## On the generalization of Theorems from Riemannian Geometry to Finsler Geometry

### November 27, 2006

Ricardo Gallego Torrome Mathematisches Institut der Universität Bonn, Beringstraße 1, D-53115 Bonn, Germany
Abstract
In the present work the construction found in [1] is used to obtain the Finslerian version of some Riemannian results. Among them are Hopf-Rinow's theorem, de Rham's theorem on reducibility of Riemannian structures, the Cartan-Hadamard theorem, Schur's lemma, an “Homeomorphic” Sphere theorem, a “Diffeomorphic” Sphere theorem for dimension $3$  and a generalization of the Gauss-Bonnet-Chern theorem for Finsler structures. We formulate a general procedure to generate global “Riemannian” results in Finsler Geometry.

1 Introduction

In reference [2], important results and methods are translated from Riemannian to Finsler geometry.
Although making emphasis on quantities living in $\mathbf{S}\mathbf{M}$  , the sphere bundle associated to the manifold $\mathbf{M}$  , the proofs are very similar to the Riemannian case but adapted to the Finsler geometry characteristics like non-symmetry of the Finslerian norm $F$  .
Several questions arise: what is the nature of these similar results in Riemannian and Finsler geometry?
Are all the Riemannian results valid also in the finslerian generalization? $\text{1}$  If not, which are suitable to be generalized from Riemannian geometry to Finsler geometry? Several of these questions start from the born and also departure of both, Finsler and Riemannian geometry and some of them are partially addressed in the present work.
It was performed in [1] the construction of a Riemannian metric and its Levi-Civita's connection in terms of the initial Finsler structure, its associated Chern's connection and the canonical projections associated with it. In addition, some structural results were found relating curvatures and parallel transports. These results are rather general and they provide the possibility to generalize Riemannian theorems to the Finsler category in a systematic way. Some insights in the above questions are argued in this work.
The main procedure is based on the existence of properties and notions which are independent of the “details” of the Finsler structure, depending only on the Riemannian “skeleton” and also being true for the initial Finsler structure.
If our approach it is enough or not to answer the above questions depends strongly on whether an arbitrary Riemannian metric can be considered as the skeleton for a Finsler metric, a question not addressed here. We think that the methods for that should be rather different to the one presented in this note.
In addition, we note that the possible failure of the above “invariance” property could be a source for strictly Finslerian results. This fact will be partially explored in this work.

$\text{1}$  There are several few deep essential Finslerian results. One of them is the counterexample found by Busemann concerning a Riemannian result of Burago and Ivanov. (For an introduction and references see [4])

The structure of this note is the following: in Section $2$  , we introduce the notation and recall the structure theorems of [1]. In Section $3$  , we relate the notions of complete Finsler manifold $\left(\mathbf{M},F\right)$  with the “average” complete Riemannian manifold $\left(\mathbf{M},h\right)$  . We give proofs for the Hopf-Rinow theorem for Finsler manifolds, a Finslerian version of de Rham's theorem concerning reducibility of structures and other “metric” results.
We introduce the notion of convex invariance and explore some examples of convex invariance properties and notions. Among them are completeness, convex envelope and the notion of center of mass. In Section $4$  , a geometric interpretation for the current flag curvature is presented and a new flag curvature is defined.
The utility of this curvature is argued and related when it is possible with the current flag curvature ([2]).
We “average” the new flag curvature obtaining the sectional curvature of $h$  . The behavior of the exponential map under the average operation is also partially explored. As consequence, the Cartan-Hadamard theorem and Schur's lemma for Finsler Geometry are proved. We present an “Homeomorphism” Sphere theorem and a “Diffeomorphism” Sphere theorem for 3-dimensional Finsler structures that are Finslerian versions of the Riemannian results. A general Gauss-Bonnet-Chern's Theorem for Finsler structures is also presented and discussed. Section $5$  is dedicated to few general remarks concerning the general implications of the present note.

2 The Levi-Civita connection of a Finsler Geometry and Structure Theorems

The main sources for the definitions and notations used in this section are [1] and [2]. Let $\left(x,\mathbf{U}\right)$  be a local coordinate system over the point $x\in \mathbf{M}$  , where $x\in \mathbf{U}$  have local coordinates $\left({x}^{1},...,{x}^{n}\right)$  and $\mathbf{U}\subset \mathbf{M}$  is an open set. A tangent vector at the point $x\in \mathbf{M}$  is denoted by $y={y}^{i}\frac{\partial }{\partial {x}^{i}},{y}^{i}\in \mathbf{R}$  (we use Einstein's convention for equal upper and lower repeated indices). We can identify the point $x$  with its coordinates $\left({x}^{1},...,{x}^{n}\right)$  and the tangent vector $y\in {\mathbf{T}}_{x}\mathbf{M}$  with its components, $y=\left({y}^{1},...,{y}^{n}\right)$  .

Let us consider the split tangent bundle $\mathbf{N}:=\mathbf{T}\mathbf{M}\\left\{0\right\}$  . The fundamental notion in Finsler geometry is contained in the following
Definition 2.1 A Finsler structure $F$  on the manifold $\mathbf{M}$  is a non-negative, real function $F:\mathbf{T}\mathbf{M}\to \left[0,\infty \left[$  such that
• 1. It is smooth in the split tangent bundle $\mathbf{N}$  .
• 2. Positive homogeneity holds: $F\left(x,\lambda y\right)=\lambda F\left(x,y\right)$  for every $\lambda >0$  .
• 3. Strong convexity holds: the Hessian matrix  $\begin{array}{c}{g}_{ij}\left(x,y\right):=\frac{1}{2}\frac{{\partial }^{2}{F}^{2}\left(x,y\right)}{\partial {y}^{i}\partial {y}^{j}}\end{array}$ (2.1)
is positive definite in $\mathbf{N}$  .
The matrix ${g}_{ij}\left(x,y\right)$  is the matrix of the fundamental tensor or metric tensor $g$  .
Definition 2.2Let $\left(\mathbf{M},F\right)$  be a Finsler structure and $\left(\left(x,y\right),\mathbf{U}\right)\right)$  a local coordinate system on $\mathbf{T}\mathbf{M}$  . The Cartan tensor components are defined by the set of functions $\left(\left[2\right]\right)$  ,
 $\begin{array}{c}{A}_{ijk}=\frac{F}{2}\frac{\partial {g}_{ij}}{\partial {y}^{k}},i,j,k=1,...,n.\end{array}$ (2.2)
These are homogeneous functions of degree zero in $y$  . In the Riemannian case, ${A}_{ijk}$  are zero and this fact characterizes Riemannian geometry from other types of Finsler geometries.

Since the components of the fundamental and Cartan's tensors have a dependence on ${y}^{i}$  , it is natural to use other manifold than $\mathbf{M}$  in order to study their properties. One possibility is based on the study of the vector bundle ${\pi }^{*}\left(\mathbf{T}\mathbf{M}\right)$  induced by the projection([2])
$\pi :\mathbf{N}\mathbf{⟶}\mathbf{M}.$
This is the natural bundle to work in Finsler geometry, at least if we work with Chern's connection.
There are other projections which we will use, namely
${\pi }_{1}:\mathbf{N}×\mathbf{T}\mathbf{M}\mathbf{⟶}\mathbf{N},{\pi }_{2}:\mathbf{N}×\mathbf{T}\mathbf{M}\mathbf{⟶}\mathbf{T}\mathbf{M}$
and the corresponding projections induced from the above on the bundle ${\pi }^{*}\mathbf{T}\mathbf{M}\subset \mathbf{N}×\mathbf{T}\mathbf{M}$  which are also denoted by ${\pi }_{1}$  and ${\pi }_{2}$  .

There is a non-linear connection in the bundle $\pi :\mathbf{N}\mathbf{⟶}\mathbf{M}$  . This non-linear connection is an invariant horizontal distribution $\mathcal{ℋ}$  such that the following $\mathbf{G}\mathbf{L}\left(2n,\mathbf{R}\right)$  -invariant decomposition holds:
$\mathbf{T}\mathbf{N}=\mathcal{V}\oplus \mathcal{ℋ},$
where $\mathcal{V}:=kernd\pi$  is the vertical distribution.

After these notions, the characterization of Chern's connection in terms of horizontal and vertical components is given by the following proposition ([1]). Let us denote by $V\left(\nabla g\right)$  and by $H\left(\nabla g\right)$  the vertical and horizontal projections. Then,
Proposition 2.3Let $\left(\mathbf{M},F\right)$  be a Finsler structure. Then Chern's connection is characterized by the conditions:
• 1. Almost $g$  -compatibility condition $\left(\left[2\right]\right)$  , equivalent to the equations  $\begin{array}{c}V\left(\nabla g\right)=2A,H\left(\nabla g\right)=0.\end{array}$ (2.3)
• 2. Torsion free condition $\left(\left[2\right]\right)$  , equivalent to the relations
• (a) Null vertical covariant derivative of sections of ${\pi }^{*}\left(\mathbf{T}\mathbf{M}\right)$  : let $\overline{X}\in \mathbf{T}\mathbf{N}$  and $Y\in \mathbf{T}\mathbf{M}$  , then  $\begin{array}{c}{\nabla }_{V\left(\overline{X}\right)}{\pi }^{*}Y=0.\end{array}$ (2.4)
• (b) Let us consider $X,Y\in \mathbf{T}\mathbf{M}$  and the associated horizontal vector fields $\stackrel{~}{X}$  and $\stackrel{~}{Y}$  with horizontal components ${X}^{i}$  and ${Y}^{i}$  . Then  $\begin{array}{c}{\nabla }_{\stackrel{~}{X}}{\pi }^{*}Y-{\nabla }_{\stackrel{~}{Y}}{\pi }^{*}X-{\pi }^{*}\left(\left[X,Y\right]\right)=0.\end{array}$ (2.5)

Definition 2.4Let $\left(\mathbf{M},F\right)$  be a Finsler structure. The subset ${\mathbf{I}}_{x}\subset {\mathbf{T}}_{x}\mathbf{M}$  defined by
${\mathbf{I}}_{x}:=\left\{y\in {\mathbf{T}}_{x}\mathbf{M}|F\left(x,y\right)=1\right\}$
is the indicatrix over the point $x\in \mathbf{M}$  .
In order to define the “average” metric $h$  we consider the measure in ${\mathbf{I}}_{x}$  defined by the Riemannian metric $\stackrel{~}{g}$  on ${\mathbf{I}}_{x}$  induced from $\left({\mathbf{T}}_{x}\mathbf{M}\\left\{0\right\},{g}_{x}\right).$  Therefore The pair $\left({\mathbf{I}}_{x},\stackrel{~}{g}\right)$  is a Riemannian sub-manifold of $\left({\mathbf{T}}_{x}\mathbf{M}\\left\{0\right\},{g}_{x}\right)$  .
Definition 2.5Let $f\in \mathcal{ℱ}\left({\mathbf{I}}_{x}\right)$  be a real function on the indicatrix ${\mathbf{I}}_{x}$  and ${\psi }^{2}$  an invariant measure respect Chern's connection. We define the map
$<,{>}_{{\psi }^{2}}:\mathcal{ℱ}\left({\mathbf{I}}_{x}\right)⟶\mathbf{R}$
given by
 $\begin{array}{c}}_{{\psi }^{2}}:={\int }_{{\mathbf{I}}_{x}}dvo{l}_{g}{\psi }^{2}\left(x,y\right)f\left(x,y\right).\end{array}$ (2.6)
The norm of the measure function ${\psi }^{2}$  is such that $<1{>}_{{\psi }^{2}}=1$  .

We define now the Riemannian metric $h$  . First we fix the measure by putting, ${\psi }^{2}={F}^{2}$  ,
Definition 2.6Let $\left(\mathbf{M},F\right)$  be a Finsler structure. Let us define the matrix coefficients by
 $\begin{array}{c}{h}_{ij}\left(x\right):=<{g}_{ij}\left(x,y\right)>,\forall x\in \mathbf{M}.\end{array}$ (2.7)
The first structure result associates a Riemannian structure $\left(\mathbf{M},h\right)$  to the Finsler structure $\left(\mathbf{M},F\right)$  ([1]),
Proposition 2.7Let $\left(\mathbf{M},F\right)$  be a Finsler structure. Then the coefficients $\left\{{h}_{ij}\left(x\right),i,j=1,...,n\right\}$  are the components of a Riemannian metric in $\mathbf{M}$  such that in a local coordinate system $\left(x,\mathbf{U}\right)$
 $\begin{array}{c}h\left(x\right)={h}_{ij}d{x}^{i}\otimes d{x}^{j}.\end{array}$ (2.8)
The relation between Chern's connection associated to the Finsler structure $\left(\mathbf{M},F\right)$  and the Levi-Civita connection of $h$  is the following ([1]):
Lemma 2.8Let $\left(\mathbf{M},F\right)$  be a Finsler structure and $\pi \left(u\right)=x$  .
Then there is defined on $\mathbf{M}$  a covariant derivative ${\stackrel{~}{\nabla }}_{X}$  characterized by the conditions:
• 1. $\forall X\in {\mathbf{T}}_{x}\mathbf{M}$  and $Y\in \mathbf{\Gamma }\left(\mathbf{M}\right)$  , the covariant derivative of $Y$  in the $X$  direction is  $\begin{array}{c}\left({\stackrel{~}{\nabla }}_{X}Y\right){|}_{x}=<{\pi }_{2}{|}_{v}{\nabla }_{{\stackrel{~}{X}}_{u}}{\vartheta }_{v}{\pi }_{v}^{*}\left(Y\right){>}_{u},\forall v\in {\mathbf{U}}_{u},\end{array}$ (2.9)
being ${\mathbf{U}}_{u}$  an open set containing $u$  .
• 2. For every function $f\in \mathcal{ℱ}\left(\mathbf{M}\right)$   $\begin{array}{c}\left({\stackrel{~}{\nabla }}_{X}f{\right)}_{x}=<{\pi }_{2}{|}_{v}{\nabla }_{{\stackrel{~}{X}}_{u}}{\vartheta }_{v}{\pi }_{v}^{*}f{>}_{u}.\end{array}$ (2.10)
We remark that due to the properties of the $\vartheta$  functions, equations $2.9$  and $2.10$  can be written with ${\pi }_{u}^{*}$  instead of ${\pi }_{u}^{*}$  inside the integral operation.

The linear application ${\vartheta }_{v}$  is determined by the canonical projections and it is defined as:
${\vartheta }_{v}:{\pi }_{v}^{*}\left(\mathbf{\Gamma }\left(\mathbf{M}\right)\right)⟶{\mathbf{\Gamma }}_{v}\left({\pi }_{v}^{*}\left(\mathbf{\Gamma }\mathbf{M}\right)\right)$
 $\begin{array}{c}{\vartheta }_{v}\left({T}_{vu}{S}_{u}\right)={S}_{v}\end{array}$ (2.11)
for every tensor ${S}_{v}\in {\pi }_{v}^{*}\left(\mathbf{\Gamma }\left(\mathbf{M}\right)\right)$  . The relevance of this function is that it reduces the calculations when products of several averages appear $\text{2}$  : the product of the“average” of two local operators (depending only on $u\in {\mathbf{I}}_{x}$  ) is the average of the product,
$=.$
Theorem 2.9Let $\left(\mathbf{M},F\right)$  be a Finsler structure. Consider the linear connection $\stackrel{~}{\nabla }$  on $\mathbf{M}$  .
Then $\stackrel{~}{\nabla }$  is the Levi-Civita connection of $h.$
The behavior of parallel transport under the “average” is given through the following result([1]),
Theorem 2.10Let $\left(\mathbf{M},F\right)$  be a Finsler structure. Let $\gamma$  be a path in M starting at $x$  and ${\stackrel{~}{\gamma }}_{u}$  the horizontal lift in N starting at $u,\pi \left(u\right)=x.$  Then the following relation holds:
 $\begin{array}{c}{\stackrel{~}{\tau }}_{{\gamma }_{x}}S=<{\pi }_{2}{|}_{\stackrel{~}{\gamma }\left(1\right)}{\tau }_{{\stackrel{~}{\gamma }}_{v}}{\vartheta }_{v}{\pi }_{v}^{*}S{>}_{v},\forall S\in {\mathbf{\Gamma }}_{x}\left(\mathbf{M}\right).\end{array}$ (2.12)
A similar result holds for the $hh$  -curvature endomorphisms with the Riemannian endomorphisms from $h$  :
Theorem 2.11Let $\left(\mathbf{M},F\right)$  be a Finsler structure. Let ${\stackrel{~}{X}}_{1},{\stackrel{~}{X}}_{2}$  be the horizontal lifts in ${\mathbf{T}}_{u}\mathbf{N}$  of the linear independent vectors ${X}_{1},{X}_{2}\in {\mathbf{T}}_{x}\mathbf{M}$  . Then for every section $Y\in \mathbf{\Gamma }\mathbf{M}$  ,
 $\begin{array}{c}{\stackrel{~}{R}}_{x}\left({X}_{1},{X}_{2}\right)Y=<{\pi }_{2}{R}_{u}\left({\stackrel{~}{X}}_{1}{|}_{u},{\stackrel{~}{X}}_{2}{|}_{u}\right){\vartheta }_{v}{\pi }_{u}^{*}Y{>}_{u}.\end{array}$ (2.13)

$\text{2}$  It could be interesting to rewrite a measure with such property in terms of distribution. One possibility one can try is a measure of the type

$d\mu :={lim}_{k\to \infty }{\prod }_{k}\delta \left({\xi }_{k}-{y}_{k}\right)d{\xi }_{k},\text{with}{y}_{k}\in {\mathbf{I}}_{x}.$
The complexity of this measure is apparent. However, the product of delta functions implies a strong locality property. Nevertheless, some problems appear with this measure, for instance, diffeomorphism invariance or the continuous limit. Apparently, an affine structure is also needed in the manifold ${\mathbf{I}}_{x}$  to define this measure, which is not obvious from the given Finsler structure.
On the other side, this kind of measures can be useful in field theory in physics. They are so strongly local that they can be useful to describe deterministic systems using path integrals. In addition, they can be the source for a path integral formulation of physical theories containing a maximal acceleration.

3 Generalization of Metric Theorems from Riemannian Geometry to Finsler Geometry

We start this section proving a version of the Hopf-Rinow theorem for Finsler spaces. Not all the statements as in the classical theorem ([2]) are considered, but just the statement relating geodesic completeness with metric completeness and forward bounded sets. After this, a Finslerian version of de Rham's theorem concerning the reducibility of Finslerian structures is also obtained. Additional results concerning the metric geometry of a Finsler space are considered. This serves as a test of the method of generalizing results from Riemannian to Finsler geometry. We present the general procedure to generalize Riemannian results to the Finsler category. All the results are obtained without using the exponential map, which is briefly discussed in the next section.
Definition 3.1(Definition of geodesic ) Let ${\nabla }_{\stackrel{~}{T}}$  denote the covariant derivative of the Chern connection, $T$  an unitary tangent vector along the curve $\gamma \left(s\right)\subset M$  and $\stackrel{~}{T}$  its horizontal lift in TN. This curve is geodesic when at any point $x\left(t\right)$  of the curve $\gamma \left(s\right)$  the following equation holds,
 $\begin{array}{c}{\nabla }_{\stackrel{~}{T}}{\pi }^{*}T=0.\end{array}$ (3.1)
As consequence, the following equation holds for $t\in I\subset \mathbf{R}$  in local coordinates and with ${T}^{i}=\frac{d{\gamma }^{i}}{ds}:$
 $\begin{array}{c}\frac{{d}^{2}{\gamma }^{i}}{d{s}^{2}}+{\Gamma }_{jk}^{i}\left(x,T\right)\frac{d{\gamma }^{j}}{ds}\frac{d{\gamma }^{k}}{ds}=0,i,j,k=1,...,n.\end{array}$ (3.2)
When $I$  can be extended to $\left[a,+\infty \right)$  the geodesic is called forward complete; if $I$  can be extended to $\left(-\infty ,a\right]$  , the geodesic is called backward complete.
We must note that this definition is apparently different than the usual one ([2]), where geodesic are defined through the first variation of the arc-length. However in local coordinates this condition reproduces the same equation $\left(3.2\right)$  and therefore we should consider both definitions as equivalent.
Proposition 3.2For every forward complete geodesic in $\left(\mathbf{M},F\right)$  corresponds a complete geodesic in $\left(\mathbf{M},h\right)$  .
Proof: Consider the geodesic equation ${\nabla }_{\stackrel{~}{T}}{\pi }^{*}T=0$  , where $\stackrel{~}{T}$  is the horizontal lift of $T$  in $\mathbf{T}\mathbf{N}$  . The corresponding geodesic equation for the Riemannian structure $\left(\mathbf{M},h\right)$  is just ${\stackrel{~}{\nabla }}_{T}T=0$  . It follows from equation $\left(2.10\right)$  that
$\left({\stackrel{~}{\nabla }}_{T}T\right){|}_{x}=<{\pi }_{2}{|}_{v}{\nabla }_{{\iota }_{u}\left(T\right)}{\vartheta }_{v}{\pi }_{u}^{*}T{>}_{u}=0.$
Since this happens for every point of the geodesic for arbitrary value $s$  , it follows the forward completeness of the geodesic in the Riemannian sense. However, a Riemannian geodesic is invariant under $s\to -s$  (that means that if $\gamma \left(s\right)$  lies on the geodesic, also $\gamma \left(-s\right)$  ), from which follows completeness. $\square$  Indeed the above proof shows that any geodesic in $\left(\mathbf{M},F\right)$  corresponds to a geodesic curve of $\left(\mathbf{M},h\right)$  , although not necessarily the same curve, because of the reflection property of Riemannian geodesics does not necessarily holds in the Finsler case.

A useful property will be the following:
Proposition 3.3Consider the average of the metric coefficients $<{g}_{ij}>$  and the line integral ${\int }_{p}^{q}\left({g}_{ij}{T}^{i}{T}^{j}{\right)}^{\frac{1}{2}}$  along a path joining the points $p$  and $q$  .
Then, they commute in the sense that:
${\int }_{p}^{q}\left(<{g}_{ij}\left(x,T\right){T}^{i}{T}^{j}{>}_{u}{\right)}^{\frac{1}{2}}\sim {\int }_{p}^{q}\left(<{g}_{ij}\left(x,u\right){>}_{u}{T}^{i}{T}^{j}{\right)}^{\frac{1}{2}}\sim <{\int }_{p}^{q}\left({g}_{ij}\left(x,u\right){T}^{i}{T}^{j}{\right)}^{\frac{1}{2}}{>}_{u}.$
Proof: The equivalence relation $\sim$  here means that if one side is arbitrarily bounded, the others are also arbitrarily bounded. It is a consequence of the compactness of the manifold ${\mathbf{I}}_{x}$  . The second relation comes from the positive scaling invariance $\left(2.1.2\right)$  of the integrand. It makes the average independent of the point $x\in \mathbf{M}$  , except for a bounded factor coming from the Jacobian of the transformation, even if M is non-compact. In particular, the change in the integration from ${\mathbf{I}}_{x}$  to ${\mathbf{S}}^{n-1}$  implies
${\int }_{{\mathbf{I}}_{x}}\left({g}_{ij}\left(x,u\right){T}^{i}{T}^{j}{\right)}^{\frac{1}{2}}={\int }_{{\mathbf{S}}^{n-1}}|jac\left(\phi \right)|\left({g}_{ij}\left(x,u\right){T}^{i}{T}^{j}{\right)}^{\frac{1}{2}},$
where $|jac\left(\phi \right)|$  is the Jacobian of the change of labels. $\square$  Indeed it is interesting to consider when the above property is equivalent to the Lipschitz condition. This could have some applications in quantum physics.

We recall some metric notions in Finsler geometry. The metric distance ${d}_{F}$  associated with the Finsler structure $F$  between two points $p$  and $q$  of the manifold $\mathbf{M}$  is defined by
${d}_{F}\left(p,q\right)=inf\left\{{\int }_{p}^{q}\parallel T{\parallel }_{F}\right\}.$
The integrals are performed for each piecewise smooth curves connecting $p$  and $q$  . The sequence $\left\{{x}_{i}\right\}$  is a forward Cauchy sequence if for every $\epsilon$  exists an integer $n$  such that when $n  , then ${d}_{F}\left({x}_{i},{x}_{j}\right)<\epsilon$  .
A space is forward complete if every forward Cauchy sequence is convergent. A forward metric ball ${\mathbf{B}}_{p}^{+}\left(r\right)$  centered at $p$  and with radii $r$  is defined as the set ${\mathbf{B}}_{p}^{+}\left(r\right):=\left\{x\in \mathbf{M}:{d}_{F}\left(p,x\right)
Corollary 3.4(“First” Hopf-Rinow's Theorem for Finsler geometry) Let $\left(\mathbf{M},F\right)$  be a Finsler structure and $\mathbf{M}$  connected.
$\left(\mathbf{M},{d}_{F}\right)$  is forward metric complete iff $\left(\mathbf{M},F\right)$  is forward geodesic complete.
Proof: Denote by ${d}_{F}^{2}\left(p,q\right)$  and ${d}_{h}^{2}\left(p,q\right)$  the metric distances obtained from the Finsler structure $\left(\mathbf{M},F\right)$  and the Riemannian structure $\left(\mathbf{M},h\right)$  respectively. From the proof of proposition $3.2$  it follows that if $\left(\mathbf{M},F\right)$  is forward geodesic complete, $\left(\mathbf{M},h\right)$  is also forward geodesic complete and therefore, because it is a Riemannian geodesic, geodesic complete. Then, from the Riemannian Hopf-Rinow theorem, $\left(\mathbf{M},{d}_{h}\right)$  is metric-complete and in particular, forward metric complete. However if this is the case, due to the commutativity property of the average operation $<,>$  and the line integral (proposition $3.3$  ), $\left(\mathbf{M},{d}_{F}\right)$  is also forward metric complete (it is only forward complete because in the definition of the Riemannian metric $h$  , the non-symmetry property of the metric is relevant).
Suppose that $\left(\mathbf{M},{d}_{F}\right)$  is forward metric complete. Then by proposition $3.3$  $\left(\mathbf{M},{d}_{h}\right)$  is metric complete and by the Hopf-Rinow theorem, $\left(\mathbf{M},h\right)$  is geodesic complete. This condition is resumed in the equation ${\stackrel{~}{\nabla }}_{T}T=0$  , which holds for every $s\in {\mathbf{R}}^{+}$  . Therefore $g\left({\stackrel{~}{\nabla }}_{T}T,{\stackrel{~}{\nabla }}_{T}T\right)=0$  . Taking the average $=0$  and as a result of the equation $2.10$  and the properties of the average operation, we get $<<>>=0$  . Because the positive measure it is only possible if ${\nabla }_{\stackrel{~}{T}}{\pi }^{*}T=0$  . (Note that not necessarily is complete because the curve can be not well defined at $-s$  as a Finsler geodesic.) $\square$  The second half of the proof together with the proof of proposition $3.2$  does not implies that the geodesic curves of $\left(\mathbf{M},F\right)$  are a subset of the geodesic curves of $\left(\mathbf{M},h\right)$  . This difference is based on the connections used and on the reflection property of Riemannian geodesics. A counterexample where the inclusion is strict is based on the result of Skorniakov ([10]). In addition, systems of Finsler geodesic curves are obtained which are not Riemannian.

Recall the notion of forward ball ([2]): a sub-set $K\subset \mathbf{M}$  is forward bounded if it is contained in a forward metric ball. Then the following result holds,
Corollary 3.5(“Second” Hopf-Rinow's Theorem for Finsler geometry) Let $\left(\mathbf{M},F\right)$  be a Finsler structure such that $\left(\mathbf{M},{d}_{F}\right)$  is forward metric complete and $\mathbf{M}$  connected. Then every closed and forward bounded subset of $\mathbf{M}$  is compact.
Proof: If $\left(\mathbf{M},{d}_{F}\right)$  is forward complete, then $\left(\mathbf{M},h\right)$  is also forward complete. As consequence of the Hopf-Rinow theorem in Riemannian geometry ([3]), every closed and forward bounded subset of $\left(\mathbf{M},{d}_{h}\right)$  is compact. That every closed, forward bounded subset in $\left(\mathbf{M},h\right)$  is forward bounded in $\left(\mathbf{M},{d}_{F}\right)$  can be proved by contradiction. Suppose that $\left(\mathbf{M},{d}_{h}\right)$  is a forward complete metric space. Then if
${d}_{h}\left({x}_{i},{x}_{j}\right)=inf\left\{{\int }_{{x}_{i}}^{{x}_{j}}\parallel T{\parallel }_{h}\right\}<\epsilon ,{n}_{0}
using proposicion $3.3$  one obtains
$inf\left\{{\int }_{{x}_{i}}^{{x}_{j}}\parallel T{\parallel }_{h}\right\}\sim inf\left\{{\int }_{{x}_{i}}^{{x}_{j}}<\parallel T{\parallel }_{g}>\right\}\sim inf\left\{{\int }_{{x}_{i}}^{{x}_{j}}\parallel T{\parallel }_{g\left(x,T\right)}\right\}<\stackrel{~}{\epsilon },$
where $\stackrel{~}{\epsilon }$  is also arbitrary. Therefore we get the thesis of the corollary. $\square$

From these results we guess the mechanism to generalize theorems from Riemannian geometry to Finsler geometry: we pass from Finsler to Riemannian structures using the structure theorems of section $2$  , then we use Riemannian results whose thesis concerns properties independent of the geometry defined by the Finsler data $F$  and then we come-back to the Finsler category using this independence. We formalize this idea as follows:
Definition 3.6Let $\left(\mathbf{M},F\right)$  be a Finsler structure and consider the 1-parameter family of Finsler structures with fundamental tensors ${g}_{t}=\left(1-t\right)g+t$  . A property is called convex-invariant if it holds for every $t\in \left[0,1\right]$  .
Associated with $t$  we have not only a Finsler metric ${g}_{t}$  but also other geometric objects like connections and curvatures. These are all called generically Finsler quantities.
Definition 3.7Consider an arbitrary Riemannian structure $\left(\mathbf{M},h\right)$  . A property will be called Riemannian if it is completely specified from the Riemannian structure $\left(\mathbf{M},F\right)$  . An analogous notion is associated to the Finsler case.
An example of convex invariant property is a topological property, not depending of the metric, but only on the underlying topology of the manifold $\mathbf{M}$  . Other example, coming from corollary 3.4 and corollary 3.5, is the notion of forward Cauchy sequence and forward bounded subset.
The general mechanism for translate results from Finsler geometry to Riemannian geometry is the following theorem:
Theorem 3.8Let $\left(\mathbf{M},F\right)$  be a Finsler structure. Then a Riemannian property is convex invariant iff it is a Finsler property.
Proof: Consider a convex invariant property expressed in terms of Riemannian notions and quantities that are also Finsler. So we can enunciate the property for every $t$  as before and in particular for $T=0$  , which means, due to the convex invariance, that the property holds in the Finsler case. The converse statement is proved in analogous way. $\square$

As direct application we prove the generalized de Rham's theorem on reducibility of Finsler structures. A Riemannian structure is said reducible (and a similar notion for locally reducible holds) when the Riemannian metric is the product of metrics: $\left({\mathbf{M}}_{1}×{\mathbf{M}}_{2},{h}_{1}×{h}_{2}\right)$  and the norm of a vector is given by $\parallel \left({y}_{1},{y}_{2}\right){\parallel }_{h}=\parallel {y}_{1}{\parallel }_{{h}_{1}}^{2}+\parallel {y}_{2}{\parallel }_{{h}_{2}}^{2}$  . A similar notion can be applied to the finslerian case, but the Riemannian metrics ${h}_{i},i=1,2$  should be substituted by fundamental tensors ${g}_{i},i=1,2$  . Then the following result holds:
Corollary 3.9(de Rham's theorem on reducibility of Finsler structures) A complete Finslerian manifold which is locally reducible and simply connected is a Finslerian product.
Proof: Completeness is a convex invariant property, as well as simply connection. We show that reducibility, global and local are also convex invariant properties. Let us denote by ${g}_{1}×{g}_{2}$  the total fundamental tensor in ${\mathbf{M}}_{1}×{\mathbf{M}}_{2}$  and by ${g}_{t}=\left(1-t\right)g+t$  the interpolating Finsler metric. Avoiding projections in the integrals, we obtain
$<{g}_{1}×{g}_{2}>={\int }_{{\mathbf{I}}_{x}}{g}_{1}×{g}_{2}={\int }_{{\stackrel{~}{\mathbf{I}}}_{1}}{g}_{1}×{\int }_{{\stackrel{~}{\mathbf{I}}}_{2}}{g}_{2}.$
The manifold ${\stackrel{~}{\mathbf{I}}}_{1}\left({x}_{1}\right)$  are defined by
${\stackrel{~}{\mathbf{I}}}_{1}\left({x}_{1}\right):=\left\{y\in {\mathbf{T}}_{{x}_{1}}{\mathbf{M}}_{1}|{F}_{1}^{2}\left({x}_{1},{y}_{1}\right)\right\},{\stackrel{~}{\mathbf{I}}}_{2}\left({x}_{2}\right):=\left\{y\in {\mathbf{T}}_{{x}_{2}}{\mathbf{M}}_{2}|{F}_{2}^{2}\left({x}_{2},{y}_{2}\right)=1-{F}_{1}^{2}\left({x}_{1},{y}_{1}\right)\right\}.$
The points where ${F}_{1}=1$  is a set of zero measure in ${\mathbf{T}}_{{x}_{1}}{\mathbf{M}}_{1}$  and ${F}_{2}=0$  is also of measure zero in ${\mathbf{T}}_{{x}_{2}}{\mathbf{M}}_{2}$  .
This fact and the positive invariance of the fundamental tensor implies
 $\begin{array}{c}<{g}_{1}^{t}×{g}_{2}^{t}>={\int }_{{\mathbf{I}}_{1}}|jac\left({\stackrel{~}{\phi }}_{1}\right)|{g}_{1}^{t}×{\int }_{{\mathbf{I}}_{2}}|jac\left({\stackrel{~}{\phi }}_{2}|\right){g}_{2}^{t}.\end{array}$ (3.3)
$|jac\left({\stackrel{~}{\phi }}_{1}\right)|$  and $|jac\left({\stackrel{~}{\phi }}_{2}\right){|}_{2}$  are the jabobian functions from the change of labels produce by ${\stackrel{~}{\mathbf{I}}}_{1}\left({x}_{1}\right)⟶{\mathbf{I}}_{1}$  and ${\stackrel{~}{\mathbf{I}}}_{2}\left({x}_{2}\right)⟶{\mathbf{I}}_{2}$  . From this equation, we get for the Finslerian norm from ${g}_{t}$  that ${F}_{t}^{2}$  is of the form ${\stackrel{~}{F}}_{1t}^{2}+{\stackrel{~}{F}}_{2t}^{2}$  with fundamental tensors $|jac\left({\stackrel{~}{\phi }}_{1}\right)|{g}_{1}^{t}$  and $|jac\left({\stackrel{~}{\phi }}_{2}\right)|{g}_{2}^{t}$  respectively. From here it follows the convex invariance of the reducibility property. The corollary is a consequence of theorem 3.8 and the Riemannian de Rham's theorem on reducibility([4]). $\square$  This result should play a similar role to the Riemannian theorem in the problem of the classification of holonomies of Chern's connection, reducing the problem to the classification of irreducible structures.

We can define the diameter of the manifold by $dia{m}_{F}\left(\mathbf{M}\right):=sup\left\{{d}_{F}\left(p,q\right)|p,q\in \mathbf{M}\right\}$  . Then the following property holds,
Corollary 3.10A complete Finsler structure $\left(\mathbf{M},F\right)$  is compact iff $dia{m}_{F}\left(\mathbf{M}\right)<\infty$  .
Proof: $dia{m}_{F}\left(\mathbf{M}\right)<\infty$  is a convex invariant property; consider the distance function
${d}_{t}\left(p,q\right):=inf\left\{{\int }_{p}^{q}\left(\left({g}_{t}{\right)}_{ij}{T}_{i}{T}_{j}{\right)}^{\frac{1}{2}}\right\}.$
$T\left(s\right)$  is the tangent field along the particular curve. This definition implies
${d}_{t}\left(p,q\right)=inf\left\{{\int }_{p}^{q}\left(t{g}_{ij}{T}^{i}{T}^{j}+\left(1-t\right)<{g}_{ij}>{T}^{i}{T}^{j}{\right)}^{\frac{1}{2}}\right\}\sim tinf\left\{{\int }_{p}^{q}\left({g}_{ij}{T}^{i}{T}^{j}{\right)}^{\frac{1}{2}}\right\}+$
$+\left(1-t\right)inf\left\{{\int }_{p}^{q}\left(<{g}_{ij}>{T}^{i}{T}^{j}{\right)}^{\frac{1}{2}}\right\}.$
In order to prove this statement, we take the average operation:
$<{\int }_{p}^{q}\left({g}_{ij}{T}^{i}{T}^{j}{\right)}^{\frac{1}{2}}>={\int }_{p}^{q}\left(<{g}_{ij}{T}^{i}{T}^{j}>{\right)}^{\frac{1}{2}}\sim {\int }_{p}^{q}\left(<{g}_{ij}>{T}^{i}{T}^{j}{\right)}^{\frac{1}{2}}.$
that is of the same size as the second term(again using proposition $3.3$  ).
This proves the convex invariance of the finiteness of the diameter. Completeness is also convex invariant as well as compactness. The corollary is a consequence of the Riemannian result ([4]) and theorem 3.8 $\square$  The next result concerns a generalization of a result of Whitehead,
Corollary 3.11Let $\left(\mathbf{M},F\right)$  be a Finsler structure and $K\subset \mathbf{M}$  compact. Then there is a positive $\epsilon$  such that ${\mathbf{M}}_{p}^{+}\left(r\right)$  with $p\in K$  and $r<\epsilon$  is strictly convex.
Proof: We prove the convex invariance of the notion on convex metric balls for the family of Finslerian metrics defined by $\left\{{g}_{t}\right\}$  . Instead of considering the forward balls, let us consider metric forward spheres $\left\{{\mathbf{S}}_{p}^{+}\left(t,s\right)\right\}$  , where $t$  is the interpolating parameter and $s$  is a parameter labeling the spheres but not necessarily the radii function. The family $\left\{{g}_{t}\right\}$  produces the homotopy between the Riemannian sphere and the metric sphere defined for the initial Finsler structure ${g}_{t}$  . Then the metric forward ball is ${\mathbf{B}}_{p}^{+}\left(t,s\right)={\cup }_{s}{\mathbf{S}}_{p}^{+}\left(t,s\right),0  . From this decomposition, it follows that any geodesic segment contained in the Riemannian ball is transformed to another geodesic segment contained in the Finslerian ball and that this transformation is a quasi-isometry. From this follows the convex invariance of convex balls. From the Riemannian theorem ([4]) and from theorem 3.8 follows the result. $\square$  From the proof of corollary $3.11$  it follows that:
Corollary 3.12Any connected Finsler structure induces a metric space $\left(\mathbf{M},{d}_{F}\right)$  whose topology is the same as the manifold topology.
Therefore a topological property or notion is convex invariant.

Let us consider the right-center of mass of a compact sub-set $K\subset \mathbf{M}$  defined as the point minimizing the function:
$C{M}_{r}:K⟶\mathbf{R}$
$p⟶{\int }_{K}{d}_{F}^{2}\left(p,a\right)da.$
$da$  is a measure defined on $K$  . A similar notion can be defined by the use of ${d}_{F}^{2}\left(a,p\right)$  in the integration.
Let us call this new function $C{M}_{l}$  (the left function center of mass).
The same construction can be done for the interpolation metric ${g}_{t}$  and in addition let usconsider the symmetric function:
 $\begin{array}{c}p⟶\frac{1}{2}\left({\int }_{K}{d}_{t}^{2}\left(p,a\right)da+{\int }_{K}{d}_{t}^{2}\left(a,p\right)da\right).\end{array}$ (3.4)
From the definition of the interpolating metric ${g}_{t}$  , the above integral can be decomposed in a Riemannian and non-Riemannian components, denoted by $C{M}_{0}$  and $\delta CM$  :
$\frac{1}{2}\left(C{M}_{r}+C{M}_{l}\right)\left(t\right)=C{M}_{0}+\delta CM,C{M}_{0}\left(t\right):=t{\int }_{K}{d}_{h}^{2}\left(p,a\right)da,$
 $\begin{array}{c}\delta CM:=\frac{1}{2}\left(1-t\right)\left({\int }_{K}{d}_{t}^{2}\left(p,a\right)da+{\int }_{K}{d}_{t}^{2}\left(p,a\right)da\right).\end{array}$ (3.5)
From equations $3.4$  and $3.5$  it follows that $\frac{\partial }{\partial {x}^{i}}\left(\frac{1}{2}\left(C{M}_{r}+C{M}_{l}\right)\left(t\right)\right)=0$  iff $\frac{\partial }{\partial {x}^{i}}C{M}_{0}=0$  . However, by a theorem of Cartan ([4]) this is the case. Therefore,
Theorem 3.13( Existence of the center of mass) Let $\left(\mathbf{M},F\right)$  be a Finsler manifold and let $K\subset \mathbf{M}$  a compact sub-set. Then there is a point ${p}_{1}$  minimizing the function
$\frac{1}{2}\left(C{M}_{r}+C{M}_{l}\right)\left(t\right):K⟶\mathbf{R}$
$p⟶\frac{1}{2}\left({\int }_{K}{d}_{t}^{2}\left(p,a\right)da+{\int }_{K}{d}_{t}^{2}\left(a,p\right)da\right)$
Proof: Because the above discussion, the existence of a minimum is reduced to the Riemannian case, which holds under the conditions of the theorem ([4]). $\square$  Similar ideas can be applied to the left and right components, giving ${m}_{l}\left(t\right)$  and ${m}_{r}\left(t\right)$  . Let us note that an interesting phenomenon happens. Consider the transformations ${\phi }_{t}$  producing the evolution ${m}_{r}⟶{m}_{r}\left(t\right)$  and ${m}_{l}⟶{m}_{l}\left(t\right)$  . Then ${p}_{1}$  is a fixed point and indeed an attractor for ${m}_{r}\left(t\right)$  and ${m}_{l}\left(t\right)$  . Indeed the whole set collapses to the point ${p}_{1}$  . Now let us consider a vector field defined through the curve. One could ask what is the natural evolution of fields for the transformation induced from ${\phi }_{t}$  . There is at least one solution,
Definition 3.14Consider the set ${\gamma }_{l}\left(t\right):{m}_{l}\left(t\right)\cup {m}_{r}\left(t\right)\subset \mathbf{M}$  and that the tangent vector at this point is $T$  . A tensor field $\mathcal{T}$  along the path ${\gamma }_{l}\left(t\right)$  evolves through the equations ${\mathcal{ℒ}}_{T}\mathcal{T}=0$  . A similar definition can be applied to the right path ${\gamma }_{r}\left(t\right)$  .
With this definition we avoid the problem of giving an arbitrary value at the point ${p}_{1}$  , that is a fixed point of the evolution: if we give any other evolution we must decide whether we take the value evolved from the right or from the left or we take the initial constant value from the begin. This is a contradiction if we consider arbitrary fields.

The notion of totally geodesic sub-manifold is convex invariant because all the notions involved on its definition are convex-invariant. Also the notion of convex set and convex envelope are convex invariant. In fact, it is easy to prove the following result from the Riemannian case:
Proposition 3.15Consider the Finsler structure $\left(\mathbf{M},F\right)$  and $K\subset \mathbf{M}$  . The convex envelope of $K$  is a convex invariant sub-set.
Proof: Let us suppose that $p$  and $q$  are points of the manifold $\mathbf{M}$  joined by a geodesic segment of the initial Finsler structure. Then due to proposition $3.2$  they are also joined by a Riemannian geodesic. The other direction of the statement comes from the maximal extension of the geodesics, where it is possible. $\square$  The last result of this section is a direct consequence of the Riemannian case. We said that a Finsler structure $\left(\mathbf{M},F\right)$  is a space form if the associated $\left(\mathbf{M},h\right)$  is a Riemannian space form. Then:
Theorem 3.16A Finsler structure $\left(\mathbf{M},F\right)$  with a totally geodesic sub-manifold of dimension $k$  passing through any point with any specified $k$  -dimensional plane as tangent space must be a space form.
This result is a generalization of a Riemannian theorem, also due to Cartan ([4]).

4 Generalization of theorems involving Curvature

We start with a result relating the $hv-$  curvature of the Chern connection with the $hv-$  curvature of the average Riemannian structure, which is just null. This will be useful to illustrate the type of calculations involving the product of several average operations.
Proposition 4.1Let $\left(\mathbf{M},F\right)$  be a Finsler structure.
Let ${\stackrel{~}{X}}_{1},{\overline{X}}_{2}$  be the horizontal and vertical lifts in ${\mathbf{T}}_{u}\mathbf{N}$  of the linear independent vectors ${X}_{1},{X}_{2}\in {\mathbf{T}}_{x}\mathbf{M}$  . Then for every section $Y\in \mathbf{T}\mathbf{M}$  ,
 $\begin{array}{c}{\stackrel{~}{P}}_{x}\left({X}_{1},{X}_{2}\right)Y=<{\pi }_{2}{P}_{u}\left({\stackrel{~}{X}}_{1}{|}_{u},{\overline{X}}_{2}{|}_{u}\right){\vartheta }_{v}{\pi }_{u}^{*}Y{>}_{u}=0.\end{array}$ (4.1)
Proof: It is just a calculation. In our notation, the intrinsic formula for the $hv$  -curvature is given by
${P}_{u}\left({\stackrel{~}{X}}_{1},{\overline{X}}_{2}\right)Z:={\nabla }_{{\overline{X}}_{2}}{\nabla }_{{\stackrel{~}{X}}_{1}}Z,$
for any section $Z\in {\pi }^{*}\mathbf{T}\mathbf{M}$  (to prove this relation it is only necessary to consider the above expression for $P$  in a convenient basis and use the characterization of Chern's connection. ). Using the definition of average,
$<{\pi }_{2}{P}_{u}\left({\stackrel{~}{X}}_{1}{|}_{u},{\overline{X}}_{2}{|}_{u}\right){\vartheta }_{v}{\pi }_{u}^{*}Y{>}_{u}=<{\pi }_{2}{\nabla }_{{\overline{X}}_{2}}{|}_{u}{\nabla }_{{\stackrel{~}{X}}_{1}}{|}_{v}{\vartheta }_{v}{\pi }_{u}^{*}Y{>}_{u}.$
Just note that ${\pi }_{2}{\nabla }_{{\stackrel{~}{X}}_{1}}{|}_{v}{\vartheta }_{v}{\pi }_{u}^{*}Y$  is a section of $\mathbf{T}\mathbf{M}$  , due to the equation $\left(2.4\right)$  ,
${\nabla }_{{\overline{X}}_{2}}{|}_{u}{\nabla }_{{\stackrel{~}{X}}_{1}}{|}_{u}{\vartheta }_{v}{\pi }_{u}^{*}Y={\nabla }_{{\overline{X}}_{2}}{|}_{u}{\pi }^{*}{|}_{v}{\pi }_{2}{|}_{v}{\nabla }_{{\stackrel{~}{X}}_{1}}{|}_{u}{\vartheta }_{v}{\pi }_{u}^{*}Y=0.$
$\square$  A similar result holds for the $vv$  -curvature $Q$  . Therefore,
Theorem 4.2Let $\left(\mathbf{M},F\right)$  be a Finsler structure with curvature endomorphism $\left\{{\Omega }_{j}^{i}\left(X,Y\right),X,Y\in \mathbf{T}\mathbf{N}\right\}$  . Then the set of average curvature endomorphism $\left\{{\stackrel{~}{\Omega }}_{j}^{i}\left(V,W\right),V,W\in \mathbf{T}\mathbf{M}\right\}$  is the set of Riemannian curvature endomorphism of $\left(\mathbf{M},h\right)$  .
Proof: It is a consequence of theorem $2.11$  and proposition $4.1$  ( $Q=0$  is also convex invariant). $\square$

We try to relate the flag curvature of the Finsler structure $\left(\mathbf{M},F\right)$  with the sectional curvature of the associated Riemannian structure $\left(\mathbf{M},h\right)$  . The sectional curvature associated with the plane $\Pi \subset {\mathbf{T}}_{x}\mathbf{M}$  is given by
$\stackrel{~}{K}\left(x,\Pi \right):={lim}_{A\to 0}\frac{{A}_{N}}{A},$
where $A$  is the area associated to the image of the exponential map measured with the Riemannian metric $h$  and ${A}_{N}$  is the area of the image under the Gauss map. This definition suggest the following flag-curvature:
Definition 4.3Let $\left(\mathbf{M},F\right)$  be a Finsler structure. For every $y\in {\mathbf{T}}_{x}\mathbf{M}$  it is defined a Gauss map associated with the direction $y$  and denote it by ${N}_{y}$  by taking the metric ${g}_{y}=g\left(\cdot ,y\right)$  . The flag curvature associated with the plane $\Pi \left(y,V\right)\subset {\mathbf{T}}_{x}\mathbf{M}$  with flag pole $y$  and edge $V$  is given by
$K\left(x,\Pi \right){lim}_{A\to 0}\frac{{A}_{N\left(y\right)}}{A\left(y\right)},$
where $A\left(y\right)$  is the area measured by the metric $g\left(\cdot ,y\right)$  in the image of the exponential map.
The next step is to show the relation between sectional curvature and flag curvature. In the general case it seems that our average is rather problematic in this case because its essential dependence of the flag-pole.
However, for the case of flag pole independence the average is well defined,
Proposition 4.4Let $\left(\mathbf{M},F\right)$  be a Finsler structure with flag curvature at $x$  given by $\left\{K\left(x,y\right)\right\}$  . Then the sectional curvature associated with $\left(\mathbf{M},F\right)$  is the averaged flag curvature,
$\stackrel{~}{K}\left(x,\Pi \right)=}_{y},$
where $\Pi$  is generated by the vectors $y$  and $V$  .
In addition, for a general Finsler space we define for each interpolating metrics ${g}_{t}$  the associated flag curvature ${K}^{t}\left(x,\Pi \right)$  with the plane $\Pi \subset {\mathbf{T}}_{x}\mathbf{M}$  . The following proposition follows,
Proposition 4.5Let $\left(\mathbf{M},F\right)$  be a Finsler structure. Then the following proposition holds,
 $\begin{array}{c}{lim}_{t\to 0}{K}^{t}\left(x,\Pi \right)=\stackrel{~}{K}\left(x,\Pi \right).\end{array}$ (4.2)
Proof: It follows from the commutativity of the double limit (for smooth structures the commutativity of the limits is allowed),
${lim}_{t\to 0}{lim}_{A\to 0}\frac{{A}_{N\left(y\right)}^{t}}{{A}^{t}\left(y\right)}={lim}_{A\to 0}{lim}_{t\to 0}\frac{{A}_{N\left(y\right)}^{t}}{{A}^{t}\left(y\right)}={lim}_{A\to 0}\frac{{A}_{N\left(y\right)}^{0}}{{A}^{0}\left(y\right)}.$
$\square$  From here it follows the convex invariance of the notion of sectional curvature at least for the case where there is not dependence on the flag-pole.
There is an alternative definition of flag curvature, which implies the selection of two orthonormal vectors as flag pole and edge. Denote them by ${e}_{1}$  and ${e}_{2}$  . Then,
Definition 4.6Let $\left(\mathbf{M},F\right)$  be a Finsler structure and $\overline{K}\left(x,\Pi \right)$  the flag curvature associated to the plane $\Pi \subset {\mathbf{T}}_{x}\mathbf{M}$  generated by the ${g}_{u}$  -orthonormal basis $\left({e}_{1},{e}_{2}\right)$  respect ${g}_{u}$  . Then on this basis we define,
 $\begin{array}{c}\overline{K}\left(u,\Pi \right):={R}_{1212}\left(u\right).\end{array}$ (4.3)
This is rather similar to the sectional curvature in Riemannian geometry. Although it seems that the definition is basis dependent, it should be stressed that, due to the tensor character of the curvature respect the flag pole, it does not really depends on that.
In relation with the average operation, one obtains the general result,
Proposition 4.7Let $\left(\mathbf{M},F\right)$  be a Finsler structure and $\overline{K}\left(u,\Pi \right)$  the flag curvature associated to the plane $\Pi \subset {\mathbf{T}}_{x}\mathbf{M}$  generated by $\left({e}_{1},{e}_{2}\right)$  . Then the corresponding sectional curvature associated to the Riemannian structure $\left(\mathbf{M},F\right)$  is given by the average
$\stackrel{~}{K}\left(u,\Pi \right):={\stackrel{~}{R}}_{1212}=<{R}_{1212}{>}_{u}.$
Proof: Nothing that ${R}_{1212}\left(u,{e}_{1}\right)={g}_{i1}\left(u\right){R}_{212}^{i}\left(u\right)$  , we average this equality:
$<{R}_{1212}\left(u\right){>}_{u}=<{R}_{1212}\left(u\right){>}_{u}=<{g}_{i1}\left(u\right){R}_{212}^{i}\left(u\right){>}_{u}=<{g}_{i1}\left(u\right){>}_{u}<{R}_{212}^{i}\left(u\right){>}_{u}={h}_{i1}\left(x\right){\stackrel{~}{R}}_{212}^{i}\left(x\right).$
$\square$  This alternative definition of flag curvature is always convex-invariant.

The first generalization of a Riemannian result involving curvature corresponds to Shur's lemma:
Theorem 4.8Let $\left(\mathbf{M},F\right)$  be a Finsler structure with $dim\mathbf{M}\ge 3$  such that its flag curvature $K$  does not depend on the flagpole nor on the transverse edge. Then the flag curvature is constant on $\mathbf{M}$  .
Proof: It is just a consequence of the Riemannian result, since $K$  only depends on $x\in \mathbf{M}$  , the average can be performed for both, the current and the new definitions of sectional curvature, giving the same result.
The non-dependence on $u$  after integration is used to apply the Riemannian result. $\square$
Proposition 4.9Let $\left(\mathbf{M},F\right)$  be a Finsler structure and consider the associated flag curvature $\overline{K}$  . The bound property of flag curvature is a convex-invariant property.
Proof: It is a direct consequence of the integration on a compact manifold because if $K\left(u,\Pi \right)<\epsilon$  , then $<\overline{K}\left(u,\Pi \right){>}_{u}<\epsilon .$  $\square$
Theorem 4.10( Cartan-Hadamard's theorem for Finsler spaces) Let $\left(\mathbf{M},F\right)$  be a forward geodesic complete Finsler structure of non-positive Flag curvature $\overline{K}\left(u,\Pi \right)$  . Then the exponential map $ex{p}_{x}:{\mathbf{T}}_{x}\mathbf{M}\mathbf{⟶}\mathbf{M}$  is a covering map. When $\mathbf{M}$  is simply connected, then it is diffeomorphic to ${\mathbf{R}}^{n}$  .
Proof: Due to the convex invariance of the exponential map, the application of the Riemannian theorem associated with the Riemannian structure $\left(\mathbf{M},h\right)$  implies the result. $\square$

We said that a Finsler manifold is strictly $\delta -$  pinching when $\delta \frac{1}{{R}^{2}}<\overline{K}\left(u,\Pi \right)<\frac{1}{{R}^{2}}$  for some positive $\delta$  and $R$  . We present now an analogous result to the Sphere Theorem of Riemannian geometry,
Theorem 4.11Let $\left(\mathbf{M},F\right)$  be a forward complete, simply-connected Finsler structure of dimension $n$  that is strictly $\frac{1}{4}$  -pinched. Then $\mathbf{M}$  is homeomorphic to the sphere ${\mathbf{S}}^{n}$  .
Proof: Again it is a direct consequence of the convex-invariance of the notions involved and the Riemannian result. $\square$

It is interesting to consider the following “Ricci” and scalar curvatures in the Finslerian context:
Definition 4.12Let $\left(\mathbf{M},F\right)$  be a Finsler structure with sectional curvature $\overline{K}\left(u,\Pi \right)$  . Consider a ${g}_{y}$  -orthonormal basis $\left\{V,{e}_{2},...,{e}_{n}\right\}$  of ${\mathbf{T}}_{x}\mathbf{M}$  . Then the Ricci curvature is defined by bi-linearity in $V$  and
 $\begin{array}{c}Ri{c}_{V}\left(u,V\right):={\sum }_{k=2}^{n}\overline{K}\left(u,\Pi \left(V,{e}_{k}\right)\right).\end{array}$ (4.4)
The scalar curvature is defined by
 $\begin{array}{c}S\left(u\right):={\sum }_{k\ne j}^{n}\overline{K}\left(u,\Pi \left({e}_{j},{e}_{k}\right)\right).\end{array}$ (4.5)
This definition of the Ricci curvature and scalar curvature are convex-invariant in the following sense:
Proposition 4.13Let $\left(\mathbf{M},F\right)$  be a Finsler structure with sectional curvature $\overline{K}.\left(u,\Pi \right)$  . Then the following relations hold:
• 1. The Ricci curvature $Ri{c}_{y}\left(u,y\right)$  is convex invariant,  $\begin{array}{c}=\stackrel{~}{R}ic\left(x,y\right).\end{array}$ (4.6)
• 2. The scalar curvature $S\left(u\right)$  is convex invariant:  $\begin{array}{c}=\stackrel{~}{S}\left(x\right).\end{array}$ (4.7)
• 3. The following relation holds between the $hh-$  curvature and $\overline{K}$  curvature:
${g}_{u}\left(R\left(X,Y\right)Z,W\right)=1/6\left\{\overline{K}\left(u,X+W,Y+Z\right)|\left(X+W\right)\wedge \left(Y+Z\right){|}^{2}-\overline{K}\left(u,Y+W,X+Z\right)|\left(Y+W\right)\wedge \left(X+Z\right){|}^{2}-$
$-\overline{K}\left(u,X,Y+Z\right)|X\wedge \left(Y+W\right){|}^{2}-\overline{K}\left(u,Y,Z+W\right)|Y\wedge \left(X+W\right){|}^{2}-\overline{K}\left(u,Z,X+W\right)|Z\wedge \left(X+W\right){|}^{2}-$
$-\overline{K}\left(u,w,Y+Z\right)|W\wedge \left(Y+Z\right){|}^{2}-\overline{K}\left(u,X,Y+W\right)|X\wedge \left(Y+W\right){|}^{2}+\overline{K}\left(u,Y,Z+W\right)|Y\wedge \left(Z+W\right){|}^{2}-$
$-\overline{K}\left(u,Z,Y+W\right)|Z\wedge \left(Z+W\right){|}^{2}+\overline{K}\left(u,W,X+Z\right)|W\wedge \left(X+Z\right){|}^{2}+\overline{K}\left(u,X,Z\right)|X\wedge Y{|}^{2}+$
 $\begin{array}{c}+\overline{K}\left(u,Y,W\right)|Y\wedge W{|}^{2}-\overline{K}\left(u,X,Y\right)|Z\wedge W{|}^{2}-\overline{K}\left(u,Y,Z\right)|Y\wedge Z{|}^{2}\right\}.\end{array}$ (4.8)
From this relation it follows the convex invariance of the $hh-$  curvature.
Proof: We perform the following calculation:
$=<{\sum }_{k=2}^{n}\overline{K}\left(u,\Pi \left(V,{e}_{k}\right)\right)>={\sum }_{k=2}^{n}\overline{K}\left(x,\Pi \left(V,{e}_{k}\right)\right)==\stackrel{~}{R}ic\left(x,y\right).$
A similar calculation could be performed for the scalar curvature. The prove of the equation $\left(4.8\right)$  it is just an expansion of the equation $\left(4.3\right)$  for arbitrary vectors. Convex invariance is proved taking the average.
The proof for the third part consist on computing both sides in a convenient basis. $\square$

This result points directly to the following result, a “diffeomorphism” pinching theorem in Finsler geometry,
Theorem 4.14Let $\left(\mathbf{M},F\right)$  be a forward complete, simply-connected Finsler structure of dimension $3$  with strictly positive scalar curvature. Then M is diffeomorphic to the sphere ${\mathbf{S}}^{3}$  .
Proof: It follows as a consequence of the analogous Riemannian theorem and from the convex invariance of the Ricci curvature defined above. $\square$

From this results, these new definitions of flag curvature and Ricci curvature are arguably interesting because they do not require more work to be computed as in the Riemannian case, they reduce to the corresponding Riemannian notions and they are convex invariant. As consequence, most results can be translate from the Riemannian to Finsler the category. Some of these results should be compared with the current ones found in [2]. For instance, Shur's lemma here is the same as in [2] while, for example, the Cartan-Hadamard theorem presented here is different. Our result is less strong. Nevertheless our results seems good enough to us because the content of our result is equivalent to the content of the Riemannian one. Similar remarks are applicable to the sphere theorems.

The last results of this section concerns a generalization of the Gauss-Bonnet-Chern theorem ([5]).
Consider the Paffian form defined by
$\Omega :=\frac{\left(-1{\right)}^{p-1}}{{2}^{2p}{\pi }^{p}p!}{\epsilon }^{{i}_{1}\cdot \cdot \cdot {i}_{n}}{\Omega }_{{i}_{1}{i}_{2}}\wedge \cdot \cdot \cdot \wedge {\Omega }_{{i}_{n-1}{i}_{n}},\text{when}n=2p;$
$\Omega :=0,\text{when}n=2q+1.$
Proposition 4.15The Paffian form $\Omega$  is a convex invariant form.
Proof: In order to proof this fact we must perform the following calculation:
$<\Omega >=<\frac{\left(-1{\right)}^{p-1}}{{2}^{2p}{\pi }^{p}p!}{\epsilon }^{{i}_{1}\cdot \cdot \cdot {i}_{n}}{\Omega }_{{i}_{1}{i}_{2}}\wedge \cdot \cdot \cdot \wedge {\Omega }_{{i}_{n-1}{i}_{n}}>\text{when}n=2p;$
$<\Omega >=0,\text{when}n=2q+1.$
For odd dimension it is obvious the result and it implies the convex invariance. For even dimension, the following equality is implied by the properties of the average:
$<\frac{\left(-1{\right)}^{p-1}}{{2}^{2p}{\pi }^{p}p!}{\epsilon }^{{i}_{1}\cdot \cdot \cdot {i}_{n}}{\Omega }_{{i}_{1}{i}_{2}}\wedge \cdot \cdot \cdot \wedge {\Omega }_{{i}_{n-1}{i}_{n}}>=\frac{\left(-1{\right)}^{p-1}}{{2}^{2p}{\pi }^{p}p!}{\epsilon }^{{i}_{1}\cdot \cdot \cdot {i}_{n}}<{\Omega }_{{i}_{1}{i}_{2}}>\wedge \cdot \cdot \cdot \wedge <{\Omega }_{{i}_{n-1}{i}_{n}}>.$
Again, taking into account the definition of the $2$  -form ${\Omega }_{ij}$  ,
$<{\Omega }_{ij}>=<{g}_{ik}{\Omega }_{j}^{k}>=<{g}_{ij}><{\Omega }_{j}^{k}>={h}_{ik}{\stackrel{~}{\Omega }}_{j}^{k}.$
After this, the following equation holds
 $\begin{array}{c}<\Omega >=\stackrel{~}{\Omega }.\end{array}$ (4.9)
This relation implies the convex invariance of the Paffian. $\square$  The Gauss-Bonnet-Chern theorem involves in an intrinsic way the notion of volume of the indicatrix, defined by $Vol\left(x\right)={\int }_{{\mathbf{I}}_{x}}dvo{l}_{g}$  , where $dvo{l}_{g}$  is the volume form
$dvo{l}_{g}=\sqrt{g}{\sum }_{j=1}^{n}\left(-1{\right)}^{j-1}{y}^{j}d{y}^{1}\wedge \cdot \cdot \cdot \wedge d{y}^{j-1}\wedge d{y}^{j}\wedge ...\wedge d{y}^{n}$
and the coordinates ${y}^{j}$  are required to satisfy $F\left(x,y\right)=1$  . Let us denote by $\chi \left(\mathbf{M}\right)$  the Euler characteristic of M. Then,
Theorem 4.16(Gauss-Bonnet-Chern's theorem) Let $\mathbf{M}$  be a closed, oriented Finsler manifold.
Then the following equality holds:
 $\begin{array}{c}{\int }_{\mathbf{S}\mathbf{M}}\frac{1}{{F}^{n}Vol\left(x\right)}\left(-\Omega \right)=\chi \left(\mathbf{M}\right).\end{array}$ (4.10)
Proof: Let us apply the Gauss-Bonnet-Chern theorem for the Riemannian structure $\left(\mathbf{M},h\right)$  :
${\int }_{\mathbf{M}}-\stackrel{~}{\Omega }=\chi \left(\mathbf{M}\right).$
Therefore, developing the left side and since $\Omega$  is invariant under scaling on $y$  ,
$\chi \left(\mathbf{M}\right)={\int }_{M}-\stackrel{~}{\Omega }={\int }_{\mathbf{M}}\frac{1}{Vol\left(x\right)}{\int }_{{\mathbf{I}}_{x}}\left(-\Omega \right)={\int }_{\mathbf{M}}\frac{1}{Vol\left(x\right)}{\int }_{{\mathbf{S}}_{x}\mathbf{M}}\frac{dvol}{{F}^{n}}\left(-\Omega \right)={\int }_{\mathbf{S}\mathbf{M}}\frac{1}{{F}^{n}Vol\left(x\right)}\left(-\Omega \right).$
$\square$  As special case we get the following formula when the Paffian $\Omega$  lives in $\mathbf{M}$  ,
Corollary 4.17Let $\mathbf{M}$  be a closed, oriented Finsler manifold such that the Paffian $\Omega$  lives in $\mathbf{M}$  . Then the following formula holds:
${\int }_{\mathbf{M}}-\Omega =\chi \left(\mathbf{M}\right).$
Proof: Considering the equality $Vol\left(x\right)={\int }_{{\mathbf{S}}_{x}\mathbf{M}}\frac{dvol}{{F}^{n}\left(x,y\right)}$  ([6]), we get
$\chi \left(\mathbf{M}\right)={\int }_{\mathbf{S}\mathbf{M}}\frac{1}{{F}^{n}Vol\left(x\right)}\left(-\Omega \right)={\int }_{\mathbf{M}}\left(-\Omega \right){\int }_{{\mathbf{S}}_{x}\mathbf{M}}\frac{1}{{F}^{n}\left(x,y\right)Vol\left(x\right)}={\int }_{\mathbf{M}}-\Omega .$
$\square$  Let us remark the similarity of our result and the one found in [7] on Berwald spaces. There it was based on the construction of a Riemannian structure using the linear connection coming from the initial Berwald structure ([8]). The Gauss-Bonnet-Chern formula is then
 $\begin{array}{c}{\int }_{\mathbf{M}}-{\Omega }_{h}=\chi \left(\mathbf{M}\right).\end{array}$ (4.11)
The Riemannian metric $h$  is the Szabó construction. Apparently our corollary $4.17$  is a generalization of this one; a priori the requirement that $\Omega$  lives in $\mathbf{M}$  is more general than the structure being Berwald. We must also compare our result with Dazord ([9]). There, a general theorem for Landsberg surfaces where the integration is performed in the Sphere bundle similarly to equation $\left(4.9\right)$  . However we differ in the treatment of the volumes and on the connection used and in our conclusion, because our 4.16 is completely general.

5 Conclusion and Remarks

Theorem $3.8$  is the main tool in this work to generalize Riemannian theorems to the Finsler category presented in this work. That this can be done for many interesting results using the same procedure is rather remarkable, hoping the existence of a deep reason for that.
For this reason it seems interesting to generalize also Riemannian notions to Finsler geometry in such a way that:
• 1. They reduce to the Riemannian case when $\left(\mathbf{M},F\right)$  it is just a Riemannian geometry.
• 2. They produce after the average the Riemannian notion.
• 3. They are computable with complexity comparable to the Riemannian case.
The first and second conditions are essential while the third condition is useful, although not so fundamental for geometry. However, adopting a computational point of view, it could be argued the real importance of this condition: although not a fundamental property, we should adopt it in order to obtain sensible results.
Unfortunately we do not know presently how to perform the generalization of arbitrary Riemannian notions to the Finsler category, so we can not really state what is the exact difference between global Finsler and global Riemannian geometry. Nevertheless, all the results stated in this work implies that they are rather similar. We also noted the existence of some differences, but the amount of results presented here and the common way they are proved, claims for the existence of a common ground, between differential geometry and topology, where both kinds of geometries should be considered as equivalent.
References

1. R. Gallego, A Riemannian structure associated to a Finsler structure, math.DG/0501058.
2. D.Bao, S.S.Chern and Z.Shen, An Introduction to Riemann-Finsler Geometry, Graduate Texts in Mathematics 200, Springer-Verlag.
3. B.O'Neill, Semi-Riemannian Geometry, with applications to Relativity, Academic Press, 1983.
4. M.Berger, A Panoramic View of Riemannian Geometry, Springer-Verlag, 2003.
5. S.S.Chern, On the Curvature Integra in a Riemannian Manifold, Ann.Math.46 (1945), 674-684.
6. D.Bao, Z.Shen, On the volume of unit spheres in a Finsler Manifold, Results in Math. 26(1994), 1-17.
7. D.Bao, S.S.Chern and Z.Shen, On the Gauss-Bonnet Integrand for 4-Dimensional Landsberg Spaces, Cont. Math. 196, 15-25.
8. Z.I.Szabó, Positive definite Berwald Spaces (Structure theorems on Berwald Spaces), Tensor, N.S. 35, 25-39.
9. P.Dazord,Tores Finslériens sans points conjugués , Bull.Soc.Math.France 99 (1971), 171-192.
10. L.Skorniakov, Metrization of the projective plane in connection with a given system of curves Iyv.Akad.Nauk SSSR Sr.Mat. 19 (1955), 471-482.