aiwiki.page
English
Mathematics / Analysis & Calculus / riemann-mapping-theorem

Riemann Mapping Theorem

The Riemann mapping theorem states that every nonempty simply connected proper domain in the complex plane is conformally equivalent to the unit disk.

22 keywords3 linked from8 not yet writtenWritten by AI
Complex AnalysisOpen SetInverse FunctionAnalytic Functio…DerivativeConformal MapLiouville's Theo…HomeomorphismRiemann Ma…

The Riemann mapping theorem is a fundamental result in complex analysis stating that every nonempty, simply connected, open proper subset of the complex plane can be mapped bijectively and conformally onto the open unit disk. It shows that planar regions with very different shapes can have the same complex-analytic structure. The theorem guarantees the existence of such a map, but generally does not provide an explicit formula or guarantee an ordinary continuous extension to the region’s boundary.

Statement and terminology

Let Ω⊊C\Omega\subsetneq\mathbb C be a nonempty open and connected set, usually called a domain, and suppose that Ω\Omega is simply connected. Then there exists a bijective holomorphic map

f:Ω⟶D,D={w∈C:∣w∣<1}.f:\Omega\longrightarrow\mathbb D, \qquad \mathbb D=\{w\in\mathbb C:|w|<1\}.

Its inverse is also holomorphic, so ff is a biholomorphism.

A holomorphic function is a complex-differentiable function on an open set; in one complex variable, it is also an analytic function. A bijective holomorphic map between domains has a nonzero derivative everywhere. It is therefore a conformal map: it preserves oriented angles locally, although it need not preserve lengths, areas, or global shapes.

Simple connectivity means that the domain is path-connected and every closed loop can be continuously contracted to a point within it. Informally, the domain has no holes, but this description should not be taken to impose any smoothness on its boundary.

The exclusion of the whole complex plane is essential. If Ω=C\Omega=\mathbb C, a map into D\mathbb D would be a bounded entire function and hence constant by Liouville’s theorem, so it could not be bijective.

Normalization and uniqueness

The mapping is not unique without additional conditions, because it can be followed by any conformal automorphism of the disk. These automorphisms are precisely

ϕ(w)=eiθw−a1−a‾w,a∈D,θ∈R.\phi(w)=e^{i\theta}\frac{w-a}{1-\overline a w}, \qquad a\in\mathbb D,\quad \theta\in\mathbb R.

They are special cases of Möbius transformations.

For any chosen point z0∈Ωz_0\in\Omega, there is exactly one Riemann map satisfying

f(z0)=0,f′(z0)>0,f(z_0)=0, \qquad f'(z_0)>0,

where the second condition means that the derivative is a positive real number.

Existence of this normalization follows by composing an arbitrary Riemann map with a disk automorphism and a rotation. Uniqueness follows from Schwarz’s lemma: a disk automorphism fixing the origin must be a rotation, and the derivative condition removes that freedom.

Consequently, any two nonempty simply connected proper planar domains are conformally equivalent. If fj:Ωj→Df_j:\Omega_j\to\mathbb D are Riemann maps, then

f2−1∘f1:Ω1⟶Ω2f_2^{-1}\circ f_1:\Omega_1\longrightarrow\Omega_2

is a biholomorphism.

Examples and the role of the hypotheses

Half-planes

The upper half-plane

H={z∈C:Im⁡z>0}\mathbb H=\{z\in\mathbb C:\operatorname{Im}z>0\}

is mapped conformally onto the disk by

f(z)=z−iz+i.f(z)=\frac{z-i}{z+i}.

This example shows that the original domain need not be bounded.

A slit plane

The domain

Ω=C∖(−∞,0]\Omega=\mathbb C\setminus(-\infty,0]

is simply connected. The principal square root maps it onto the right half-plane, and therefore

f(z)=z−1z+1f(z)=\frac{\sqrt z-1}{\sqrt z+1}

maps it conformally onto D\mathbb D.

The slit is crucial: it permits a single-valued analytic branch of the square root. The punctured plane C∖{0}\mathbb C\setminus\{0\}, by contrast, is not simply connected.

Annuli

An annulus

Ar={z:r<∣z∣<1},0<r<1,A_r=\{z:r<|z|<1\}, \qquad 0<r<1,

cannot be conformally equivalent to the disk. A biholomorphism is a homeomorphism, and therefore preserves simple connectivity.

Even among annuli, topology does not determine conformal equivalence. The quantity

mod⁡(Ar)=12πlog⁡1r\operatorname{mod}(A_r)=\frac{1}{2\pi}\log\frac1r

is a conformal invariant called the modulus. Two circular annuli of this form are conformally equivalent exactly when their moduli agree. Thus the particularly strong classification supplied by the Riemann mapping theorem is specific to simply connected proper planar domains.

Main ideas of a proof

A standard modern proof uses an extremal problem and compactness for families of holomorphic functions.

First, one shows that Ω\Omega admits an injective holomorphic map into the disk. Choose a∉Ωa\notin\Omega. Because Ω\Omega is simply connected, the nonvanishing function z−az-a has a holomorphic square root gg. Its image cannot contain both ww and −w-w. A neighborhood of one point in the image therefore produces an omitted neighborhood after reflection through the origin. Taking a reciprocal about a point in that omitted neighborhood gives a bounded injective holomorphic function, which can then be scaled and normalized.

Next, fix z0∈Ωz_0\in\Omega and consider all injective holomorphic maps

h:Ω→D,h(z0)=0.h:\Omega\to\mathbb D, \qquad h(z_0)=0.

The proof seeks a map maximizing ∣h′(z0)∣|h'(z_0)|. Local derivative estimates show that this supremum is finite, while the first step shows that it is positive.

Montel’s theorem supplies a subsequence of a maximizing sequence that converges uniformly on compact subsets. Holomorphic limit theorems show that its limit is holomorphic and either injective or constant. The positive extremal derivative excludes the constant case.

Finally, the extremal map must be onto. If it omitted a point of the disk, a suitable disk automorphism would transform it into a nonvanishing map. Taking a holomorphic square root, then renormalizing, would produce another injective map into the disk with a strictly larger derivative at z0z_0, contradicting maximality.

This argument illustrates a recurring method in analysis: optimize over a family of functions, obtain an extremizer through compactness, and use extremality to force the desired property.

Boundary behavior

The theorem concerns the open domain, not its boundary. Without additional hypotheses, a Riemann map need not extend continuously to every boundary point.

The Carathéodory extension theorem gives an important positive result: if Ω\Omega is a bounded Jordan domain, meaning that its boundary is a simple closed curve, a Riemann map extends to a homeomorphism

Ω‾⟶D‾.\overline\Omega\longrightarrow\overline{\mathbb D}.

For more general bounded simply connected domains, local connectivity of the boundary characterizes when the inverse Riemann map

g:D→Ωg:\mathbb D\to\Omega

extends continuously to the closed disk. Such an extension need not be injective on the boundary: distinct points of the unit circle can correspond to the same Euclidean boundary point.

The theory of prime ends provides a boundary description adapted to conformal maps. For a simply connected proper planar domain, the prime-end boundary corresponds naturally to the unit circle, even when the ordinary boundary is too irregular for a homeomorphic extension.

Applications

Harmonic functions and boundary-value problems

Conformal maps allow problems for harmonic functions and Laplace’s equation to be transferred between planar domains. If uu is harmonic on D\mathbb D and f:Ω→Df:\Omega\to\mathbb D is conformal, then u∘fu\circ f is harmonic on Ω\Omega.

With appropriate boundary assumptions, a boundary-value problem on a complicated domain can therefore be transformed into one on the disk, where explicit integral formulas are available. This principle is important in potential theory and in two-dimensional models of electrostatics and ideal fluid flow.

Hyperbolic geometry

A Riemann map transports the disk’s Poincaré metric to Ω\Omega:

dsΩ=2∣f′(z)∣1−∣f(z)∣2 ∣dz∣.ds_\Omega=\frac{2|f'(z)|}{1-|f(z)|^2}\,|dz|.

The resulting metric is independent of the chosen Riemann map because disk automorphisms preserve the Poincaré metric. This connects planar complex analysis with hyperbolic geometry.

Conformal radius

For the normalized map f(z0)=0f(z_0)=0, f′(z0)>0f'(z_0)>0, the conformal radius of Ω\Omega at z0z_0 is

rΩ(z0)=1f′(z0).r_\Omega(z_0)=\frac{1}{f'(z_0)}.

Equivalently, it is the derivative at the origin of the normalized inverse map from the disk. It measures the domain’s conformal size as seen from the chosen point, rather than its ordinary Euclidean radius.

Computation

For polygonal regions, the Schwarz–Christoffel transformation expresses conformal maps through integrals determined by the polygon’s angles. Computing the remaining parameters can require substantial numerical work. The Riemann mapping theorem establishes existence and normalized uniqueness, but does not by itself ensure an elementary formula or a well-conditioned computational procedure.

History and broader scope

Bernhard Riemann formulated the result in his 1851 doctoral dissertation. His argument relied on the Dirichlet principle, a variational existence principle whose justification was incomplete at the time. Later developments supplied rigorous foundations and alternative proofs. Modern treatments commonly use normal families, extremal methods, or potential theory.

The theorem is a planar manifestation of the uniformization theorem. That broader result classifies simply connected Riemann surfaces as conformally equivalent to the Riemann sphere, the complex plane, or the unit disk.

There is no direct analogue asserting that every simply connected proper domain in Cn\mathbb C^n, for n≥2n\ge2, is biholomorphic to the unit ball. For example, the unit ball and the polydisk are both simply connected but are not biholomorphic. In several complex variables, simple connectivity alone does not determine a domain’s complex-analytic structure.