (a) F is smooth on TM^{0} , (b) F is a positively 1-homogeneous on the fibers of tangent bundle TM , (c) The Hessian of \frac{1}{2}F^{2} with element g_{ij}=\frac{1}{2}\frac{\partial^{2}F^{2}}{\partial y^{i}\partial y^{j}} is positive definite on TM^{0} . The pair (M,F) is called a Finsler manifold and g_{ij} is called the fundamental metric tensor. The notion of Finslerian geodesics extends to that of the Riemannian geodesics. The Finslerian geodesic equation involves the spray coefficients which generalize the Riemannian Christoffel symbols. Definition 2 (Finsler geodesic ( SSZ , p. 78 ) ) . A smooth curve in a Finsler manifold (M,F) is a geodesic , if it has constant speed and is locally length minimizing. Thus, a geodesic in a Finsler manifold (M,F) is a curve \gamma:I=(a,b)\rightarrow M with F(\gamma(t),\dot{\gamma}(t))= constant and for any t_{0}\in I , there is a small number \epsilon>0 , such that \gamma is length minimizing on [t_{0}-\epsilon,t_{0}+\epsilon]\cap I . Definition 3 (Geodesic equation, Spray coefficients ( SSZ , p. 78,79 ) ) . A smooth curve \gamma in a Finsler manifold (M,F) is a geodesic, if and only if in local coordinates \gamma(t)=(x^{i}(t)) satisfies the following second order non linear differential equation: \displaystyle\frac{d^{2}x^{i}(t)}{dt^{2}}+{G}^{i}\left(x^{i},\frac{dx^{i}}{dt}\right)=0,\qquad 1\leq i\leq n, (1) where G^{i}=G^{i}(x,y) are local functions on TM defined by \displaystyle{G}^{i}=\frac{1}{4}{g}^{i{\ell}}\left\{\left[{F}^{2}\right]_{x^{k}y^{\ell}}y^{k}-\left[{F}^{2}\right]_{x^{\ell}}\right\}. (2) The coefficients G^{i} are called the spray coefficients and the quantity G:=y^{i}\frac{\partial}{\partial x^{i}}+G^{i}\frac{\partial}{\partial y^{i}} is called spray on M . Now we discuss the notion of Riemann curvature tensor in the Finslerian settings in order to define the flag and the Ricci curvature. Definition 4 ( The Riemann curvature tensor ( CXSZ , p. 50 ) ) . Let (M^{n},F) be a Finsler manifold. The Riemann curvature tensor on M is a mapping R={{R}_{\xi}:T_{x}M\rightarrow T_{x}M} given by {R}_{\xi}(u)={R}^{i}_{k}(x,\xi)u^{k}\frac{\partial}{\partial x^{i}} , u=u^{k}\frac{\partial}{\partial x^{k}} , where {R}^{i}_{k}={R}^{i}_{k}(x,\xi) denote the coefficients of the Riemann curvature tensor of the Finsler metric F and are given by {R}^{i}_{k}=2\frac{\partial{G}^{i}}{\partial x^{k}}-\xi^{j}\frac{\partial^{2}{G}^{i}}{\partial x^{j}\partial\xi^{k}}+2{G}^{j}\frac{\partial^{2}{G}^{i}}{\partial\xi^{j}\partial\xi^{k}}-\frac{\partial{G}^{i}}{\partial\xi^{j}}\frac{\partial{G}^{j}}{\partial\xi^{k}}, (3) for 1\leq i,j,k\leq n. The flag curvature K=K(x,\xi,P) , extends the concept of sectional curvature from Riemannian to Finsler geometry, independent of whether employing the Berwald, Chern, or Cartan connection. Definition 5 ( Flag Curvature ( SSZ , p. 98 ) ) . Let P\subset T_{x}M be a tangent plane. For a vector y\in P\backslash\{0\} , define \mbox{K}(P,y):=\frac{\mbox{g}_{y}\left(\mbox{R}_{y}(u),u\right)}{\mbox{g}_{y}(y,y)\mbox{g}_{y}(u,u)-\mbox{g}_{y}(y,u)\mbox{g}_{y}(y,u)}, where u\in P such that P=\operatorname{span}\{y,u\} . We can easily show that \mbox{K}(P,y) is independent of u\in P with P=\operatorname{span}\{y,u\} . The number \mbox{K}(P,y) is called the flag curvature of the flag (P,y) in T_{x}M . Definition 6 ( The Ricci curvature ( SSZ , p. 98 ) ) . Define {\operatorname{Ric}}(y):=\sum_{i=1}^{n}R_{i}^{i}(y). Ric is a scalar function on TM^{0} , which satisfies the following homogeneity property: {\operatorname{Ric}}(\lambda y)=\lambda^{2}{\operatorname{Ric}}(y),\quad\lambda>0. We call Ric the Ricci curvature. The notion of Einstein Finsler manifolds have been studied extensively, see for eg., xz1 . Definition 7 ( Einstein manifold ( SSZ , p. 118 ) ) . Let (M^{n},F) be a Finsler manifold of dimension n , then M is said to be an Einstein manifold if \operatorname{Ric}(x,y)=\lambda(x)F^{2}(x,y). By differentiating twice with respect to y_{i} and y_{j} , we infer \operatorname{Ric}_{ij}=\lambda(x)g_{ij}, (4) where \lambda is a scalar function on M . (M,F) is said to be of Ricci constant, if the function \lambda is constant. For Riemannian manifolds of dimension \geq 3 , Schur’s lemma ensures that every Riemannian Einstein metric is necessarily of constant Ricci curvature. However, it is not known whether such a Schur lemma holds for Einstein Finsler metrics in general. But, if we restrict Finsler metrics to those of Randers type, then Schur’s lemma holds for dimension \geq 3 . See boa , p. 216 , for more details. Finsler manifolds differ drastically from Riemannian manifolds and have rich geometry. Distortion, S -Curvature, Berwald curvature, E -curvature etc., measure how far a Finsler manifold is away from being Riemannian. We need these quantities for our geometric analysis and now we are probing into them. A Finsler measure space , (M^{n},F,d\mu) is a Finsler manifold equipped with volume as measure. Definition 8 ( Distortion, ( SSZ , p. 117 ) ) . Suppose (M^{n},F,d\mu) is a Finsler measure space. Let \{b_{i}\}_{i=1}^{n} be an arbitrary basis for T_{x}M and \{\theta^{i}\}_{i=1}^{n} be its dual basis for T_{x}^{\ast}M . Then we can write the volume form as d\mu={\sigma}_{\mu}(x)\;\theta^{1}\wedge...\wedge\theta^{n} . For a vector y\in T_{x}M_{0} , the distortion of a Finsler metric F with measure \mu is defined by \tau_{\mu}(x,y):=\ln\frac{\sqrt{\operatorname{det}(g_{ij}(y))}}{\sigma_{\mu}(x)}, where g_{ij}(y):=g_{y}(b_{i},b_{j}). Definition 9 ( S-Curvature ( SSZ , p. 118 ) ) . (M^{n},F,d\mu) be a Finsler measure space. For x\in M and a vector y\in T_{x}M_{0} , let \gamma=\gamma(t) be the geodesic with \gamma(0)=x and \dot{\gamma}(0)=y . Then S_{\mu}(x,y):=\frac{d}{dt}\left[\tau(\gamma(t),\dot{\gamma}(t))\right]\Bigr|_{t=0}, is called the S -curvature of the Finsler metric F with distortion {\tau}_{\mu} . Definition 10 ( The Berwald curvature tensor ( FMSBC , p. 127 ) ) . Let (M^{n},F) be a Finsler manifold. For a non-zero tangent vector y\in T_{x}M , in local coordinate system, define B_{y}:T_{x}M\times T_{x}M\times T_{x}M\rightarrow T_{x}M by \displaystyle B_{y}(u,v,w):=B^{i}_{jk{\ell}}(x,y)u^{j}v^{k}w^{\ell}\frac{\partial}{\partial x^{i}}, (5) where \displaystyle B^{i}_{jk{\ell}}=[G^{i}]_{y^{j}y^{k}y^{\ell}},\quad u=u^{i}\frac{\partial}{\partial x^{i}},\quad v=v^{i}\frac{\partial}{\partial x^{i}},\quad w=w^{i}\frac{\partial}{\partial x^{i}}, (6) for 1\leq i,j,k,l\leq n. Then B=\left\{B_{y}|\ y\in TM^{0}\right\} is called the Berwald curvature tensor and the corresponding coefficients B^{i}_{jk{\ell}} are called as Berwald curvature. A Finsler metric is called Berwald metric , if B\equiv 0. Definition 11 ( E-Curvature ( SSZ , p. 118 ) ) . Let E_{ij}:=\frac{1}{2}S_{y^{i}y^{j}}=\frac{1}{2}\left[\frac{\partial G^{s}}{\partial y^{s}}\right]_{y^{i}y^{j}}. The symmetric tensor E:=E_{ij}(x,y)\mathrm{d}x^{i}\otimes\mathrm{d}x^{j} is called the E curvature tensor. E=\{E_{y}|y\in TM^{0}\} is called the E -curvature or the mean Berwald curvature. A local formula for E -curvature is given by SSZ , E_{y}(u,v)=\frac{1}{2}S_{y^{i}y^{j}}(y)u^{i}v^{j}=\frac{1}{2}\frac{\partial^{3}G^{\gamma}}{\partial y^{i}\partial y^{j}\partial y^{\gamma}}(y)u^{i}v^{j}=\frac{1}{2}B^{\gamma}_{ij\gamma}, where E_{y}(u,v):=E_{ij}(x,y)u^{i}v^{j}, u=u^{i}\partial_{i} , v=v^{i}\partial_{i} . Finsler manifolds with mean Berwald curvature zero are also called as weakly Berwald Finsler manifolds. It should be noted that clearly, if S=0 , then E=0 . Moreover, the following result is well-known. Proposition 2 (Proposition 5.1.2 ( VCRFG , p. 90 ) ) . For any Berwald metric, the S -curvature vanishes, S\equiv 0 . There are numerous volume forms in Finsler geometry unlike in the Riemannian case. We will be working with the well-known volume forms, the Busemann Hausdorff, Holmes-Thompson, maximum and minimum. Now we define them. Let \mathbb{B}^{n}(1) denote the Euclidean unit ball in \mathbb{R}^{n} and {\operatorname{Vol}(\mathbb{B}^{n}(1))} denote the canonical volume of the Euclidean ball. Definition 12 (Busemann-Hausdorff Volume form ( SSZ , p. 22 ) ) . The Busemann-Hausdorff volume form on a Finsler manifold (M^{n},F) is defined as: \operatorname{dV}_{BH}=\sigma_{BH}(x) \operatorname{dx} , where \sigma_{BH}(x)=\frac{\emph{Vol}(\mathbb{B}^{n}(1))}{\textnormal{Vol}\left\{(\xi^{i})\in T_{x}M:F(x,\xi)<1\right\}}. (7) Definition 13 ( Holmes-Thompson Volume form ( sc , p. 6 ) ) . For a Finsler manifold (M^{n},F) , the Holmes-Thompson volume form is defined as \operatorname{dV}_{HT}=\sigma_{HT}(x) \operatorname{dx} , where \sigma_{HT}(x)=\frac{1}{\operatorname{Vol}({\mathbb{B}}^{n}(1))}\int_{F(x,y)<1}\det(g_{ij}(x,y))\operatorname{dy}. (8) Definition 14 ( Max and Min Volume forms ( BYW , p. 727 ) ) . The maximum volume form \operatorname{dV}_{max} and the minimum volume form \operatorname{dV}_{min} of a Finsler metric F is defined as, \operatorname{dV}_{\max}=\sigma_{\max}(x)\operatorname{dx},\quad\operatorname{dV}_{\min}=\sigma_{\min}(x)\operatorname{dx}, (9) where \sigma_{\max}(x)=\max\limits_{y\in I_{x}}\sqrt{\det(g_{ij}(x,y))} and \sigma_{\min}(x)=\min\limits_{y\in I_{x}}\sqrt{\det(g_{ij}(x,y))} ; I_{x}=\left\{y\in T_{x}M|F(x,y)=1\right\} is the indicatrix of the Finsler metric F at the point x . 3 Admissible class of Einstein Finsler warped product manifolds of finite volume In this section, we explore the concept of the warped product of two Finslerian manifolds, which is a generalization of the analogous Riemannian concept of warped product manifold. Then we recall the relationships between Ricci curvatures and scalar curvatures of the warped product, its base and fiber manifold. Our main theorems rely heavily on these computations. See subsection 3.1 . Further, in subsection 3.2 , we establish that Finsler warped products of finite Holmes-Thompson/maximum/minimum volume are of finite volume in the same category. Finally, in subsection 3.3 , we confirm the existence of Einstein Finsler warped product manifolds of finite volumes, in the aforementioned categories, thereby establishing the admissible class of Einstein Finsler warped product manifolds of finite volume. 3.1 Einstein Finslerian warped product manifolds We begin with the definition of warped product Finsler manifolds. Definition 15 ( Warped product Finsler manifolds ( wlt , p. 10 ) ) . On the product manifold M=M_{1}\times M_{2} , we consider a function on TM_{1}^{0}\times TM_{2}^{0} , where TM_{i}^{0}=TM_{i}\backslash\{0\} , i=1,2 , defined as F^{2}(x_{1},y_{1},x_{2},y_{2})=F_{1}^{2}(x_{1},y_{1})+f^{2}(x_{1})F_{2}^{2}(x_{2},y_{2}), (10) for (x_{1},y_{1},x_{2},y_{2})\in TM_{1}^{0}\times TM_{2}^{0} . The function F is smooth and positively homogeneous. For the function F , we have \displaystyle\mathbf{g}_{ab}(x_{1},y_{1},x_{2},y_{2})=\Bigg(\frac{1}{2}\frac{\partial^{2}F^{2}(x_{1},y_{1},x_{2},y_{2})}{\partial\mathbf{y}^{a}\partial\mathbf{y}^{b}}\Bigg) \displaystyle\quad\quad\quad\quad\quad\quad\quad\quad=\Bigg(\begin{matrix}g_{ij}(x_{1},y_{1})&0\\ 0&f^{2}(x_{1})g_{\alpha\beta}(x_{2},y_{2})\end{matrix}\Bigg), (11) where g_{ij}(x_{1},y_{1})=\frac{1}{2}\frac{\partial^{2}F_{1}^{2}(x_{1},y_{1})}{\partial y_{1}^{i}\partial y_{2}^{j}},\quad\quad g_{\alpha\beta}(x_{2},y_{2})=\frac{1}{2}\frac{\partial^{2}F_{2}^{2}(x_{2},y_{2})}{\partial y_{1}^{\alpha}\partial y_{2}^{\beta}}, and \mathbf{y}^{a}=(y_{1}^{i},y_{2}^{\alpha}),\>\mathbf{y}^{b}=(y_{1}^{j},y_{2}^{\beta}),\>\mathbf{g}_{ij}=g_{ij},\>\mathbf{g}_{\alpha\beta}=g_{\alpha\beta} . The set of indices satisfy: \displaystyle i,j,k,...\in\{1,2,...,n_{1}\},\quad\alpha,\beta,\gamma,...\in\{1,2,...,n_{2}\},\quad a,b,c,...\in\{1,2,...,n_{1}+n_{2}\}. Thus (M,F) defines a Finsler manifold and is called the Finsler warped product of M_{1} and M_{2} , and is denoted by M=M_{1}\times_{f}M_{2}. M_{1} is called the base manifold and M_{2} is called as the fiber manifold. Note: The warped product is called as trivial , if the warping function is constant. In this case, the product will be termed as Riemannian product of two Finsler manifolds. The following result by Tokura et al., wlt , which formulates the Ricci curvature of the warped product of Finslerian manifolds lays the foundation for our research in this paper. In what follows, we will use \Delta f=-\mbox{tr}_{g}H^{f}=-\mbox{div}(\nabla f) , where H^{f} is the Hessian of the function f . Proposition 3 (Proposition 3.6 ( wlt , p. 12 ) ) . Let M=M_{1}^{n_{1}}\times_{f}M_{2}^{n_{2}} be a Finslerian warped product manifold. If (M_{1}^{n_{1}},g_{1}) is a Riemannian manifold, then \displaystyle\operatorname{Ric}_{ij} \displaystyle=(\operatorname{Ric}^{M_{1}})_{ij}-\frac{{n_{2}}}{f}\bigg(\frac{\partial^{2}f}{\partial x_{1}^{i}\partial x_{2}^{j}}-G_{ij}^{k}\frac{\partial f}{\partial x_{1}^{k}}\bigg), (12) \displaystyle\operatorname{Ric}_{i\alpha} \displaystyle=0, (13) \displaystyle\operatorname{Ric}_{\alpha\beta} \displaystyle=\operatorname{Ric}_{\alpha\beta}^{M_{2}}+\frac{1}{f}y_{1}^{h}\frac{\partial f}{\partial x_{1}^{h}}G_{\alpha\beta\gamma}^{\gamma}-[-f\Delta f+(n_{2}-1)g_{1}(\nabla f,\nabla f)]g_{\alpha\beta}, (14) where \nabla f and \Delta f represent the gradient and the Laplacian of f on the base manifold respectively, and (\operatorname{Ric}^{M_{1}})_{ij} is the component of the Riemannian Ricci curvature on M_{1} . Hence, Einstein Finslerian warped products satisfy: Corollary 3.1 . Let M=M_{1}^{n_{1}}\times_{f}M_{2}^{n_{2}} be a Finslerian warped product manifold, where (M_{1}^{n_{1}},g_{1}) is a Riemannian manifold , then M is Einstein Finslerian warped product with \operatorname{Ric}_{ab}=\lambda g_{ab} if and only if \displaystyle(\operatorname{Ric}^{M_{1}})_{ij} \displaystyle=\lambda g_{1}{{}_{ij}}+\frac{{n_{2}}}{f}\bigg(\frac{\partial^{2}f}{\partial x_{1}^{i}\partial x_{2}^{j}}-G_{ij}^{k}\frac{\partial f}{\partial x_{1}^{k}}\bigg), (15) \displaystyle\operatorname{Ric}_{\alpha\beta}^{M_{2}}= \displaystyle[\lambda f^{2}-f\Delta f+(n_{2}-1)g_{1}(\nabla f,\nabla f)]g_{\alpha\beta} (16) \displaystyle-\frac{1}{f}y_{1}^{h}\frac{\partial f}{\partial x_{1}^{h}}G_{\alpha\beta\gamma}^{\gamma}, where, \lambda and \mu respectively, are Ricci scalar functions on M and fiber of M , viz., (M_{2}^{n_{2}},F_{2}) respectively. Proof. Suppose M is Einstein, then \operatorname{Ric}_{ab}=\lambda g_{ab}. (17) Now using equations ( 12 ) and ( 14 ), we obtain ( 15 ) and ( 16 ). ∎ Consequently, we obtain a corollary that will significantly aid our forthcoming calculations. Corollary 3.2 . If M_{1}^{n_{1}}\times_{f}M_{2}^{n_{2}} is a Finslerian warped product manifold, where (M_{1}^{n_{1}},g_{1}) is a Riemannian manifold and (M_{2}^{n_{2}},F_{2}) is a weakly Berwald Finsler manifold, then M is Einstein Finsler warped product with \operatorname{Ric}_{ab}=\lambda g_{ab} if and only if \displaystyle(\operatorname{Ric}^{M_{1}})_{ij}=\lambda g_{1}{{}_{ij}}+\frac{{n_{2}}}{f}H_{ij}^{f}, (18) \displaystyle(M_{2},g_{M_{2}})\,\text{is Einstein with }\,\operatorname{Ric}_{\alpha\beta}^{M_{2}}=\mu g_{\alpha\beta}, (19) \displaystyle\mu=\lambda f^{2}-f\Delta f+(n_{2}-1)g_{1}(\nabla f,\nabla f). (20) Proof. As (M_{1},g_{1}) is a Riemannian manifold, the spray coefficients G_{ij}^{k} actually reduce to Christoffel symbols \Gamma^{k}_{ij} . And the expression \bigg(\frac{\partial^{2}f}{\partial x_{1}^{i}\partial x_{2}^{j}}-G_{ij}^{k}\frac{\partial f}{\partial x_{1}^{k}}\bigg) is a coordianate expression for the Hessian of f . Hence, we obtain ( 18 ). As fiber manifold is weakly Berwald, G_{\alpha\beta\gamma}^{\gamma}=0 . Thus we get ( 20 ). ∎ It should be noted that, we recover Ricci curvatures of Riemannian warped product manifolds, in particular, Corollary 3 kim , about Einstein warped manifolds. Next, we will obtain relations between \operatorname{Scal}_{M} , \operatorname{Scal}_{M_{1}} and \operatorname{Scal}_{M_{2}} . These relationships will be useful in finding obstructions to the existence of Einstein warped product manifolds. Corollary 3.3 . If M_{1}^{n_{1}}\times_{f}M_{2}^{n_{2}} is an Einstein Finslerian warped product manifold with \operatorname{Ric}_{ab}=\lambda g_{ab} , where (M_{1}^{n_{1}},g_{1}) is a Riemannian manifold and (M_{2}^{n_{2}},F_{2}) is a weakly Berwald Finsler manifold. Then the scalar curvatures of warped product, base and fiber satisfy: \operatorname{Scal}_{M_{1}}=n_{1}\lambda-n_{2}\frac{\Delta f}{f}, (21) \operatorname{Scal}_{M_{2}}=n_{2}\mu, (22) \operatorname{Scal}_{M_{1}}+\dfrac{\operatorname{Scal}_{M_{2}}}{f^{2}}=n_{2}(n_{2}-1)\dfrac{\|\nabla f\|^{2}}{f^{2}}-2n_{2}\frac{\Delta f}{f}+\operatorname{Scal}_{M}. (23) Proof. Now contracting the first two equations of the aforementioned corollary, we confirm \operatorname{Scal}_{M_{1}}=n_{1}\lambda-n_{2}\frac{\Delta f}{f}, and \operatorname{Scal}_{M_{2}}=n_{2}\mu. If we use ( 20 ), ( 21 ) and ( 22 ), we conclude \operatorname{Scal}_{M_{1}}+\dfrac{\operatorname{Scal}_{M_{2}}}{f^{2}}=n_{2}(n_{2}-1)\dfrac{\|\nabla f\|^{2}}{f^{2}}-2n_{2}\frac{\Delta f}{f}+\operatorname{Scal}_{M}. ∎ We will need the following important observation about Einstein Finslerian warped product manifolds which tells us that the dimension of non-trivial Finslerian manifolds must be at least 4 . Proposition 4 . Let (M_{1}^{n_{1}},g_{1}) be a compact Riemannian manifold, the base, f as the warping function, and (M_{2}^{n_{2}},F_{2}) be a weakly Berwald Finsler manifold, the fiber. Then, for an Einstein warped product manifold M=M_{1}\times_{f}M_{2} with n_{1}=1 or n_{2}=1 and a non-negative (or non-positive) Ricci scalar \lambda(x_{1},x_{2}) , the warped product is simply a Riemannian product of two Finslerian manifolds, and is in fact, also a weakly Berwald manifold. Proof. We prove it in two cases: Case 1 : n_{1}=1. As n_{1}=1 so (\operatorname{Scal}_{M_{1}})_{ij}=0. If we take trace in ( 18 ), then we obtain \frac{\lambda(x_{1},x_{2})}{n_{2}}=\frac{\Delta f(x_{1})}{f(x_{1})}. Observe that from the last equation, if \lambda(x_{1},x_{2}) is non-negative (or non-positive), then f(x_{1}) is constant, as the base manifold is compact. Case 2 : n_{2}=1. As n_{2}=1 so (\operatorname{Scal}_{M_{2}})_{ij}=0. If we take trace in ( 19 ) then we obtain, \mu(x_{2})=0. Thus, by ( 20 ) we get \lambda(x_{1},x_{2})f(x_{1})=\Delta f(x_{1}). Hence, by the similar argument as in case 1 , f(x_{1}) is constant. In both the cases, M is the Riemannian product of two Finsler manifolds and is weakly Berwald by Corollary 4.1 proved in § 4 ∎ Thus, we conclude that, the dimension of Finslerian warped product can not be 1,2, or 3 . Corollary 3.4 . Let M=M_{1}^{n_{1}}\times_{f}M_{2}^{n_{2}} be as in the above proposition. Then, for Einstein Finslerian manifold M , the warped product will be nontrivial, if \operatorname{dim}M\geq 4 and the Ricci scalar \lambda(x_{1},x_{2}) is non-negative (or non-positive). Remark 1 . In view of the above corollary, in the sequel we will work with n_{1},n_{2}\geq 2 and n\geq 4 . 3.2 Finsler warped product manifolds of finite volume Warped product of Finsler manifolds of finite Holmes-Thompson, maximum and minimum Volume: Theorem 5 . Let (M_{1}^{n_{1}},F_{1}) and (M_{2}^{n_{2}},F_{2}) be base and fiber Finsler manifolds respectively, of the warped product M=M_{1}\times_{f}M_{2} . Suppose that M_{1},M_{2} have finite Holmes-Thompson volume, then the warped product M=M_{1}\times_{f}M_{2} also has finite Holmes-Thompson volume provided that the warping function f is bounded, with a\leq f\leq b for some a,b>0 and b<\infty . Proof. From Lemma 3.1 of sc it follows that, \operatorname{Vol}_{\operatorname{\operatorname{HT}}}(M^{n},F)=\frac{1}{\operatorname{Vol}(\mathbb{B}^{n}(1))}\int_{BM^{1}}\det g_{ab}\;\operatorname{dx}\;\operatorname{dy}, (24) where BM^{1}=\{(x,y)|F(x,y)\leq 1\} . Then clearly, F_{1}^{2}(x_{1},y_{1})+f^{2}(x_{1})F_{2}^{2}(x_{2},y_{2})\leq 1, (25) implies that BM^{1}\subseteq BM_{1}^{1}\times BM_{2}^{1/a}. (26) Therefore, \displaystyle\int_{BM^{1}}\det g_{ab}\;\operatorname{dx}\;\operatorname{dy} \displaystyle\leq \displaystyle b^{2n_{2}}\int_{BM_{1}^{1}\times BM_{2}^{1/a}}\det(g_{ij})\det(g_{\alpha\beta})\;\operatorname{dx}\;\operatorname{dy}. (27) Equivalently, by using Fubini-Tonelli Theorem, \displaystyle\frac{1}{\operatorname{Vol}(\mathbb{B}^{n}(1))\operatorname{Vol}(\mathbb{B}^{n_{1}}(1))\operatorname{Vol}(\mathbb{B}^{n_{2}}(1/a))} \displaystyle\int_{BM^{1}}\det g_{ab}\;\operatorname{dx}\;\operatorname{dy}\leq (28) \displaystyle\frac{b^{2n_{2}}}{\operatorname{Vol}(\mathbb{B}^{n}(1))}\operatorname{Vol}_{\operatorname{HT}}(M_{1}^{n_{1}},F_{1})\operatorname{Vol}_{\operatorname{HT}}(M_{2}^{n_{2}},F_{2}). The last step follows as \operatorname{Vol}_{\operatorname{HT}}(M^{n_{2}},F_{2})=\frac{1}{\operatorname{Vol}(\mathbb{B}^{n_{2}}(1/a))}\int_{BM_{2}^{1/a}}\det g_{ab}\;\operatorname{dx}\;\operatorname{dy}. Consequently, \operatorname{Vol}_{\operatorname{HT}}(M^{n},F)\leq\frac{b^{2n_{2}}\operatorname{Vol}(\mathbb{B}^{n_{1}}(1))\operatorname{Vol}(\mathbb{B}^{n_{2}}(1/a))}{(\operatorname{Vol}_{\operatorname{HT}}(M_{1}^{n_{1}},F_{1}))(\operatorname{Vol}_{\operatorname{HT}}(M_{2}^{n_{2}},F_{2}))}. Thus, the assertion follows. ∎ Theorem 6 . If (M_{1}^{n_{1}},F_{1}) and (M_{2}^{n_{2}},F_{2}) be base and fiber Finsler manifolds respectively, with finite maximum (minimum) volume, then the warped product M=M_{1}\times_{f}M_{2} has also finite maximum (minimum) volume, provided that warping function f is bounded as a\leq f\leq b for some a,b>0 . Proof. Clearly, \sigma_{\max}(x)=\max\limits_{y\in I_{x}M}\sqrt{\det(g_{ab}(x,y))}=\max\limits_{y\in{BM^{1}}}\sqrt{\det(g_{ab}(x,y))}. (29) Thus we confirm that, \operatorname{Vol}_{\max}(M,F)=\int_{M}\max\limits_{(y_{1},y_{2})\in{BM^{1}}}\sqrt{\det(g_{ab}(x_{1},x_{2},y_{1},y_{2}))}\;\operatorname{dx}. (30) Then by ( 25 ) and ( 26 ) we have, \operatorname{Vol}_{\max}(M,F)\leq b^{2n_{2}}\left(\int_{M_{1}}\max\limits_{y_{1}\in BM_{1}^{1}}\sqrt{\det(g_{ij}(x_{1},y_{1}))}\operatorname{dx}_{1}\right)\\ \left(\int_{M_{2}}\max\limits_{y_{2}\in BM_{2}^{1/a}}\sqrt{\det(g_{\alpha\beta}(x_{2},y_{2}))}\operatorname{dx}_{2}\right). (31) For minimum volume form the result can be proved similarly. ∎ Corollary 3.5 . If (M_{i}^{n_{i}},F_{i}), for i=1,2 be base and fiber Finsler manifolds respectively, of finite Holmes-Thompson volumes, and if the warped product (M,F) is a Randers manifold, then M is also of finite Busemann Haudorff volume. Proof. This follows as by Remark 3.2 (ii) p. 5 sc , it follows that \operatorname{Vol}_{BH}(M,F)\leq\operatorname{Vol}_{HT}(M,F) . ∎ 3.3 Existence of Einstein Finsler warped product of finite volume Inspired by the proof of Proposition 5 of kim , we affirm the following result which establishes the admissible class of Einstein Finsler manifolds. We recall the following identity proved in Lemma 4 ( kim , p. 2574 ) ]: \displaystyle{\operatorname{div}(H^{f})}(X)=\operatorname{Ric}(\nabla f,X)-\Delta(df)(X), (32) where \Delta=d\delta+\delta d denotes the Laplacian on base manifold acting on differential forms. Theorem 7 . Let (M_{1}^{n_{1}},g_{1}) be a Riemannian manifold of finite volume and f be a non-constant, positive smooth bounded function on M_{1} , that is a\leq f\leq b . If f satisfies ( 18 ) for a scalar function \lambda and for natural numbers n_{2}\in\mathbb{N} , then f satisfies ( 20 ) for a Ricci scalar function \mu . Hence, for a finite volume Holmes-Thompson, maximum or minimum volume, weakly Berwald Einstein Finsler manifold ({M_{2}}^{n_{2}},F_{2}) with \operatorname{Ric}_{\alpha\beta}^{M_{2}}=\mu g_{\alpha\beta} , for scalar function \mu , we can construct a finite volume Finsler warped product manifold M=M_{1}^{n_{1}}\times_{f}M_{2}^{n_{2}} with \operatorname{Ric}_{ab}=\lambda g_{ab} . Proof. Contracting both sides of ( 18 ), we have \operatorname{Scal}_{M_{1}}=n_{1}\lambda-\frac{n_{2}}{f}\Delta f. The above equation imply that \displaystyle d\operatorname{Scal}_{M_{1}}=n_{1}d\lambda+\frac{n_{2}}{f^{2}}\{\Delta fdf-fd(\Delta f)\}. (33) From the second Bianchi’s identity, we conclude that \displaystyle d\operatorname{Scal}_{M_{1}}=2\mbox{div}(\operatorname{Ric}^{M_{1}}). (34) Now, equations ( 33 ) and ( 34 ) provide that \displaystyle\mbox{div}(\operatorname{Ric}^{M_{1}})=\frac{n_{1}}{2}d\lambda+\frac{n_{2}}{2f^{2}}\{\Delta fdf-fd(\Delta f)\}. (35) From proof of Proposition 5 of kim we determine, \displaystyle\mbox{div}\left(\frac{1}{f}H^{f}\right)=-\frac{1}{2f^{2}}d(g_{1}(\nabla f,\nabla f))+\frac{1}{f}\mbox{div}(H^{f}). (36) From equations ( 32 ) and ( 18 ), we calculate \mbox{div}(H^{f})=\lambda df+\frac{n_{2}}{f}H^{f}-\Delta df. (37) After using ( 37 ) in ( 36 ), we derive \displaystyle\mbox{div}\left(\frac{1}{f}H^{f}\right) \displaystyle=\frac{1}{2f^{2}}\big[(n_{2}-1)d(g_{1}(\nabla f,\nabla f))+2\lambda fdf (38) \displaystyle-2fd(\Delta f)\big]. Now taking divergent of both side of ( 18 ), we have \displaystyle\mbox{div}(\operatorname{Ric}^{M_{1}})=\mbox{div}\left(\frac{n_{2}}{f}H^{f}\right)+\mbox{div}(\lambda g_{1}). (39) Thus from equations ( 35 ), ( 38 ) and ( 39 ), we infer that n_{2}d[(n_{2}-1)(g_{1}(\nabla f,\nabla f))-(f\Delta f)+\lambda f^{2}]+2f^{2}\operatorname{div}(\lambda g_{1})-(2n_{1}+{n_{2}})(d\lambda)f^{2})=0. Integrating the above equation yields that, \mu=(n_{2}-1)(g_{1}(\nabla f,\nabla f))-(f\Delta f)+\lambda f^{2}, (40) for some scalar function \mu . Thus, we proved the first part of this theorem. Next, for given a finite volume Holmes-Thompson, maximum or minimum volume Finslerian manifold (M_{2}^{n_{2}},F_{2}) with \operatorname{Ric}_{\alpha\beta}^{M_{2}}=\mu g_{\alpha\beta} , we can construct a finite volume Einstein Finslerian warped product manifold M=M_{1}^{n_{1}}\times_{f}M_{2}^{n_{2}} with \operatorname{Ric}_{ab}=\lambda g_{ab} , by sufficient part of Corollary 3.2 , Theorem 5 and Theorem 6 , provided that warping function f is bounded. ∎ Remark 2 . Let \mathcal{M}=\{M=M_{1}^{n_{1}}\times_{f}M_{2}^{n_{2}}|\;(M_{i}^{n_{i}},F_{i}),i=1,2,\emph{offinitevolume}\\ \emph{(withaforementionedmeasures),}M\emph{isEinsteinand\;}a\leq f\leq b\}. In summary, we have shown that the collection \mathcal{M} constitutes the admissible class of all Finsler manifolds, and henceforth, we will work with them. 4 Warped product of weakly Berwald Finsler manifolds It is interesting to know whether the warped product of two weakly Berwald or Berwald Finslerian manifold is again weakly Berwald or Berwald, respectively. We explore this in this section. Theorem 8 . If the base (M_{1}^{n_{1}},F_{1}) and fiber (M_{2}^{n_{2}},F_{2}) respectively, of the warped product are Berwald Finsler manifolds, then the warped product (M,F) is a Berwald Finsler manifold if and only if the warping function f is constant. Proof. The geodesic spray of the warped product Finsler metric ( 10 ) are given by ( wlt , p. 11) \mathbf{G}^{i}=G^{i}-\frac{1}{4}g^{ih}\frac{\partial f^{2}}{\partial x_{1}^{h}}F_{2}^{2},\quad\mathbf{G}^{\alpha}=G^{\alpha}-\frac{1}{2}\frac{1}{f^{2}}y_{1}^{h}\frac{\partial f^{2}}{\partial x_{1}^{h}}y_{2}^{\alpha}. (41) Now we use, \displaystyle\mathbf{G}^{a}_{b}:=\frac{\partial\mathbf{G}^{a}}{\partial\mathbf{y}^{b}},\quad G^{i}_{j}:=\frac{\partial G^{i}}{\partial y_{1}^{j}},\quad G^{\alpha}_{\beta}:=\frac{\partial G^{\alpha}}{\partial y_{2}^{\beta}}. Then the spray coefficients of warped product are given by, \displaystyle\mathbf{G}^{i}_{j}=G^{i}_{j}-\frac{1}{4}\frac{\partial g^{ih}}{\partial y_{1}^{j}}\frac{\partial f^{2}}{\partial x_{1}^{h}}F_{2}^{2},\quad\mathbf{G}^{i}_{\beta}=-\frac{1}{4}g^{ih}\frac{\partial f^{2}}{\partial x_{1}^{h}}\frac{\partial F^{2}}{\partial y_{2}^{\beta}}. \displaystyle\mathbf{G}^{\alpha}_{j}=\frac{1}{2}\frac{1}{f^{2}}y_{2}^{\alpha}\frac{\partial f^{2}}{\partial x_{1}^{j}},\quad\mathbf{G}^{\alpha}_{\beta}=G^{\alpha}_{\beta}+\frac{1}{2}\frac{1}{f^{2}}y_{1}^{h}\frac{\partial f^{2}}{\partial x_{1}^{h}}\delta^{\alpha}_{\beta}. Since we have, \displaystyle\mathbf{G}^{i}_{jk}=\frac{\partial}{\partial y_{1}^{k}}\Bigg(\mathbf{G}^{i}_{j}\Bigg)=G^{i}_{jk}-\frac{1}{4}\frac{\partial^{2}g^{ih}}{\partial y^{j}_{1}\,\partial y_{1}^{k}},\frac{\partial f^{2}}{\partial x_{1}^{h}}F_{2}^{2}. (42) \displaystyle\mathbf{G}^{i}_{j\alpha}=\frac{\partial}{\partial y_{2}^{\alpha}}\Bigg(\mathbf{G}^{i}_{j}\Bigg)=-\frac{1}{4}\frac{\partial g^{ih}}{\partial y_{1}^{j}}\,\frac{\partial f^{2}}{\partial x_{1}^{h}}\frac{\partial F_{2}^{2}}{\partial y_{2}^{\alpha}}. (43) \displaystyle\mathbf{G}^{i}_{\alpha\beta}=\frac{\partial}{\partial y_{2}^{\beta}}\Bigg(\mathbf{G}^{i}_{\alpha}\Bigg)=-\frac{1}{4}g^{ih}\,\frac{\partial f^{2}}{\partial x_{1}^{h}}\frac{\partial F_{2}^{2}}{\partial y_{2}^{\alpha}}. (44) \displaystyle\mathbf{G}^{\alpha}_{jk}=\frac{\partial}{\partial y_{1}^{k}}\Bigg(\mathbf{G}^{\alpha}_{j}\Bigg)=0. (45) \displaystyle\mathbf{G}^{\alpha}_{\beta\gamma}=\frac{\partial}{\partial y_{2}^{\gamma}}\Bigg(\mathbf{G}^{\alpha}_{\beta}\Bigg)=G^{\alpha}_{\beta\gamma}. (46) Now we explicitly compute the Berwald curvatures in warped product Finsler manifold. Peyghan and Tayebi have already computed the Berwald curvatures of a doubly warped product Finsler manifold in Theorem 2 , MLA . We compute them here for the sake of completeness. \displaystyle\mathbf{G}^{i}_{jkl}=\frac{\partial}{\partial y_{1}^{l}}\Bigg(\mathbf{G}^{i}_{jk}\Bigg)=G_{jkl}^{i}-\frac{1}{4}\frac{\partial^{3}g^{ih}}{\partial y_{1}^{j}\partial y_{1}^{k}\partial y_{1}^{l}}\,\frac{\partial f^{2}}{\partial x_{1}^{h}}F_{2}^{2}, (47) where l\in\{1,2,3,...,n_{1}\}. \displaystyle\mathbf{G}^{\alpha}_{\beta\gamma\delta}=G^{\alpha}_{\beta\gamma\delta}, (48) where \delta\in\{1,2,3,...,n_{2}\}. \displaystyle\mathbf{G}^{i}_{jk\alpha}=\frac{\partial}{\partial y_{2}^{\alpha}}\Bigg(\mathbf{G}^{i}_{jk}\Bigg)=-\frac{1}{4}\frac{\partial^{2}g^{ih}}{\partial y_{1}^{j}\partial y_{1}^{k}}\,\frac{\partial f^{2}}{\partial x_{1}^{h}}\frac{\partial F_{2}^{2}}{\partial y_{2}^{\alpha}}. (49) \displaystyle\mathbf{G}^{i}_{j\alpha\beta}=\frac{\partial}{\partial y_{2}^{\beta}}\Bigg(\mathbf{G}^{i}_{j\alpha}\Bigg)=-\frac{1}{4}\frac{\partial g^{ih}}{\partial y_{1}^{j}}\,\frac{\partial f^{2}}{\partial x_{1}^{h}}\frac{\partial^{2}F_{2}^{2}}{\partial y_{2}^{\alpha}\partial y_{2}^{\beta}}. (50) \displaystyle\mathbf{G}^{i}_{\alpha\beta\gamma}=\frac{\partial}{\partial y_{2}^{\gamma}}\Bigg(\mathbf{G}^{i}_{\alpha\beta}\Bigg)=-\frac{1}{4}g^{ih}\,\frac{\partial f^{2}}{\partial x_{1}^{h}}\frac{\partial^{3}F_{2}^{2}}{\partial y_{2}^{\alpha}\partial y_{2}^{\beta}\partial y_{2}^{\gamma}}. (51) \displaystyle\mathbf{G}^{\alpha}_{ijk}=0 (52) \displaystyle\mathbf{G}^{\alpha}_{ij\gamma}=0 (53) \displaystyle\mathbf{G}^{\alpha}_{i\beta\gamma}=0. (54) Clearly, (M,F) is Berwald if and only if all the Berwald curvatures vanish. Thus ( 47 )-( 54 ) all vanish. Also observe that ( 50 ) vanishes if and only if the warping function f is constant. Hence, the result follows. ∎ Corollary 4.1 . If the base (M_{1}^{n_{1}},F_{1}) and fiber (M_{2}^{n_{2}},F_{2}) respectively, of the warped product are weakly Berwald Finsler manifolds, then the warped product (M,F) is a weakly Berwald Finsler manifold, if the warping function f is constant. From Theorem 8 , Corollary 4.1 , Theorem 5 and Theorem 6 , we infer: Corollary 4.2 . If the base (M_{1}^{n_{1}},F_{1}) and fiber (M_{2}^{n_{2}},F_{2}) respectively, of the warped product are Berwald or weakly Berwald Finsler manifolds of finite Holmes-Thompson, maximum or minimum volume, then the warped product (M,F) is also Berwald or weakly Berwald Finsler manifold with respect to the same volume measure, if the warping function f is constant. 5 Einstein Finsler warped product manifold with Ricci a non-positive scalar function In this section, we investigate the obstructions to the existence of (M,F)\in\mathcal{M} , when Ricci of (M,F) is a non-positive scalar function. We will be using the following generalized maximum principle which guarantees the existence of a maximizing sequence. The more details can be found in PRS . Definition 16 ( Omori-Yau maximum principle ( PRS , p. 486 ) ) . A Riemannian manifold (M,g) is said to satisfy the Omori-Yau maximum principle, if for any smooth function u\in C^{2}(M) with \sup u<\infty , there exist a sequence \{x_{n}\} such that (i) \displaystyle{\lim_{n\to\infty}}u(x_{n})={\sup}_{M}u, (ii) ||\nabla u(x_{n})||\leq\frac{1}{n}, (iii) \Delta u(x_{n})\geq\frac{1}{n} , where \Delta u=-\operatorname{tr}(\Delta^{2}u) . Theorem 9 ( ( PRS , p. 486 ) ) . The Omori-Yau maximum principle holds on every complete Riemannian manifold with Ricci curvature bounded below. Now we prove the main theorem of this section. Theorem 10 . Let the base (M_{1}^{n_{1}},g_{1}) be a finite volume complete, connected Riemannian manifold, and the fiber (M_{2}^{n_{2}},F_{2}) be a weakly Berwald Finsler manifold, with the warped product M=M_{1}\times_{f}M_{2} . Suppose the warping function f satisfies the following conditions: (i) f is bounded as a\leq f\leq b (ii) ||\nabla f||\in L^{1}(M_{1}) and (iii) \nabla^{2}f\geq l , for some constant 00. This implies that the warping function f is a subharmonic function on the base manifold. Using the maximum principle, f is constant, if it attains its maximum or if the base manifold is compact. ∎ Under the hypothesis of the above theorem, the Ricci scalar function of the fiber cannot be negative. We conclude: Corollary 6.1 . Let M=M_{1}^{n_{1}}\times_{f}M_{2}^{n_{2}} be an Einstein Finslerian warped product manifold of positive Ricci scalar function with base, a compact Riemannian manifold, and fiber a weakly Berwald Finsler manifold. Then in order that the warped product M is a legitimate Einstein Finsler warped product manifold, Ricci scalar function of the fiber must be positive. Dan Dumitru delves into the specific conditions related to the scalar curvature of the base and fiber of Riemannian warped product manifolds. These conditions lead to the elegant revelation that the resulting manifold is, in fact, a product of Riemannian manifolds, as demonstrated in Theorem 2.5 of dam . We extend this result to the Finslerian settings. In view of the above corollary, it suffices to explore obstructions to the triviality of Finslerian warped product manifolds when \operatorname{Scal}_{M} and \operatorname{Scal}_{M_{2}} are strictly positive. Theorem 13 . Let M=M_{1}^{n_{1}}\times_{f}M_{2}^{n_{2}} be an Einstein warped product Finsler manifold with positive scalar curvature. Let the base be a Riemannian manifold and fiber be a weakly Berwald Finsler manifold with positive scalar curvature. Suppose that any of the following conditions holds, then the warping function is either subharmonic or superharmonic. Consequently, if the base manifold is compact or if either f attains its maximum or minimum, then it is a constant function. In this case, the warped product M=M_{1}\times_{f}M_{2} is a weakly Berwald Finsler manifold, which is simply a Riemannian product of two Finslerian manifolds. (a) There exist a point \tilde{x_{2}}\in M_{2} such that \|\nabla f(x_{1})\|\geq\sqrt{\frac{\operatorname{Scal}_{M_{2}}(\tilde{x_{2}})}{n_{2}(n_{2}-1)}} for every x_{1} , (b) \operatorname{Scal}_{M}\leq\operatorname{Scal}_{M_{1}} , (c) \operatorname{Scal}_{M}-\operatorname{Scal}_{M_{2}}\geq\operatorname{Scal}_{M_{1}} and \frac{\operatorname{Scal}_{M_{2}}(x_{2})}{\operatorname{Scal}_{M}(x_{1},x_{2})}\geq\frac{n_{2}}{(n_{1}+n_{2})} , for all (x_{1},x_{2})\in M_{1}\times M_{2} , (d) \operatorname{Scal}_{M}-\dfrac{\operatorname{Scal}_{M_{2}}}{f^{2}}\geq\operatorname{Scal}_{M_{1}} , (e) \operatorname{Scal}_{M_{1}}\leq 0 , (f) \operatorname{Scal}_{M_{1}}\geq\dfrac{n_{1}\operatorname{Scal}_{M_{2}}(x_{2})}{n_{2}f^{2}(x_{1})} . Proof. (a) For (x_{1},x_{2})\in M availing ( 20 ), -f(x_{1})\Delta f(x_{1})+(n_{2}-1)\|\nabla f(x_{1})\|^{2}+\lambda(x_{1},x_{2})f^{2}(x_{1})=\mu(x_{2}). Then for x_{2}=\tilde{x_{2}} , the above equation implies, -f(x_{1})\Delta f(x_{1})+\lambda(x_{1},\tilde{x_{2}})f^{2}(x_{1})=\mu(\tilde{x_{2}})-(n_{2}-1)\|\nabla f(x_{1})\|^{2}\leq 0, by using the hypothesis (a) which can be written as \|\nabla f(x_{1})\|\geq\sqrt{\frac{\mu(\tilde{x_{2}})}{n_{2}-1}} . This yields, f(x_{1})\Delta f(x_{1})\geq\lambda(x_{1},\tilde{x_{2}})f^{2}(x_{1}). Hence \Delta f>0 so warping function f is constant, as base manifold is compact. (b) By ( 21 ), \operatorname{Scal}_{M_{1}}=n_{1}\lambda(x_{1},x_{2})-n_{2}\frac{\Delta f(x_{1})}{f(x_{1})}, equivalently \operatorname{Scal}_{M_{1}}+n_{2}\lambda(x_{1},x_{2})=(n_{1}+n_{2})\lambda(x_{1},x_{2})-n_{2}\frac{\Delta f(x_{1})}{f(x_{1})}. Consequently, \operatorname{Scal}_{M}-\operatorname{Scal}_{M_{1}}=n_{2}\left(\lambda(x_{1},x_{2})+\frac{\Delta f(x_{1})}{f(x_{1})}\right), which implies \dfrac{\Delta f(x_{1})}{f}+\lambda(x_{1},x_{2})\leq 0, by using assumption (b). Hence \Delta f\leq 0 , so warping function f is constant. (c) By ( 22 ), \displaystyle\operatorname{Scal}_{M_{2}}=n_{2}\mu(x_{2})\geq n_{2}\lambda(x_{1},x_{2}), by availing the part of the supposition (c) which is \mu(x_{2})\geq\lambda(x_{1},x_{2}) for all (x_{1},x_{2})\in M_{1}\times M_{2} . Now by using ( 21 ) and the above equation, \operatorname{Scal}_{M_{1}}+\operatorname{Scal}_{M_{2}}\geq(n_{1}+n_{2})\lambda(x_{1},x_{2})-n_{2}\frac{\Delta f(x_{1})}{f(x_{1})}, thus, \operatorname{Scal}_{M_{1}}+\operatorname{Scal}_{M_{2}}\geq\operatorname{Scal}_{M}-n_{2}\frac{\Delta f(x_{1})}{f(x_{1})}, which gives, n_{2}\frac{\Delta f(x_{1})}{f(x_{1})}\geq\operatorname{Scal}_{M}-(\operatorname{Scal}_{M_{1}}+\operatorname{Scal}_{M_{2}})\geq 0, by hypothesis in (c) . Hence, \Delta f\geq 0 so warping function f is constant. (d) By ( 23 ), we have \operatorname{Scal}_{M_{1}}+\dfrac{\operatorname{Scal}_{M_{2}}}{f(x_{1})^{2}}=\operatorname{Scal}_{M}+n_{2}(n_{2}-1)\dfrac{\|\nabla f\|^{2}}{f(x_{1})^{2}}-2n_{2}\frac{\Delta f(x_{1})}{f(x_{1})}. Hence, \operatorname{Scal}_{M}-\left(\operatorname{Scal}_{M_{1}}+\dfrac{\operatorname{Scal}_{M_{2}}}{f(x_{1})^{2}}\right)=2n_{2}\frac{\Delta f(x_{1})}{f(x_{1})}-n_{2}(n_{2}-1)\dfrac{\|\nabla f\|^{2}}{f(x_{1})^{2}}\geq 0, which employing hypothesis in (d), 2n_{2}\frac{\Delta f(x_{1})}{f(x_{1})}\geq n_{2}(n_{2}-1)\dfrac{\|\nabla f\|^{2}}{f(x_{1})^{2}}\geq 0. This yields, \Delta f\geq 0 so warping function f is constant. (e) By ( 21 ), \displaystyle n_{2}\dfrac{\Delta f(x_{1})}{f(x_{1})} \displaystyle=n_{1}\lambda(x_{1},x_{2})-\operatorname{Scal}_{M_{1}}\geq 0, by hypothesis in (e) . That is, \Delta f\geq 0 so warping function f is constant. (f) By ( 21 ) we have, \operatorname{Scal}_{M_{1}}=n_{1}\lambda(x_{1},x_{2})-n_{2}\dfrac{\Delta f(x_{1})}{f(x_{1})}, thus, \operatorname{Scal}_{M_{1}}f^{2}(x_{1})=n_{1}\lambda(x_{1},x_{2})f^{2}(x_{1})-n_{2}f(x_{1})\Delta f(x_{1}).\\ Now by using ( 20 ), \operatorname{Scal}_{M_{1}}f^{2}(x_{1})-n_{1}\mu(x_{2})=(n_{1}-n_{2})f(x_{1})\Delta f(x_{1})-n_{1}(n_{2}-1)\|\nabla f(x_{1})\|^{2}\geq 0, by hypothesis in (f) . Finally, we affirm that (n_{1}-n_{2})f(x_{1})\Delta f(x_{1})\geq n_{1}(n_{2}-1)\|\nabla f(x_{1})\|^{2}\geq 0. Here the following three cases occur. (i) n_{1}=n_{2} , then clearly \|\nabla f(x_{1})\|=0, and the warping function f must be constant as base manifold is connected. (ii) n_{1}\geq n_{2} , then clearly, \Delta f\geq 0, so warping function f is constant. (iii) n_{1}\leq n_{2} , then \Delta f\leq 0 consequently, warping function f is constant. ∎ Final Conclusions: In summary, for the base (M_{1}^{n_{1}},g_{1}) Riemannian and the fiber (M_{2}^{n_{2}},F_{2}) a weakly Berwald Finsler manifold, whether Einstein warped product (M,F) is trivial or not is depicted in the following tables. Thus, we have answered the Besse’s question partially in the Finslerian settings as follows. (i) The base (M_{1}^{n_{1}},g_{1}) is a complete, Riemannian manifold of finite volume with warping function bounded. Dimension of Warped Product Ricci scalar Function \lambda of M Ricci Scalar Function \mu of M_{2} Triviality of M Any Non-positive Constant Any Constant Trivial n\geq 3 (Randers) Non-positive Scalar Function Any Scalar Function Trivial (ii) The base (M_{1}^{n_{1}},g_{1}) is a compact Riemannian manifold. Dimension of Warped product Ricci Scalar Function \lambda of M Ricci Scalar Function \mu of M_{2} Triviality of M 1-3 Positive (or Negative) Scalar Function Any Scalar Function Trivial n_{1},n_{2}\geq 2 Positive Scalar Function Any Scalar Function Nontrivial Any Non-negative Scalar Function Non-positive Scalar Function at One Point Trivial Any Non-negative Scalar Function Non-negative Scalar Function If (a)-(f) Holds of Theorem 13 , then Trivial Declarations • Funding: Not applicable. • Conflict of interest/Competing interests: The authors have no conflict of interest and no financial interests for 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. • Authors’ contributions: Conceptualization, methodology, investigation, validation, writing-original draft, review, editing, and reading have been performed by all the authors of the paper.