Riemann Surfaces — Introduction

A Riemann surface is a connected one-dimensional complex manifold — a topological surface equipped with an atlas whose transition maps are holomorphic. Key result: every Riemann surface carries a unique complex-analytic structure compatible with locally defined holomorphic functions, and the global theory bridges analysis, topology, and algebraic geometry. Applies in complex analysis, algebraic curves, string theory, and number theory. Introduced by Bernhard Riemann in his 1851 dissertation.

A Riemann surface $X$ is a connected Hausdorff topological space together with a collection of charts $φ_{α} : U_{α} \to C$ whose transition functions $φ_{β} \circ φ_{α}^{- 1}$ are biholomorphic on overlaps. This makes precise the idea of a "multi-valued function" like $z$ or $lo g z$ becoming single-valued on an appropriate covering space. Riemann surfaces matter because they unify three apparently distinct worlds: the analysis of holomorphic functions, the topology of orientable surfaces, and the algebra of polynomial equations in two variables. The surprising fact is the deep trichotomy: every simply connected Riemann surface is biholomorphic to exactly one of three models — the Riemann sphere, the complex plane, or the unit disc — a uniformization that secretly classifies all Riemann surfaces by their universal cover.

The Core Insight

A Riemann surface is precisely the structure needed to make "holomorphic" a local notion that glues globally. Because biholomorphic transition maps are angle-preserving and orientation-preserving, every Riemann surface is automatically a smooth orientable real 2-manifold equipped with a conformal structure. Conversely, in dimension two the data of a conformal structure (a way to measure angles) is equivalent to the data of a complex structure. This is why Riemann surfaces sit at the crossroads: complex analysis, conformal geometry, and orientable topology are three faces of one object.

Formal Statement

Holomorphic Atlas and Riemann Surface

Let $X$ be a connected Hausdorff topological space. A complex chart is a homeomorphism $φ_{α} : U_{α} \to V_{α} \subseteq C$ from an open set $U_{α} \subseteq X$ onto an open set $V_{α} \subseteq C$ . Two charts $(U_{α}, φ_{α})$ , $(U_{β}, φ_{β})$ are holomorphically compatible if either $U_{α} \cap U_{β} = \emptyset$ or the transition map

φ_{β} \circ φ_{α}^{- 1} : φ_{α} (U_{α} \cap U_{β}) ⟶ φ_{β} (U_{α} \cap U_{β})

is biholomorphic (holomorphic with holomorphic inverse). A complex atlas $A = {(U_{α}, φ_{α})}$ is a collection of pairwise compatible charts covering $X$ . A Riemann surface is the pair $(X, Σ)$ where $Σ$ is a maximal complex atlas (a complex structure).

⟳The transition map is well-defined on the image

φ_{α} (U_{α} \cap U_{β}) \subseteq C

; requiring it to be holomorphic (Cauchy–Riemann) imposes the complex structure independently of chart choice.

Holomorphic and Meromorphic Functions

A function $f : X \to C$ is holomorphic at $p \in X$ if for some (hence every) chart $φ_{α}$ around $p$ , the composite $f \circ φ_{α}^{- 1}$ is holomorphic at $φ_{α} (p)$ in the usual sense. A meromorphic function is a holomorphic map $f : X \to P^{1} = C \cup {\infty}$ not identically $\infty$ . The field of meromorphic functions is denoted $M (X)$ .

Topology Forces Orientability

Every Riemann surface is a smooth, orientable, real two-dimensional manifold. Moreover, the conformal structure induced by the charts is intrinsic.

We argue in numbered steps that holomorphicity yields orientability.

Step 1. Write a transition map

T = φ_{β} \circ φ_{α}^{- 1}

in real coordinates

T (x, y) = (u (x, y), v (x, y))

with

z = x + i y

. Holomorphicity gives the Cauchy–Riemann equations

u_{x} = v_{y}, u_{y} = - v_{x}

. Step 2. The Jacobian determinant of

T

as a real map is

J_{T} = u_{x} v_{y} - u_{y} v_{x} = u_{x}^{2} + v_{x}^{2} = ∣ T^{'} (z) ∣^{2} \geq 0.

⟳substitute the CR relations

v_{y} = u_{x}

v_{x} = - u_{y}

into the

2 \times 2

real Jacobian and recognize

∣ T^{'} (z) ∣^{2} = u_{x}^{2} + u_{y}^{2}

Step 3. Since

T

is biholomorphic,

T^{'} (z) \neq = 0

, so

J_{T} > 0

everywhere. All transition maps are orientation-preserving, so the local orientations from

C

glue into a global orientation. Step 4. Holomorphic maps with nonzero derivative are conformal (preserve oriented angles):

T^{'} (z)

acts as multiplication by a complex number

r e^{i θ}

, a rotation–scaling. Thus angle measurement transfers consistently between charts, defining an intrinsic conformal structure.

■

Bernhard Riemann | 1826–1866

Riemann introduced these surfaces in his 1851 Göttingen dissertation Grundlagen für eine allgemeine Theorie der Functionen einer veränderlichen complexen Grösse, supervised informally in the spirit of Gauss. His radical idea: rather than wrestle with multi-valued functions like $z$ over the plane, build a new domain — a layered surface — on which the function becomes single-valued and well-behaved. He connected the topology (genus, connectivity) of these surfaces to the analytic behavior of integrals of algebraic functions (Abelian integrals), culminating in the Riemann–Roch theorem with his student Roch. Riemann's surfaces were initially viewed as a heuristic device; only with Weyl's 1913 Die Idee der Riemannschen Fläche did they receive the modern axiomatic manifold definition.

Building Intuition

Think of $w = z$ . Over each nonzero $z$ there are two values $\pm z$ ; following a loop around $z = 0$ swaps them. To make $w$ single-valued, take two copies of the plane (sheets), slit each along the negative real axis, and cross-glue them so that walking around the origin moves you from sheet 1 to sheet 2 and back. The resulting surface — when the branch point and infinity are added — is topologically a sphere, and on it $w$ is a genuine single-valued holomorphic function. A Riemann surface is exactly this idea made rigorous: the "natural geometric home" where a multi-valued analytic function lives without ambiguity. The genus (number of handles) measures how topologically complicated this home is.

Two-sheeted Riemann surface of $w=\sqrt z$: cross-gluing along a branch cut

Worked Examples

example[The Riemann Sphere $P^{1}$ ] Given: $X = C \cup {\infty}$ , the one-point compactification, with two charts.

Steps.

Chart 1: $U_{0} = C$ , $φ_{0} (z) = z$ . Chart 2: $U_{\infty} = (C ∖ {0}) \cup {\infty}$ , $φ_{\infty} (z) = 1/ z$ (with $φ_{\infty} (\infty) = 0$ ).
On the overlap $U_{0} \cap U_{\infty} = C ∖ {0}$ , the transition map is

φ_{\infty} \circ φ_{0}^{- 1} (z) = \frac{1}{z},

⟳

φ_{0}^{- 1} (z) = z

, then apply

φ_{\infty} (z) = 1/ z

; this is holomorphic and nonvanishing on

C ∖ {0}

The map $z \mapsto 1/ z$ is biholomorphic on $C ∖ {0}$ , so the two charts are compatible. $X$ is compact, connected, Hausdorff.
Compute the genus: stereographic projection shows $X$ is homeomorphic to $S^{2}$ .

Answer:

P^{1} is a compact Riemann surface of genus g = 0, χ = 2.

Its meromorphic function field is

M (P^{1}) = C (z)

, the rational functions.

A Complex Torus $\mathbb{C}/\Lambda$

Given: Lattice

Λ = Z \cdot 1 + Z \cdot τ

with

τ = i

(so

Λ = Z + Z i

); form

X = C /Λ

. Steps.

Charts: for each $p \in X$ pick a small disc in $C$ mapping injectively under the quotient $π : C \to X$ ; let $φ = (π ∣_{disc})^{- 1}$ .
Two such inverse charts differ by a translation $z \mapsto z + λ$ , $λ \in Λ$ :

φ_{β} \circ φ_{α}^{- 1} (z) = z + λ, λ \in Λ,

⟳overlapping local inverses of

π

are sections that differ by a deck transformation of the covering

C \to C /Λ

, which are exactly translations by lattice vectors.

Translations are biholomorphic, so the atlas is holomorphic. $X$ is compact (continuous image of the compact fundamental square $[0, 1]^{2}$ ).
Topologically $R^{2} / Z^{2} ≅ S^{1} \times S^{1}$ , a genus-1 surface: $χ = 0$ .

Answer:

C /Λ is a compact Riemann surface of genus g = 1.

Its meromorphic functions are generated by the Weierstrass

℘

-function and

℘^{'}

, satisfying

(℘^{'})^{2} = 4 ℘^{3} - g_{2} ℘ - g_{3}

. For

τ = i

one has

g_{3} = 0

and the torus is biholomorphic to the curve

y^{2} = 4 x^{3} - g_{2} x

Common Misconceptions

A Riemann surface is one-complex-dimensional, not one-real-dimensional. The word "surface" refers to real dimension 2; the complex dimension is 1. Beginners conflate the two because both contain a "1." Always remember:

dim_{C} = 1 ⟺ dim_{R} = 2

. Compactness is not automatic.

C

, the unit disc

D

, and any open subset of

C

are perfectly good (non-compact) Riemann surfaces. Many global theorems (Riemann–Roch, finiteness of meromorphic functions) hold only in the compact case. Non-constant holomorphic functions need not exist globally. On a compact Riemann surface the only globally holomorphic functions are constants (by the maximum modulus principle); the interesting global objects are meromorphic functions and differentials.

One might guess that any orientable surface admits a unique complex structure. False. The torus $C /Λ$ admits a one-complex-parameter family of inequivalent complex structures: tori $C / (Z + Z τ)$ and $C / (Z + Z τ^{'})$ are biholomorphic iff $τ^{'} = \frac{a τ + b}{c τ + d}$ for some $(a c b d) \in S L_{2} (Z)$ . Thus the moduli space of genus-1 Riemann surfaces is $H / S L_{2} (Z)$ , which is one-dimensional and far from a point. Topology underdetermines complex structure: the moduli space $M_{g}$ of genus- $g$ surfaces has complex dimension $3 g - 3$ for $g \geq 2$ .

Connections

Riemann surfaces are the meeting point of several theories. Via the equivalence between compact Riemann surfaces and smooth projective algebraic curves over $C$ , they connect to Algebraic Curves and Algebraic Geometry — every compact Riemann surface is the analytification of a unique smooth projective curve (Riemann's existence theorem / Chow's theorem). The Uniformization Theorem links them to Hyperbolic Geometry: surfaces of genus $\geq 2$ carry a metric of constant negative curvature. The study of holomorphic 1-forms ties into Hodge Theory and Abelian Varieties through the Jacobian. The covering-space and monodromy aspects connect to Fundamental Group and Galois Theory via the dictionary between branched covers and field extensions of $C (z)$ . Finally, Sheaf Cohomology organizes the global existence questions (Riemann–Roch, Serre duality).

Real-World Applications

String theory (worldsheet path integral): A propagating closed string sweeps out a Riemann surface; the genus- $g$ surface gives the $g$ -loop contribution to scattering amplitudes, integrated over moduli space $M_{g}$ (Polyakov 1981).
Conformal field theory: Correlation functions are defined on Riemann surfaces; modular invariance on the torus constrains the spectrum of 2D CFTs.
Integrable systems and solitons: Finite-gap solutions of the KdV and KP equations are expressed via theta functions of hyperelliptic Riemann surfaces (Novikov–Krichever theory).
Number theory: Modular curves $X_{0} (N) = H^{*} / Γ_{0} (N)$ are compact Riemann surfaces; their geometry underlies modular forms and the modularity theorem used in Fermat's Last Theorem.
Fluid dynamics and electrostatics: Two-dimensional potential flow and conductor problems use conformal maps between Riemann surface domains; the uniformization principle yields explicit solutions for multiply connected regions.
Computer graphics and surface parameterization: Discrete conformal mapping (based on the conformal structure of Riemann surfaces) provides angle-preserving texture mapping and mesh flattening.
Coding and complex analysis software: Algorithms computing periods of Abelian integrals and Riemann theta functions (used in geophysics, water-wave modeling) rest directly on the period-matrix theory of compact Riemann surfaces.

Historical Development

1851 — Bernhard Riemann's dissertation introduces Riemann surfaces as multi-sheeted coverings to resolve multi-valuedness; defines genus via connectivity.
1857 — Riemann's Theorie der Abel'schen Functionen develops Abelian integrals, the Riemann inequality, and the role of genus; Gustav Roch (1865) completes it to the Riemann–Roch theorem.
1882 — Felix Klein and Henri Poincaré, in competition, establish early forms of the uniformization theorem using automorphic and Fuchsian functions.
1900 — Hilbert's 22nd problem asks for the uniformization of analytic relations via automorphic functions, spurring rigorous proofs.
1907 — Poincaré and Paul Koebe independently prove the general Uniformization Theorem: every simply connected Riemann surface is $P^{1}$ , $C$ , or $D$ .
1913 — Hermann Weyl's Die Idee der Riemannschen Fläche gives the first axiomatic, manifold-theoretic definition, founding modern surface topology.
1931–1933 — Teichmüller begins the theory of quasiconformal deformations, later (1939–43) producing Teichmüller space and the modern theory of moduli.
1950s — Sheaf-theoretic and cohomological reformulations by Cartan, Serre, and Grauert (e.g., Serre's GAGA, 1956) recast Riemann surfaces within algebraic geometry.
1960s onward — Ahlfors, Bers, and others develop quasiconformal mappings and Kleinian groups, while the algebraic-geometric viewpoint matures into the theory of $M_{g}$ (Deligne–Mumford, 1969).

Subtleties & Limits

While the classification of compact Riemann surfaces by genus is complete, the fine geometry of moduli spaces remains deep. The Kobayashi/Weil–Petersson geometry of $M_{g}$ , the precise structure of its cohomology ring, and the asymptotics of Weil–Petersson volumes (Mirzakhani's recursion) connect to open questions in 2D quantum gravity and intersection theory on $\overline{M}_{g}$ . In the non-compact and infinite-genus settings, the type problem — deciding whether a given simply connected surface is parabolic ( $C$ ) or hyperbolic ( $D$ ) from local data — has no general algorithmic solution. For arithmetic curves, effective bounds in the theory of heights and explicit computation of periods to high precision remain active. The relationship between the analytic moduli space and its arithmetic/Galois-theoretic counterparts (Grothendieck's dessins d'enfants and the action of $Gal (\overline{Q} / Q)$ on Belyi covers) is only partially understood.