Applications

In general relativity, the metric $g$ on a 4-dimensional manifold encodes gravity. Geodesics of the metric describe the trajectories of freely falling particles. The curvature of spacetime is determined by the distribution of matter and energy via the Einstein field equations.

Example (Schwarzschild Metric). The metric outside a spherically symmetric mass $M$ is:

$ds^2 = -\left(1 - \frac{2GM}{c^2 r}\right)c^2 dt^2 + \left(1 - \frac{2GM}{c^2 r}\right)^{-1} dr^2 + r^2 d\Omega^2$

This metric describes non-rotating black holes and predicts phenomena such as gravitational redshift, perihelion precession, and gravitational lensing.

Connections to Curvature. The Riemann curvature tensor $R^\rho_{\ \sigma\mu\nu}$ encodes the tidal forces experienced by extended bodies. The Ricci tensor $R_{\mu\nu} = R^\rho_{\ \mu\rho\nu}$ and scalar curvature $R = g^{\mu\nu}R_{\mu\nu}$ appear directly in the Einstein field equations:

$R_{\mu\nu} - \frac{1}{2}Rg_{\mu\nu} + \Lambda g_{\mu\nu} = \frac{8\pi G}{c^4}T_{\mu\nu}$

In Yang-Mills gauge theory, the connection (gauge field) on a principal $G$-bundle over spacetime plays a role analogous to the Levi-Civita connection in Riemannian geometry. The curvature of this connection is the field strength $F = dA + A \wedge A$.

The Chern-Weil theory relates the curvature of a connection to topological invariants (characteristic classes) of the bundle. For example, the first Chern class $c_1(E) \in H^2(M, \mathbb{R})$ is represented by the closed 2-form $\frac{i}{2\pi}\mathrm{tr}(F)$.

Instantons. Self-dual connections on $S^4$ (or $\mathbb{R}^4$) with finite action correspond to instantons. Their topological charge is given by the second Chern number, which is an integer classifying the gauge field configuration.

A minimal surface is a surface with zero mean curvature $H = 0$. Such surfaces locally minimise area and arise as soap films spanning wire frames.

Plateau’s Problem. Given a closed curve in $\mathbb{R}^3$, find the surface of minimal area spanning it. The solution is a minimal surface, and its existence follows from geometric measure theory. The helicoid, catenoid, and Enneper surface are classic examples.

Applications. Minimal surfaces appear in lipid bilayer membranes (biological cell walls), architecture (lightweight tensile structures), and materials science (block copolymer interfaces).

The configuration space of a robotic arm is a manifold. For a $n$-joint arm, the space of possible configurations is an $n$-dimensional manifold, often a product of circles $S^1$ for rotational joints.

Motion Planning. Paths in the configuration manifold correspond to motions of the robot. Obstacles in physical space create obstacles in the configuration manifold. Finding a collision-free path is equivalent to finding a curve in the manifold that avoids obstacle regions.

Riemannian metrics on the configuration space encode the kinetic energy of the robot arm, leading to geodesic equations for minimum-energy motion.

Discrete differential geometry approximates smooth manifolds by polyhedral meshes. Curvature is discretised using angle defects at vertices: the Gaussian curvature at a vertex of a triangle mesh is $K(v) = 2\pi - \sum_i \theta_i$, where $\theta_i$ are the interior angles meeting at $v$.

Applications. Shape analysis uses the Laplace-Beltrami operator $\Delta$ on a manifold for spectral shape descriptors, mesh denoising, and shape correspondence. The eigenfunctions of $\Delta$ generalise Fourier modes to curved surfaces.

Texture mapping uses parametrisations of surfaces onto planar domains, minimising distortion measured by the first fundamental form.

The Gauss-Bonnet theorem relates the total Gaussian curvature of a surface to its Euler characteristic:

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

For a compact orientable surface without boundary, $\int_M K\,dA = 2\pi\chi(M)$. This explains why a sphere (Euler characteristic 2) cannot be mapped to a plane without distortion: the total curvature is fixed at $4\pi$.

Consequences. No flat map of the Earth can preserve both angles and areas. The Mercator projection preserves angles (it is conformal) but greatly distorts areas near the poles.

In string theory, the worldsheet of a propagating string is a 2-dimensional Lorentzian manifold embedded in a higher-dimensional spacetime. The equations of motion derive from the Polyakov action, whose symmetries include diffeomorphism invariance and Weyl rescaling.

Calabi-Yau manifolds are compact Ricci-flat K”ahler manifolds used for the extra spatial dimensions in string compactifications. Their holonomy group $SU(n)$ preserves supersymmetry, and their Hodge numbers determine the particle spectrum of the effective 4-dimensional theory.

Mirror symmetry relates pairs of Calabi-Yau manifolds whose Hodge diamonds are reflected: $h^{p,q}(M) = h^{\dim M - p,q}(M^\vee)$. This symmetry has deep implications for enumerative geometry and has led to predictions of curve counts on Calabi-Yau threefolds.

Manifold hypothesis. Many high-dimensional data sets lie on or near a low-dimensional manifold embedded in the ambient space. Dimensionality reduction algorithms exploit this.

ISOMAP constructs geodesic distances on the data manifold by approximating them with shortest paths on a nearest-neighbour graph, then uses multidimensional scaling to find a low-dimensional embedding.

t-SNE and UMAP use local neighbourhood information to construct embeddings that preserve the local geometry of the data manifold, revealing clustering and topological structure.

1. Verify that the Schwarzschild metric satisfies the vacuum Einstein equations $R_{\mu\nu} = 0$ outside the mass.
2. Compute the Gaussian curvature of a catenoid parametrised by $\mathbf{r}(u,v) = (\cosh u \cos v, \cosh u \sin v, u)$.
3. Explain why the Euler characteristic of a torus is zero and relate this to the total curvature via Gauss-Bonnet.
4. For a 2-joint planar robot arm, describe its configuration manifold and sketch how an obstacle in physical space maps to a forbidden region in the configuration space.
5. Show that the first Chern class of a line bundle over $S^2$ is the magnetic charge of a Dirac monopole.

Differential geometry finds applications across physics, engineering, and data science. General relativity uses curvature to model gravity. Gauge theory describes fundamental forces via connections on bundles. Minimal surfaces model soap films and biological membranes. Robotics uses configuration manifolds for motion planning. Discrete differential geometry powers modern computer graphics. The Gauss-Bonnet theorem forces trade-offs in cartography. String theory uses Calabi-Yau manifolds, and manifold learning algorithms exploit geometry for data analysis.