Exterior derivative

From Wikipedia, the free encyclopedia

In mathematics, the exterior derivative operator of differential geometry extends the concept of the differential of a function to differential forms of higher degree. It is important in the theory of integration on manifolds, and is the differential (coboundary) used to define de Rham and Alexander-Spanier cohomology. Its current form was invented by Élie Cartan.

1 Definition
2 Properties
3 Invariant formula
4 The exterior derivative in calculus
5 Examples
6 See also

The exterior derivative of a differential form of degree k is a differential form of degree k + 1.

For a k-form ω = f_I dx_I over Rⁿ, the definition is as follows:

$d{\omega} = \sum_{i=1}^n \frac{\partial f_I}{\partial x_i} dx_i \wedge dx_I.$

For general k-forms Σ_I f_I dx_I (where the multi-index I runs over all ordered subsets of {1, ..., n} of cardinality k), we just extend linearly. Note that if $i = I$ above then $dx_i \wedge dx_I = 0$ (see wedge product).

Exterior differentiation satisfies three important properties:

linearity

the wedge product rule (see antiderivation)

$d(\omega \wedge \eta) = d\omega \wedge \eta+(-1)^{{\rm deg\,}\omega}(\omega \wedge d\eta)$

and d² = 0, a formula encoding the equality of mixed partial derivatives, so that always

$d^2(\omega)=d(d\omega)=0 \, \!$

It can be shown that the exterior derivative is uniquely determined by these properties and its agreement with the differential on 0-forms (functions).

The kernel of d consists of the closed forms, and the image of the exact forms (cf. exact differentials).

The exterior derivative is natural. If f: M → N is a smooth map and Ω^k and Ω^k+1 are the contravariant smooth functors that assign correspondingly to each manifold the space of k- and k+1-forms on the manifold, then the following diagram commutes

so d(f*ω) = f*dω, where f* denotes the pullback of f. Thus d is a natural transformation from Ω^k to Ω^k+1.

Given a k-form ω and arbitrary smooth vector fields V₀,V₁, …, V_k we have

$d\omega(V_0,V_1,...V_k) = \sum_i(-1)^i V_i\left(\omega(V_0, \ldots, \hat V_i, \ldots,V_k)\right)$

$+\sum_{i<j}(-1)^{i+j}\omega([V_i, V_j], V_0, \ldots, \hat V_i, \ldots, \hat V_j, \ldots, V_k)$

where $[V i, V j]$ denotes Lie bracket and the hat denotes the omission of that element: $\omega(V_0, \ldots, \hat V_i, \ldots,V_k) = \omega(V_0, \ldots, V_{i-1}, V_{i+1}, \ldots, V_k).$

In particular, for 1-forms we have:

d ω(X, Y) = X (ω(Y)) - Y (ω(X)) - ω([X, Y]).

The following correspondence reveals about a dozen formulas from vector calculus as merely special cases of the above three rules of exterior differentiation.

For a 0-form, that is a smooth function f: Rⁿ→R, we have

$df = \sum_{i=1}^n \frac{\partial f}{\partial x_i}\, dx_i.$

Therefore, for vector field $V$

$df(V) = \langle \mbox{grad }f,V\rangle,$

where grad f denotes gradient of f and < , > is the scalar product.

For a 1-form $\omega=\sum_{i} f_i\,dx_i$ ,

$d \omega=\sum_{i,j}\frac{\partial f_i}{\partial x_j} dx_j\wedge dx_i,$

$d\omega=\sum_{i<j}\left(\frac{\partial f_j}{\partial x_i}-\frac{\partial f_i}{\partial x_j}\right)dx_i\wedge dx_j$

which restricted to the three-dimensional case $\omega= u\,dx+v\,dy+w\,dz$ is

$d \omega = \left(\frac{\partial v}{\partial x} - \frac{\partial u}{\partial y} \right) dx \wedge dy + \left(\frac{\partial w}{\partial y} - \frac{\partial v}{\partial z} \right) dy \wedge dz + \left(\frac{\partial u}{\partial z} - \frac{\partial w}{\partial x} \right) dz \wedge dx.$

Therefore, for vector fields $U$ , $V = [u, v, w]$ and $W$ we have $d \omega(U,W)=\langle\mbox{curl}\, V \times U,W\rangle$ where curl V denotes the curl of V, × is the vector product, and < , > is the scalar product.

For a 2-form $\omega = \sum_{i,j} h_{i,j}\,dx_i\wedge dx_j,$

$d \omega = \sum_{i,j,k} \frac{\partial h_{i,j}}{\partial x_k} dx_k \wedge dx_i \wedge dx_j.$

For three dimensions, with $\omega = p\,dy\wedge dz+q\,dz\wedge dx+r\,dx\wedge dy$ we get

$d \omega\,$	$= \left( \frac{\partial p}{\partial x} + \frac{\partial q}{\partial y} + \frac{\partial r}{\partial z} \right) dx \wedge dy \wedge dz$
	$= \mbox{div}V\, dx \wedge dy \wedge dz,$