Introduction to Riemann Surfaces

Introduction to Riemann Surfaces Riemann surfaces represent one of the most elegant frameworks in mathematics, where complex analysis meets topology and geometry. The central idea is elegant: take functions that seem to have multiple values—like the square root or logarithm—and create a surface on which they become single-valued. This idea transforms complicated multi-valued functions into well-behaved holomorphic maps, and it reveals deep connections between analysis, topology, and algebraic geometry. What Is a Riemann Surface? A Riemann surface is a two-dimensional real manifold equipped with a compatible system of complex coordinates. In simpler terms: it's a surface that locally looks exactly like the complex plane $\mathbb{C}$. Building Intuition Imagine zooming in on a Riemann surface at any point. No matter which point you choose, the neighborhood around it looks like a small open region of $\mathbb{C}$. You could place a grid of complex coordinates on it, just as you would on the regular complex plane. The key insight is that by allowing the "shape" of this surface to twist and fold in three-dimensional space, we can make problematic multi-valued functions into honest, single-valued holomorphic functions. The Problem: Multi-Valued Functions To understand why Riemann surfaces are necessary, consider the function $w = \sqrt{z}$. This seems innocent enough, but there's a hidden complication. Why Square Root Is Troublesome In the complex plane, every nonzero complex number $z$ has two square roots. For example, if $z = 1$, then both $w = 1$ and $w = -1$ satisfy $w^2 = z$. The real issue emerges when we try to trace continuously how the square root of $z$ changes as $z$ moves around the origin in the complex plane. Start at $z = 1$ with square root $w = 1$. As you move $z$ counterclockwise around a complete circle back to $z = 1$, the square root continuously traces to $w = -1$. Trace another full circle, and you return to $w = 1$. This means we cannot define $\sqrt{z}$ as a single continuous function on all of $\mathbb{C}$ (excluding 0). We get different answers depending on how many times we've wound around the origin. The Branch Cut Solution (and Its Limitation) The standard approach in calculus is to introduce a branch cut—a ray from the origin, say along the negative real axis—and declare that we only consider values of $z$ that avoid this ray. On the domain $\mathbb{C}\setminus\{x\in\mathbb{R}\mid x\leq 0\}$, the square root becomes single-valued. But this is unsatisfying: we've artificially restricted the domain, and the function behaves pathologically at the branch cut. How Riemann Surfaces Solve This A Riemann surface solves the problem by "unfolding" the complex plane. Instead of a single copy of $\mathbb{C}$, imagine two copies glued together along a cut in a way that respects the two-fold nature of the square root. The formula $w = \sqrt{z}$ then defines a genuine single-valued holomorphic function on this two-sheeted surface. Local Charts and Complex Structure To make Riemann surfaces rigorous, we need to formalize how complex coordinates work on them. Charts: Coordinate Systems on the Surface A chart on a Riemann surface is a homeomorphism $$\phi\alpha \colon U\alpha \to V\alpha \subset \mathbb{C}$$ where $U\alpha$ is an open set on the surface and $V\alpha$ is an open set in the complex plane. Intuitively, $\phi\alpha$ is a "coordinate system" that labels points in $U\alpha$ using complex numbers. Because Riemann surfaces are locally like $\mathbb{C}$, we can always find such charts. The surface is covered by overlapping charts, just as an atlas covers the Earth with overlapping maps. Transition Maps and Holomorphic Compatibility When two charts $U\alpha$ and $U\beta$ overlap, there's a natural way to relate their coordinate systems. The transition map is $$\phi\beta \circ \phi\alpha^{-1} \colon \phi\alpha(U\alpha \cap U\beta) \to \phi\beta(U\alpha \cap U\beta)$$ This map translates from "$\phi\alpha$ coordinates" to "$\phi\beta$ coordinates" on the overlap region. The defining property of a Riemann surface is that every transition map must be a holomorphic function. This is the compatibility condition that makes everything work. Why Holomorphic Transition Maps Matter Holomorphic transition maps ensure that complex analysis is well-defined on the surface. If a function $f$ is holomorphic when expressed in $\phi\alpha$ coordinates, it remains holomorphic when expressed in $\phi\beta$ coordinates (because holomorphic means "remains holomorphic under holomorphic change of variables"). The notion of complex differentiability is intrinsic and doesn't depend on which chart you use. This collection of charts with holomorphic transition maps constitutes the complex structure of the Riemann surface—it's what allows us to do complex analysis there. Fundamental Examples Let's explore the most important Riemann surfaces. The Riemann Sphere The simplest non-trivial Riemann surface is the Riemann sphere $\mathbb{C}\cup\{\infty\}$, which adds a single point at infinity to the complex plane. We can visualize this using stereographic projection: imagine the complex plane sitting flat, and a sphere sitting on top of it with one point at the bottom touching the origin. For any point in the complex plane, draw a line to the north pole of the sphere; this line intersects the sphere at exactly one other point. This gives a one-to-one correspondence between $\mathbb{C}$ and the sphere minus the north pole. We declare that $\infty$ corresponds to the north pole. The Riemann sphere is a compact Riemann surface of genus 0. Its significance: every rational function $p(z)/q(z)$ (quotient of polynomials) extends to a holomorphic map on the Riemann sphere, including behavior at its poles and at infinity. Complex Tori: Genus 1 Surfaces Now imagine a parallelogram in the complex plane spanned by vectors $\omega1$ and $\omega2$ (where $\omega1/\omega2$ is not real). Identify opposite edges: points on the left edge are glued to corresponding points on the right edge, and points on the bottom edge are glued to corresponding points on the top edge. The resulting surface is a donut or torus. This is a complex torus, a compact Riemann surface of genus 1. Algebraically, a complex torus can be written as $\mathbb{C}/\Lambda$, where $\Lambda = \{m\omega1 + n\omega2 \mid m,n \in \mathbb{Z}\}$ is the lattice of integer combinations of $\omega1$ and $\omega2$. We identify two complex numbers $z$ and $w$ if their difference lies in $\Lambda$. Higher-Genus Surfaces By gluing together more complex polygons with appropriate identifications, we create surfaces with more "holes." A surface with $g$ holes has genus $g$. The torus has genus 1, the sphere has genus 0, and we can construct surfaces of any genus $g \geq 2$. Higher-genus surfaces are more complicated topologically, but each carries a natural complex structure. They correspond to algebraic curves of increasingly intricate type. Topological Classification An essential fact about Riemann surfaces is that compact ones are classified by a single topological invariant: genus $g$. The genus is the number of "holes" or "handles." The sphere has genus 0, the torus has genus 1, and so on. More precisely, every compact Riemann surface is homeomorphic (topologically equivalent) to a sphere with $g$ handles attached, and the genus $g$ uniquely determines the topological type. This classification means that, from a topological viewpoint, there are infinitely many "essentially different" compact Riemann surfaces—one for each non-negative integer $g$. Behavior of Holomorphic Functions Once we have Riemann surfaces, we can do complex analysis on them. Analytic Continuation Analytic continuation describes how a holomorphic function defined on a small open set can be extended along paths on the surface. Suppose we have a holomorphic function $f$ defined on a small open disk $D$ in the surface. If we have a path $\gamma$ starting at a point in $D$, we can often extend $f$ along $\gamma$. At each point along $\gamma$, we find a small disk where $f$ is still holomorphic, and we extend using the identity theorem for holomorphic functions (which says holomorphic functions are determined by their values on any open set). The key question: if we extend $f$ along two different paths from the starting point to the same endpoint, do we get the same function? Monodromy and Multivaluedness If we start with a function $f$ on a small open set and perform analytic continuation around a closed loop, we might arrive back at the starting point with a different function. The difference encodes the multivaluedness of the function. This is exactly what happens with $\log z$: analytic continuation around a small loop containing the origin increases the logarithm by $2\pi i$. The monodromy theorem states something reassuring: if the Riemann surface is simply connected (no "holes" from a topological standpoint), then analytic continuation around any closed loop gives the same function back. In simply connected spaces, there is no obstruction to single-valuedness. <extrainfo> The Uniformization Theorem The uniformization theorem is one of the deepest results in complex analysis. It states: > Every simply connected Riemann surface is conformally equivalent to exactly one of three universal models: the Riemann sphere $\mathbb{C}\cup\{\infty\}$, the complex plane $\mathbb{C}$, or the open unit disk $\mathbb{D} = \{z \in \mathbb{C} \mid |z| < 1\}$. "Conformally equivalent" means there exists a holomorphic bijection with holomorphic inverse between them. The three cases roughly correspond to: Sphere: surfaces of genus 0 (compact) Plane: non-compact surfaces with specific growth properties Unit disk: most other simply connected surfaces (including higher-genus surfaces viewed as covers) This theorem unifies enormous classes of surfaces under three canonical models. However, it typically appears in advanced courses rather than introductory ones. </extrainfo> Connections to Algebraic Geometry One of the remarkable aspects of Riemann surfaces is that they connect complex analysis to algebraic geometry. Algebraic Curves as Riemann Surfaces An algebraic curve is defined by a polynomial equation in two variables, like $y^2 = x^3 - x$. (More generally, a curve is a one-dimensional algebraic variety.) The key insight: a smooth algebraic curve naturally carries a Riemann surface structure. Conversely, every compact Riemann surface of genus $g$ can be realized as a smooth algebraic curve of some degree determined by $g$. For instance, an algebraic curve of genus 1 can be put in the form $y^2 = x^3 + ax + b$ (Weierstrass form) and corresponds to a complex torus. This equivalence means that topological/geometric properties of Riemann surfaces (like genus) translate into algebraic properties of the defining polynomial, and vice versa. It's a beautiful bridge between two major areas of mathematics. Summary Riemann surfaces are the natural arena for complex analysis because they resolve the fundamental problem of multivalued functions by unfolding the complex plane. They are defined through local complex coordinates (charts) that are compatible via holomorphic transition maps, giving them a well-defined complex structure. The main examples—the Riemann sphere, complex tori, and higher-genus surfaces—are classified by genus. Functions on Riemann surfaces can be analytically continued along paths, and analytic continuation around loops reveals whether the surface "produces" multivaluedness. Finally, compact Riemann surfaces correspond exactly to smooth algebraic curves, creating a profound link between complex analysis and algebraic geometry.