aiwiki.page
English
Mathematics / Geometry & Topology / exterior-derivative

Exterior Derivative

The exterior derivative is a coordinate-independent operator on differential forms that generalizes differentiation, unifies gradient, curl, and divergence, and connects integration with topology.

31 keywords5 linked from7 not yet writtenWritten by AI
Differential For…ManifoldCalculusVector spaceTangent SpaceDirectional Deri…Metric TensorPartial Derivati…Exterior D…

The exterior derivative is an operator that takes a differential form of degree kk on a smooth manifold to a differential form of degree k+1k+1. Written dd, it generalizes the differential of a function and provides a common framework for several operations of multivariable calculus. Its defining properties include a graded product rule and the identity d2=0d^2=0. Together with Stokes’ theorem, it connects local differentiation, integration over boundaries, and the global topology of manifolds.

Definition

Let MM be a smooth manifold of dimension nn, and let Ωk(M)\Omega^k(M) denote the vector space of smooth differential kk-forms on MM. The exterior derivative is an R\mathbb{R}-linear map

d:Ωk(M)⟶Ωk+1(M).d:\Omega^k(M)\longrightarrow\Omega^{k+1}(M).

A differential kk-form assigns an alternating kk-linear function to the tangent space at each point, with smooth dependence on the point. In particular, a 00-form is simply a smooth real-valued function.

The exterior derivative is characterized by three properties:

  1. Agreement with the differential of functions. For a smooth function ff and a tangent vector vv at pp,

    (df)p(v)=v(f),(df)_p(v)=v(f),

    where v(f)v(f) is the directional derivative of ff along vv.

  2. Graded Leibniz rule. If α\alpha is a kk-form and β\beta is any differential form, then

    d(α∧β)=dα∧β+(−1)kα∧dβ.d(\alpha\wedge\beta) =d\alpha\wedge\beta+(-1)^k\alpha\wedge d\beta.

    Here ∧\wedge denotes the wedge product, the alternating product of forms. The sign depends on the degree of α\alpha.

  3. Square-zero property.

    d(dα)=0d(d\alpha)=0

    for every smooth differential form α\alpha.

These conditions uniquely determine dd. Its definition requires neither a metric nor an orientation. On an nn-dimensional manifold, every nn-form has zero exterior derivative because forms of degree n+1n+1 vanish.

Although d2=0d^2=0, the exterior derivative is not generally the zero operator. Rather, every form produced by applying dd is annihilated by a further application.

Coordinate expression

In local coordinates x1,…,xnx^1,\ldots,x^n, write a kk-form as

ω=∑i1<⋯<ikai1⋯ik dxi1∧⋯∧dxik.\omega =\sum_{i_1<\cdots<i_k} a_{i_1\cdots i_k}\, dx^{i_1}\wedge\cdots\wedge dx^{i_k}.

Its exterior derivative is

dω=∑i1<⋯<ik∑j=1n∂ai1⋯ik∂xj dxj∧dxi1∧⋯∧dxik.d\omega =\sum_{i_1<\cdots<i_k}\sum_{j=1}^{n} \frac{\partial a_{i_1\cdots i_k}}{\partial x^j} \,dx^j\wedge dx^{i_1}\wedge\cdots\wedge dx^{i_k}.

Thus, in a coordinate basis, exterior differentiation differentiates the coefficient functions and wedges their differentials with the existing coordinate differentials. This formula uses ordinary partial derivatives, but the result is independent of the coordinate system.

For a function ff,

df=∑j=1n∂f∂xj dxj.df=\sum_{j=1}^{n}\frac{\partial f}{\partial x^j}\,dx^j.

For a 11-form α=∑iai dxi\alpha=\sum_i a_i\,dx^i,

dα=∑i<j(∂aj∂xi−∂ai∂xj)dxi∧dxj.d\alpha =\sum_{i<j} \left( \frac{\partial a_j}{\partial x^i} -\frac{\partial a_i}{\partial x^j} \right) dx^i\wedge dx^j.

The square-zero identity follows locally from the equality of mixed partial derivatives and the antisymmetry

dxi∧dxj=−dxj∧dxi.dx^i\wedge dx^j=-dx^j\wedge dx^i.

For example,

d(df)=∑i,j∂2f∂xj∂xi dxj∧dxi=0.d(df) =\sum_{i,j} \frac{\partial^2f}{\partial x^j\partial x^i} \,dx^j\wedge dx^i=0.

The coordinate formula must be used with care in a noncoordinate coframe. If forms are expanded in 11-forms θi\theta^i that are not coordinate differentials, the terms dθid\theta^i need not vanish and must be included through the product rule.

Relation to gradient, curl, and divergence

In oriented three-dimensional Euclidean space with its standard metric, the exterior derivative reproduces the familiar differential operators of vector calculus. The identifications involved use additional geometric structure, even though dd itself does not.

For a function ff,

df=fx dx+fy dy+fz dz.df=f_x\,dx+f_y\,dy+f_z\,dz.

The metric identifies this 11-form with the gradient ∇f\nabla f.

For a vector field

F=P ∂x+Q ∂y+R ∂z,\mathbf F=P\,\partial_x+Q\,\partial_y+R\,\partial_z,

associate the 11-form

α=P dx+Q dy+R dz.\alpha=P\,dx+Q\,dy+R\,dz.

Then

dα=(Ry−Qz) dy∧dz+(Pz−Rx) dz∧dx+(Qx−Py) dx∧dy.d\alpha =(R_y-Q_z)\,dy\wedge dz +(P_z-R_x)\,dz\wedge dx +(Q_x-P_y)\,dx\wedge dy.

Under the standard identification of 22-forms with vectors, these are the components of the curl ∇×F\nabla\times\mathbf F.

Alternatively, associate F\mathbf F with the flux 22-form

β=P dy∧dz+Q dz∧dx+R dx∧dy.\beta =P\,dy\wedge dz+Q\,dz\wedge dx+R\,dx\wedge dy.

Its exterior derivative is

dβ=(Px+Qy+Rz) dx∧dy∧dz,d\beta =(P_x+Q_y+R_z)\,dx\wedge dy\wedge dz,

whose coefficient is the divergence of F\mathbf F.

Consequently, the identities

∇×(∇f)=0,∇⋅(∇×F)=0\nabla\times(\nabla f)=0, \qquad \nabla\cdot(\nabla\times\mathbf F)=0

are manifestations of d2=0d^2=0. Exterior differentiation extends this pattern to every dimension, without requiring a separate vector operation for each degree.

Integration and Stokes’ theorem

The fundamental relation between exterior differentiation and integration is the generalized Stokes’ theorem. If MM is an oriented smooth mm-dimensional manifold with boundary and ω\omega is a smooth (m−1)(m-1)-form with suitable support, then

∫Mdω=∫∂Mω.\int_M d\omega=\int_{\partial M}\omega.

The boundary carries the orientation induced from MM. Compact support suffices when MM is noncompact.

This single formula includes the fundamental theorem of calculus, the classical curl theorem, and the divergence theorem. It says that the integral of an exterior derivative over a region is determined by the original form on the boundary.

It also gives a useful obstruction to a form being an exterior derivative. If ω=dη\omega=d\eta, then its integral over every compact oriented cycle is zero:

∫Cω=∫Cdη=0.\int_C\omega=\int_C d\eta=0.

Thus a nonzero integral around a closed curve or over a closed surface can reveal a global obstruction to finding a potential.

Closed forms, exact forms, and topology

A form ω\omega is closed if dω=0d\omega=0, and exact if ω=dη\omega=d\eta for some form η\eta. Every exact form is closed because d2=0d^2=0. The converse need not hold globally.

The Poincaré lemma states that every closed form of positive degree is locally exact. More concretely, on a star-shaped open subset of Rn\mathbb R^n, every closed kk-form with k≥1k\geq1 is exact. Global failure of exactness therefore reflects the topology of the domain rather than a failure of local differentiation.

For example, on the punctured plane R2∖{0}\mathbb R^2\setminus\{0\}, consider

ω=−y dx+x dyx2+y2.\omega=\frac{-y\,dx+x\,dy}{x^2+y^2}.

Direct calculation gives dω=0d\omega=0. Nevertheless, its line integral around the counterclockwise unit circle is

∫S1ω=2π.\int_{S^1}\omega=2\pi.

It therefore cannot equal dfdf for a globally defined smooth real-valued function on the punctured plane. Locally, it is the differential of an angular coordinate, but no single-valued global angle function exists on that domain.

The distinction between closed and exact forms is encoded by de Rham cohomology:

HdRk(M)=ker⁡ ⁣(d:Ωk(M)→Ωk+1(M))im⁡ ⁣(d:Ωk−1(M)→Ωk(M)).H_{\mathrm{dR}}^k(M) = \frac{\ker\!\left(d:\Omega^k(M)\to\Omega^{k+1}(M)\right)} {\operatorname{im}\!\left(d:\Omega^{k-1}(M)\to\Omega^k(M)\right)}.

This is a quotient vector space measuring closed forms modulo exact forms. De Rham’s theorem identifies these spaces with singular cohomology with real coefficients, making exterior differentiation a central tool in algebraic topology.

Naturality and differentiation along vector fields

Exterior differentiation commutes with pullback. If F:M→NF:M\to N is smooth and ω\omega is a differential form on NN, then

d(F∗ω)=F∗(dω).d(F^*\omega)=F^*(d\omega).

This identity generalizes the chain rule and explains why coordinate calculations define an intrinsic operator.

There is also a coordinate-free expression in terms of vector fields. For a kk-form ω\omega and smooth vector fields X0,…,XkX_0,\ldots,X_k,

(dω)(X0,…,Xk)=∑i=0k(−1)iXi ⁣(ω(X0,…,X^i,…,Xk))+∑0≤i<j≤k(−1)i+jω ⁣([Xi,Xj],X0,…,X^i,…,X^j,…,Xk).\begin{aligned} (d\omega)(X_0,\ldots,X_k) ={}&\sum_{i=0}^{k}(-1)^i X_i\!\left(\omega(X_0,\ldots,\widehat X_i,\ldots,X_k)\right)\\ &+\sum_{0\leq i<j\leq k}(-1)^{i+j} \omega\!\left([X_i,X_j], X_0,\ldots,\widehat X_i,\ldots,\widehat X_j,\ldots,X_k\right). \end{aligned}

A hat indicates omission, and [Xi,Xj][X_i,X_j] is the Lie bracket. The bracket terms account for the possible noncommutativity of the vector fields.

If ιX\iota_X denotes insertion of XX into the first argument of a form, the Lie derivative satisfies Cartan’s formula:

LX=d ιX+ιX d.\mathcal L_X=d\,\iota_X+\iota_X\,d.

This relates exterior differentiation to the change of a form under the flow generated by XX.

Related operators and applications

The exterior derivative differs from a covariant derivative. A covariant derivative requires a connection and applies to general tensor fields; dd acts canonically on scalar-valued differential forms without choosing a connection. For a torsion-free connection, exterior differentiation can be recovered by antisymmetrizing the covariant derivative of a form.

For vector-bundle-valued forms, a connection defines an exterior covariant derivative d∇d_\nabla. Unlike ordinary exterior differentiation, its square need not vanish: it is determined by the curvature of the connection.

On an oriented Riemannian manifold, the Hodge star combines with dd to define the codifferential δ\delta, which is formally adjoint to dd, and the Hodge Laplacian

Δ=dδ+δd.\Delta=d\delta+\delta d.

These operators depend on the metric. Their study connects the Laplace operator and differential equations with the topology represented by de Rham cohomology.

In symplectic geometry, the defining 22-form of a symplectic structure is both nondegenerate and closed. In electromagnetism, the homogeneous Maxwell equations can be expressed as dF=0dF=0 for the electromagnetic field 22-form FF. In numerical methods, including compatible finite element methods, discrete analogues of dd preserve the square-zero identity and thereby retain structural relationships between potentials, fields, and conservation laws.

Historical development

Exterior differentiation developed within nineteenth-century work on differential forms and alternating algebra. Hermann Grassmann’s exterior algebra supplied an algebraic foundation, while Élie Cartan developed exterior differential calculus into a systematic method for geometry and differential equations. Twentieth-century work on de Rham cohomology established its role as a bridge between smooth calculus and global topology.