The Gauss-Bonnet Theorem

The Gauss-Bonnet theorem is one of the most profound results in mathematics, connecting local differential geometry (curvature) to global topology (Euler characteristic). It shows that the total curvature of a surface is a topological invariant—unchanged by any smooth deformation of the surface.

Historical Context

The theorem has a long history. Gauss proved the case for geodesic triangles on surfaces in his Disquisitiones (1827), showing that the angular excess of a geodesic triangle equals its integral curvature. Pierre Ossian Bonnet extended this to arbitrary compact surfaces in 1848.

The higher-dimensional generalization was achieved by Allendoerfer and Weil (1943) and independently by Chern (1944). Chern's proof was particularly influential, as it introduced methods that would lead to the theory of characteristic classes and the Chern-Gauss-Bonnet theorem.

The Gauss-Bonnet theorem can be viewed as the simplest case of the Atiyah-Singer index theorem and serves as a bridge between differential geometry, algebraic topology, and theoretical physics.

Derivation 1: Gaussian Curvature

For a 2-dimensional Riemannian manifold, the Gaussian curvature $K$ at a point is the product of the principal curvatures $\kappa_1$ and $\kappa_2$ (for embedded surfaces), or intrinsically:

$K = \frac{R_{1212}}{g_{11}g_{22} - g_{12}^2} = \frac{R_{1212}}{\det(g)}$

Gauss's Theorema Egregium

Gauss's "remarkable theorem" states that $K$ depends only on the metric $g_{ij}$ and its first two derivatives—it is an intrinsic invariant. For a metric $ds^2 = E\,du^2 + 2F\,du\,dv + G\,dv^2$:

$K = \frac{1}{2\sqrt{EG-F^2}}\left[\frac{\partial}{\partial v}\frac{E_v - F_u}{\sqrt{EG-F^2}} + \frac{\partial}{\partial u}\frac{F_v - G_u}{\sqrt{EG-F^2}}\right]$

Examples

$\text{Sphere } S^2(R): \quad K = \frac{1}{R^2} > 0$

$\text{Plane } \mathbb{R}^2: \quad K = 0$

$\text{Hyperbolic plane } \mathbb{H}^2: \quad K = -\frac{1}{R^2} < 0$

$\text{Torus: } K \text{ varies (positive outside, negative inside, total = 0)}$

Physical intuition: Gaussian curvature measures how much a surface differs from being flat. A surface with $K > 0$ (sphere) cannot be flattened without tearing; $K = 0$ (cylinder, cone) can be unrolled to a plane; $K < 0$ (saddle) curves oppositely in different directions.

Derivation 2: The Euler Characteristic

The Euler characteristic $\chi(M)$ is a topological invariant that can be computed from any triangulation of $M$:

$\chi(M) = V - E + F$

where $V$, $E$, $F$ are the numbers of vertices, edges, and faces. This is independent of the triangulation—a deep topological fact.

Classification of Surfaces

For a closed orientable surface of genus $g$ (number of handles):

$\chi(\Sigma_g) = 2 - 2g$

$g = 0 \text{ (sphere):} \quad \chi = 2$

$g = 1 \text{ (torus):} \quad \chi = 0$

$g = 2 \text{ (double torus):} \quad \chi = -2$

Euler's polyhedron formula (1758): For convex polyhedra,$V - E + F = 2$. This was the first topological invariant ever discovered and marks the birth of topology. It holds for any polyhedron homeomorphic to the sphere.

Derivation 3: The Gauss-Bonnet Theorem

For a compact oriented 2-manifold $M$ without boundary:

$\boxed{\int_M K\,dA = 2\pi\,\chi(M)}$

Proof Sketch

Triangulate $M$ into geodesic triangles $\Delta_k$. For each triangle with interior angles $\alpha_i, \beta_i, \gamma_i$, the local Gauss-Bonnet gives:

$\int_{\Delta_k} K\,dA = (\alpha_k + \beta_k + \gamma_k) - \pi$

Summing over all $F$ triangles:

$\int_M K\,dA = \sum_{k=1}^F (\alpha_k + \beta_k + \gamma_k) - \pi F$

At each interior vertex, the angles sum to $2\pi$. Each edge is shared by two triangles. Each triangle contributes 3 edges, so $3F = 2E$. The total angle sum is $2\pi V$. Therefore:

$\int_M K\,dA = 2\pi V - \pi F = 2\pi V - 2\pi E + 2\pi F = 2\pi(V - E + F) = 2\pi\chi(M)$

Remarkable consequence: No matter how you deform a sphere (stretch, compress, wrinkle), the total curvature always equals $4\pi$. Regions of positive curvature must exactly compensate regions of negative curvature to maintain this topological constraint.

Derivation 4: Gauss-Bonnet with Boundary

For a compact oriented surface $M$ with smooth boundary $\partial M$:

$\int_M K\,dA + \oint_{\partial M} \kappa_g\,ds = 2\pi\,\chi(M)$

where $\kappa_g$ is the geodesic curvature of the boundary. If the boundary has corners with exterior angles $\theta_i$:

$\int_M K\,dA + \oint_{\partial M} \kappa_g\,ds + \sum_i \theta_i = 2\pi\,\chi(M)$

Application: Geodesic Polygons on $S^2$

For a geodesic polygon on the unit sphere ($K = 1$, $\kappa_g = 0$ along edges), with interior angles $\alpha_1, \ldots, \alpha_n$ and exterior angles $\theta_i = \pi - \alpha_i$:

$\text{Area} = \sum_{i=1}^n \alpha_i - (n-2)\pi$

This is the spherical excess formula. For a geodesic triangle with angles summing to $\pi + \epsilon$, the area equals $\epsilon$.

Navigation application: Navigators have used the spherical excess formula for centuries. On Earth (radius 6371 km), a triangle with vertices at the North Pole, Quito (Ecuador), and a point on the prime meridian at the equator has area$\frac{1}{8}$ of the total surface, confirming the formula.

Derivation 5: The Generalized Gauss-Bonnet (Chern-Gauss-Bonnet)

For a compact oriented Riemannian manifold $M^{2n}$ of even dimension $2n$ without boundary, the Chern-Gauss-Bonnet theorem states:

$\boxed{\int_{M^{2n}} \text{Pf}(\Omega) = (2\pi)^n\,\chi(M)}$

where $\text{Pf}(\Omega)$ is the Pfaffian of the curvature 2-form$\Omega$. The Pfaffian is defined for antisymmetric matrices by $\text{Pf}(A)^2 = \det(A)$.

The 4-Dimensional Case

In dimension 4, the integrand is the Euler density:

$e_4 = \frac{1}{32\pi^2}\epsilon^{ijkl}R_{ijmn}R_{kl}{}^{mn}$

which can be rewritten using the Riemann tensor decomposition:

$\chi(M^4) = \frac{1}{8\pi^2}\int_{M^4}\left(|W|^2 - 2|E|^2 + \frac{R^2}{24}\right)dV$

where $W$ is the Weyl tensor, $E$ the traceless Ricci tensor, and $R$ the scalar curvature.

In physics: The 4D Euler density appears in the gravitational action as the Gauss-Bonnet term $\mathcal{G} = R^2 - 4R_{\mu\nu}R^{\mu\nu} + R_{\mu\nu\rho\sigma}R^{\mu\nu\rho\sigma}$. In 4D, this is topological (doesn't affect equations of motion), but in higher-dimensional Lovelock gravity, it contributes dynamical equations.

Interactive Simulation

This simulation verifies the Gauss-Bonnet theorem on the sphere and torus, computes the Euler characteristic for various surfaces, demonstrates the spherical excess formula for geodesic polygons, and shows numerical convergence of the curvature integral to $4\pi$ on $S^2$.

Gauss-Bonnet Theorem: Curvature, Euler Characteristic & Topology

Python

script.py185 lines

#!/usr/bin/env python3
"""
gauss_bonnet.py - Gaussian Curvature, Euler Characteristic, and the Gauss-Bonnet Theorem
Verifies the theorem on various surfaces and visualizes curvature distributions
"""
import numpy as np
import matplotlib
matplotlib.use('Agg')
import matplotlib.pyplot as plt
import base64
from io import BytesIO

# ============================================================
# 1. Gaussian curvature of various surfaces
# ============================================================
print("=" * 72)
print("  GAUSS-BONNET THEOREM: CURVATURE MEETS TOPOLOGY")
print("=" * 72)
print()

# Sphere of radius R
print("  GAUSSIAN CURVATURE OF CLASSICAL SURFACES")
print("  " + "-" * 50)
print()

for R in [0.5, 1.0, 2.0, 5.0]:
    K = 1.0 / R**2
    integral = 4 * np.pi * R**2 * K  # = 4*pi
    print("    Sphere R={:.1f}: K = {:.4f}, integral(K dA) = {:.6f} = 4*pi".format(
        R, K, integral))
print()

# Torus: K varies
print("  Torus (R=2, r=1):")
theta_torus = np.linspace(0, 2*np.pi, 500)
R_major, r_minor = 2.0, 1.0
K_torus = np.cos(theta_torus) / (r_minor * (R_major + r_minor * np.cos(theta_torus)))
total_K_torus = np.trapz(K_torus * r_minor * (R_major + r_minor * np.cos(theta_torus)),
                          theta_torus) * 2 * np.pi
print("    K ranges from {:.4f} to {:.4f}".format(K_torus.min(), K_torus.max()))
print("    Integral of K dA = {:.6f} (should be 0 = 2*pi*chi(T^2))".format(total_K_torus))
print("    Euler characteristic chi(T^2) = 0")
print()

# ============================================================
# 2. Gauss-Bonnet theorem verification
# ============================================================
print("  GAUSS-BONNET THEOREM")
print("  " + "-" * 50)
print()
print("  For a compact surface M without boundary:")
print("    integral(K dA) = 2*pi*chi(M)")
print()

surfaces = [
    ("Sphere S^2", 2, 1.0),
    ("Torus T^2", 0, 0.0),
    ("Double torus (g=2)", -2, -1.0),
    ("Triple torus (g=3)", -4, -2.0),
]

for name, chi_val, _ in surfaces:
    total = 2 * np.pi * chi_val
    print("    {}: chi = {}, integral(K dA) = {:.4f}".format(name, chi_val, total))
print()
print("  For genus g surface: chi = 2 - 2g")
print()

# ============================================================
# 3. Gauss-Bonnet with boundary
# ============================================================
print("  GAUSS-BONNET WITH BOUNDARY")
print("  " + "-" * 50)
print()
print("  For a compact surface M with smooth boundary:")
print("    integral(K dA) + integral(kappa_g ds) = 2*pi*chi(M)")
print()
print("  Geodesic triangle on unit sphere:")

# Spherical triangle with angles A, B, C
angles = [np.pi/2, np.pi/3, np.pi/4]
area_triangle = sum(angles) - np.pi
excess = sum(angles) - np.pi
print("    Angles: pi/2, pi/3, pi/4")
print("    Angular excess = {:.6f} rad".format(excess))
print("    Area = angular excess = {:.6f} (on unit sphere)".format(area_triangle))
print("    Verifies: integral(K dA) = angular excess (K=1 on S^2)")
print()

# ============================================================
# 4. Euler characteristic from triangulations
# ============================================================
print("  EULER CHARACTERISTIC FROM TRIANGULATIONS")
print("  " + "-" * 50)
print()
print("  chi = V - E + F (vertices - edges + faces)")
print()

triangulations = [
    ("Tetrahedron (= S^2)", 4, 6, 4),
    ("Cube (= S^2)", 8, 12, 6),
    ("Octahedron (= S^2)", 6, 12, 8),
    ("Icosahedron (= S^2)", 12, 30, 20),
]

for name, V, E, F in triangulations:
    chi = V - E + F
    print("    {}: V={}, E={}, F={}, chi={}".format(name, V, E, F, chi))
print()

# ============================================================
# Visualization
# ============================================================
fig, axes = plt.subplots(2, 2, figsize=(14, 11))
fig.patch.set_facecolor('#0a0a0a')

for ax in axes.flat:
    ax.set_facecolor('#111111')
    for spine in ax.spines.values():
        spine.set_color('#2dd4bf')
    ax.tick_params(colors='#99f6e4')
    ax.grid(True, color='#134e4a', alpha=0.4)

# Plot 1: Gaussian curvature on torus
ax1 = axes[0, 0]
ax1.plot(theta_torus, K_torus, color='#2dd4bf', linewidth=2.5)
ax1.fill_between(theta_torus, K_torus, 0, where=(K_torus > 0),
                  alpha=0.2, color='#34d399', label='K > 0 (outside)')
ax1.fill_between(theta_torus, K_torus, 0, where=(K_torus < 0),
                  alpha=0.2, color='#f472b6', label='K < 0 (inside)')
ax1.axhline(y=0, color='#6b7280', linewidth=1)
ax1.set_xlabel('theta (poloidal angle)', color='#5eead4')
ax1.set_ylabel('Gaussian curvature K', color='#5eead4')
ax1.set_title('Curvature on the Torus', color='#99f6e4', fontsize=12, fontweight='bold')
ax1.legend(facecolor='#111111', edgecolor='#134e4a', labelcolor='#99f6e4')

# Plot 2: Euler characteristic vs genus
ax2 = axes[0, 1]
genus = np.arange(0, 8)
chi_vals = 2 - 2 * genus
ax2.bar(genus, chi_vals, color='#2dd4bf', alpha=0.7, edgecolor='#34d399')
ax2.set_xlabel('Genus g', color='#5eead4')
ax2.set_ylabel('Euler characteristic chi', color='#5eead4')
ax2.set_title('chi = 2 - 2g for Orientable Surfaces', color='#99f6e4', fontsize=12, fontweight='bold')

# Plot 3: Angular excess on sphere
ax3 = axes[1, 0]
excess_angles = np.linspace(0, 2*np.pi, 100)
area_from_excess = excess_angles  # area = angular excess on unit sphere
ax3.plot(excess_angles * 180 / np.pi, area_from_excess, color='#f59e0b', linewidth=2.5)
ax3.set_xlabel('Angular excess (degrees)', color='#5eead4')
ax3.set_ylabel('Area of geodesic polygon', color='#5eead4')
ax3.set_title('Area vs Angular Excess on S^2', color='#99f6e4', fontsize=12, fontweight='bold')

# Plot 4: Curvature integral convergence
ax4 = axes[1, 1]
n_points = np.arange(10, 500, 10)
integrals = []
for n in n_points:
    th = np.linspace(0, np.pi, n)
    ph = np.linspace(0, 2*np.pi, n)
    dth = th[1] - th[0]
    dph = ph[1] - ph[0]
    total = 0.0
    for t in th[:-1]:
        total += np.sin(t) * dth * 2 * np.pi  # K=1 on unit sphere
    integrals.append(total)

ax4.plot(n_points, integrals, color='#34d399', linewidth=2)
ax4.axhline(y=4*np.pi, color='#f59e0b', linestyle='--', linewidth=1.5, label='4*pi (exact)')
ax4.set_xlabel('Number of grid points', color='#5eead4')
ax4.set_ylabel('Integral of K dA', color='#5eead4')
ax4.set_title('Numerical Gauss-Bonnet on S^2', color='#99f6e4', fontsize=12, fontweight='bold')
ax4.legend(facecolor='#111111', edgecolor='#134e4a', labelcolor='#99f6e4')

plt.tight_layout(pad=2.0)
buf = BytesIO()
plt.savefig(buf, format='png', dpi=150, bbox_inches='tight', facecolor='#0a0a0a')
buf.seek(0)
img_b64 = base64.b64encode(buf.read()).decode()
print('<img src="data:image/png;base64,' + img_b64 + '" alt="Gauss-Bonnet Theorem Visualization" />')
buf.close()
print()
print("Visualization complete.")

Click Run to execute the Python code

Code will be executed with Python 3 on the server

Summary

Gaussian Curvature

An intrinsic measure of curvature for 2-manifolds. Positive on spheres, zero on planes and cylinders, negative on saddle surfaces.

Euler Characteristic

The topological invariant $\chi = V - E + F = 2 - 2g$ for orientable surfaces. Computed from any triangulation and invariant under homeomorphism.

Gauss-Bonnet Theorem

The total Gaussian curvature $\int K\,dA = 2\pi\chi$ is a topological invariant. This is the paradigmatic result connecting local geometry to global topology.

Chern-Gauss-Bonnet

The generalization to higher even dimensions, using the Pfaffian of the curvature form. The 4D version connects to the Gauss-Bonnet gravity term in Lovelock theories.

← Geodesics Next: Fiber Bundles →