Skip to content

Differential Forms, Integration, and Orientation

Scrollable formula. Use Left and Right arrow keys to reveal hidden content.

Scrollable table. Use Left and Right arrow keys to reveal additional columns.

Scrollable figure. Use Left and Right arrow keys to inspect hidden content.

Differential forms unite local calculus, oriented integration, and global topology in one metric-free language. Alternation turns Jacobians into signed geometric factors; the exterior derivative expresses differential identities without a connection; Stokes’ theorem converts bulk derivatives into boundary integrals; and de Rham cohomology records when local potentials cannot be joined globally. This chapter develops that sequence carefully, with the punctured-plane angular form and the oriented disk serving as recurring tests of local versus global reasoning and of every consequential sign.

Required background. Tensor Fields and Natural Operations — use covariant tensors, pullbacks, and weighted antisymmetrization.

Helpful background. Calculus and Linear Algebra for Relativity — review determinants, multivariable integration, and the fundamental theorem of calculus.

Alternation as the language of integration

Section titled “Alternation as the language of integration”

A differential pp-form is a smooth alternating covariant pp-tensor. It evaluates an ordered pp-tuple of tangent vectors and changes sign when two are exchanged. This behavior is exactly what an oriented infinitesimal pp-dimensional element needs. The exterior algebra and forms page builds wedge products, interior products, pullbacks, and degree counting from that observation. It keeps the normalization explicit:

ω=1p!ωi1⋯ipdxi1∧⋯∧dxip.\omega=\frac{1}{p!}\omega_{i_1\cdots i_p} dx^{i_1}\wedge\cdots\wedge dx^{i_p}.

Because a top form transforms by the signed Jacobian determinant, it is suited to integration on an oriented manifold. This conclusion needs no metric. The Hodge star, norms of forms, and metric volume are later structures and should not be imported into the present argument.

Differentiation, local potentials, and global failure

Section titled “Differentiation, local potentials, and global failure”

The exterior derivative and Poincaré lemma page characterizes the operator dd by its action on functions, its graded Leibniz rule, and d2=0d^2=0. A form is closed when dω=0d\omega=0 and exact when ω=dη\omega=d\eta. Exact forms are always closed. On a star-shaped domain, the homotopy operator proves the local converse for positive degree.

The global converse fails. On R2∖{0}\mathbb R^2\setminus\{0\},

η=−y dx+x dyx2+y2\eta=\frac{-y\,dx+x\,dy}{x^2+y^2}

is closed and locally equals dθd\theta, but its integral around a positively oriented circle is 2π2\pi. No globally single-valued real angle can be its potential. This example prevents the common overstatement that the Poincaré lemma applies on every domain. The homotopy proof and exact domain hypotheses are developed in Guillemin and Haine 2019, §2.6, Theorem 2.6.15 and Exercises 2.6.vi–vii, pp. 64–67, PDF.

The parametrized integration page defines an integral by pulling a kk-form back along a parametrization σ:D→M\sigma:D\to M. Orientation-preserving reparametrizations leave the integral invariant; orientation-reversing ones change its sign. This is the signed counterpart of the ordinary change-of-variables theorem. The page also makes multiplicity and rank failure explicit, so a many-to-one map is not silently treated as an ordinary coordinate change.

The generalized Stokes theorem page fixes the boundary sign with the outward-normal-first convention and proves the architecture behind

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

For the disk oriented by dx∧dydx\wedge dy and ω=x dy\omega=x\,dy, both sides equal π\pi. The induced boundary runs counterclockwise; reversing that traversal exposes the sign error immediately. Compact support, compactness, or separately checked convergence is essential, because an unrecorded boundary at infinity can spoil the formula.

Orientation is a coherent choice of sign for tangent bases. A top form uses that sign; a density transforms with the absolute value of the Jacobian. The orientation, densities, and volume elements page tests the distinction on the Möbius strip. Its seam reverses orientation, so a nowhere-vanishing two-form cannot descend, while the positive density ∣ds dt∣\lvert ds\,dt\rvert does. Nonorientability therefore obstructs signed top-form integration, not positive size.

The de Rham cohomology and global flux page then packages closed forms modulo exact forms:

HdRp(M)=Zp(M)/Bp(M).H^p_{\mathrm{dR}}(M)=Z^p(M)/B^p(M).

Periods over closed cycles depend only on the cohomology class. For S1S^1, integration around the circle identifies HdR1(S1)H^1_{\mathrm{dR}}(S^1) with R\mathbb R. A nonzero period is global information that no local potential detects. This result does not by itself define a physical charge: a theory must select the form, normalization, cycles, sources, and boundary conditions.

For the full foundation, read the six pages in order. Exterior algebra supplies the objects; dd supplies the differential complex; pullback defines parametrized integration; Stokes connects bulk and boundary; orientation determines the sign; cohomology identifies global obstructions.

For a field-theory route, read exterior algebra, the exterior derivative, Stokes, and de Rham cohomology. For boundary calculations, read parametrized integration, Stokes, and orientation together—the three pages share one sign convention and should not be separated. For topology, start with the Poincaré lemma’s local result, then compare it with the angular form and the period map on S1S^1.

Metric-dependent Hodge theory and general algebraic topology require further study. This chapter establishes the differential-form tools needed later for metric volume, Cartan equations, curvature forms, variational boundary terms, and characteristic classes.

Orient the unit disk DD by dx∧dydx\wedge dy and set ω=x dy\omega=x\,dy. Compute both sides of Stokes’ theorem and explain why the induced orientation of ∂D\partial D is counterclockwise.

Solution

Since dω=dx∧dyd\omega=dx\wedge dy,

∫Ddω=area⁡(D)=π.\int_Dd\omega=\operatorname{area}(D)=\pi.

Parametrize the boundary by γ(θ)=(cos⁡θ,sin⁡θ)\gamma(\theta)=(\cos\theta,\sin\theta) for 0≤θ≤2π0\le\theta\le2\pi. Then

γ∗ω=cos⁡2θ dθ,∫∂Dω=∫02πcos⁡2θ dθ=π.\gamma^*\omega=\cos^2\theta\,d\theta, \qquad \int_{\partial D}\omega =\int_0^{2\pi}\cos^2\theta\,d\theta=\pi.

At (1,0)(1,0) the outward normal is ∂x\partial_x and the counterclockwise tangent is ∂y\partial_y; the ordered pair (∂x,∂y)(\partial_x,\partial_y) has the disk orientation. Rotating this argument around the circle proves that counterclockwise is the outward-normal-first boundary orientation. A clockwise parametrization would give −π-\pi and would represent the opposite orientation.

You should now be able to calculate graded signs, distinguish local exactness from global exactness, integrate a form by pullback, diagnose an orientation reversal, apply Stokes with its support and boundary hypotheses visible, distinguish a density from a top form, and interpret a period as cohomological data rather than automatically as a physical observable. The next chapter adds a Lorentzian metric and makes clear which new constructions—causal cones, a metric volume, and the Hodge dual—depend on it.