This article needs additional citations for verification. Please help improve this article by adding citations to reliable sources. Unsourced material may be challenged and removed.Find sources: "Finsler manifold" – news · newspapers · books · scholar · JSTOR (May 2017) (Learn how and when to remove this template message)

In mathematics, particularly differential geometry, a **Finsler manifold** is a differentiable manifold *M* where a (possibly asymmetric) **Minkowski functional** *F*(*x*, −) is provided on each tangent space T_{x}*M*, that enables one to define the length of any smooth curve *γ* : [*a*, *b*] → *M* as

Finsler manifolds are more general than Riemannian manifolds since the tangent norms need not be induced by inner products.

Every Finsler manifold becomes an intrinsic quasimetric space when the distance between two points is defined as the infimum length of the curves that join them.

Élie Cartan (1933) named Finsler manifolds after Paul Finsler, who studied this geometry in his dissertation (Finsler 1918).

A **Finsler manifold** is a differentiable manifold *M* together with a **Finsler metric**, which is a continuous nonnegative function *F*: T*M* → [0, +∞) defined on the tangent bundle so that for each point *x* of *M*,

*F*(*v*+*w*) ≤*F*(*v*) +*F*(*w*) for every two vectors*v*,*w*tangent to*M*at*x*(subadditivity).*F*(λ*v*) = λ*F*(*v*) for all λ ≥ 0 (but not necessarily for λ < 0) (positive homogeneity).*F*(*v*) > 0 unless*v*= 0 (positive definiteness).

In other words, *F*(*x*, −) is an asymmetric norm on each tangent space T_{x}*M*. The Finsler metric *F* is also required to be **smooth**, more precisely:

*F*is smooth on the complement of the zero section of T*M*.

The subadditivity axiom may then be replaced by the following **strong convexity condition**:

- For each tangent vector
*v*≠ 0, the Hessian matrix of*F*^{2}at*v*is positive definite.

Here the Hessian of *F*^{2} at *v* is the symmetric bilinear form

also known as the **fundamental tensor** of *F* at *v*. Strong convexity of implies the subadditivity with a strict inequality if *u*⁄*F*(*u*) ≠ *v*⁄*F*(*v*). If *F* is strongly convex, then it is a **Minkowski norm** on each tangent space.

A Finsler metric is **reversible** if, in addition,

*F*(−*v*) =*F*(*v*) for all tangent vectors*v*.

A reversible Finsler metric defines a norm (in the usual sense) on each tangent space.

- Smooth submanifolds (including open subsets) of a normed vector space of finite dimension are Finsler manifolds if the norm of the vector space is smooth outside the origin.
- Riemannian manifolds (but not pseudo-Riemannian manifolds) are special cases of Finsler manifolds.

Let be a Riemannian manifold and *b* a differential one-form on *M* with

where is the inverse matrix of and the Einstein notation is used. Then

defines a **Randers metric** on *M* and is a **Randers manifold**, a special case of a non-reversible Finsler manifold.^{[1]}

Let (*M*, *d*) be a quasimetric so that *M* is also a differentiable manifold and *d* is compatible with the differential structure of *M* in the following sense:

- Around any point
*z*on*M*there exists a smooth chart (*U*, φ) of*M*and a constant*C*≥ 1 such that for every*x*,*y*∈*U* - The function
*d*:*M*×*M*→ [0, ∞] is smooth in some punctured neighborhood of the diagonal.

Then one can define a Finsler function *F*: *TM* →[0, ∞] by

where *γ* is any curve in *M* with *γ*(0) = *x* and *γ'*(0) = v. The Finsler function *F* obtained in this way restricts to an asymmetric (typically non-Minkowski) norm on each tangent space of *M*. The induced intrinsic metric *d*_{L}: *M* × *M* → [0, ∞] of the original quasimetric can be recovered from

and in fact any Finsler function *F*: T*M* → [0, ∞) defines an intrinsic quasimetric *d*_{L} on *M* by this formula.

Due to the homogeneity of *F* the length

of a differentiable curve *γ*: [*a*, *b*] → *M* in *M* is invariant under positively oriented reparametrizations. A constant speed curve *γ* is a geodesic of a Finsler manifold if its short enough segments *γ*|_{[c,d]} are length-minimizing in *M* from *γ*(*c*) to *γ*(*d*). Equivalently, *γ* is a geodesic if it is stationary for the energy functional

in the sense that its functional derivative vanishes among differentiable curves *γ*: [*a*, *b*] → *M* with fixed endpoints *γ*(*a*) = *x* and *γ*(*b*) = *y*.

The Euler–Lagrange equation for the energy functional *E*[*γ*] reads in the local coordinates (*x*^{1}, ..., *x*^{n}, *v*^{1}, ..., *v*^{n}) of T*M* as

where *k* = 1, ..., *n* and *g*_{ij} is the coordinate representation of the fundamental tensor, defined as

Assuming the strong convexity of *F*^{2}(*x*, *v*) with respect to *v* ∈ T_{x}*M*, the matrix *g*_{ij}(*x*, *v*) is invertible and its inverse is denoted by *g*^{ij}(*x*, *v*). Then *γ*: [*a*, *b*] → *M* is a geodesic of (*M*, *F*) if and only if its tangent curve *γ'*: [*a*, *b*] → T*M*∖{0} is an integral curve of the smooth vector field *H* on T*M*∖{0} locally defined by

where the local spray coefficients *G*^{i} are given by

The vector field *H* on T*M*∖{0} satisfies *JH* = *V* and [*V*, *H*] = *H*, where *J* and *V* are the canonical endomorphism and the canonical vector field on T*M*∖{0}. Hence, by definition, *H* is a spray on *M*. The spray *H* defines a nonlinear connection on the fibre bundle T*M*∖{0} → *M* through the vertical projection

In analogy with the Riemannian case, there is a version

of the Jacobi equation for a general spray structure (*M*, *H*) in terms of the Ehresmann curvature and nonlinear covariant derivative.

By Hopf–Rinow theorem there always exist length minimizing curves (at least in small enough neighborhoods) on (*M*, *F*). Length minimizing curves can always be positively reparametrized to be geodesics, and any geodesic must satisfy the Euler–Lagrange equation for *E*[*γ*]. Assuming the strong convexity of *F*^{2} there exists a unique maximal geodesic *γ* with *γ*(0) = x and *γ'*(0) = v for any (*x*, *v*) ∈ T*M*∖{0} by the uniqueness of integral curves.

If *F*^{2} is strongly convex, geodesics *γ*: [0, *b*] → *M* are length-minimizing among nearby curves until the first point *γ*(*s*) conjugate to *γ*(0) along *γ*, and for *t* > *s* there always exist shorter curves from *γ*(0) to *γ*(*t*) near *γ*, as in the Riemannian case.