Define and explore surfaces through parametrization
A surface is a sheet drawn in space by two dials, u and v, much like latitude and longitude. At every point it has a flat tangent plane and an arrow sticking straight out.
Switch surfaces and hover to read (u, v) at any point. Cyan lines hold v fixed, magenta lines hold u fixed; notice the grid bunching up near the sphere's poles.
The classic 2-sphere
A unit sphere centered at the origin
A parametric surface is a 2D manifold embedded in 3D space, defined by a function S(u, v) = (x(u,v), y(u,v), z(u,v)). The parameters u and v serve as coordinates on the surface, like latitude and longitude on Earth.
At each point, the surface has a tangent plane spanned by the partial derivatives Su and Sv, and a normal vector N perpendicular to it. These form the foundation for measuring curvature.
Click anywhere on the surface to place its tangent plane, with Su (cyan), Sv (pink), and the normal N (green). Move the point and watch the plane tilt.
These vectors are tangent to the surface: Su points in the direction of increasing u (along u-curves), Sv points in the direction of increasing v.
The unit normal vector is perpendicular to both tangent vectors. It determines the “outward” direction at each point (for orientable surfaces).
At each point p on a smooth surface, the tangent plane TpS is the best linear approximation to the surface near p. It contains all vectors tangent to curves passing through p.
Any tangent vector can be written as aSu + bSv for some coefficients a and b. This makes {Su, Sv} a natural basis for the tangent space — though not necessarily orthonormal!
Spot a mistake or have an idea?