Surface (mathematics)

In mathematics, a surface is a generalization of a plane which needs not be flat (curvature is not necessarily zero). This is analogous to a curve generalizing a straight line. There are several more precise definitions, depending on the context and the mathematical tools that are used for the study.

Definitions

Often, a surface is defined by equations that are satisfied by the coordinates of its points. This is the case of the graph of a continuous function of two variables. The set of the zeros of a function of three variables is a surface, which is called an implicit surface.^[1] If the defining three-variate function is a polynomial, the surface is an algebraic surface. For example the unit sphere is an algebraic surface, as it may be defined by the implicit equation

$x^2+y^2+z^2 -1= 0.$

A surface may also be defined as the image, in some space of dimension at least 3, of a continuous function of two variables (some further conditions are required to insure that the image is not a curve). In this case, one says that one has a parametric surface, which is parametrized by these two variables, called parameters, in this case. For example the unit sphere may be parametrized by the Euler angles, also called longitude u and latitude v by

$\begin{align} x&= \cos(u)\cos(v)\\ y&=\sin(u)\cos(v)\\ z&=\sin(v)\,. \end{align}$

Parametric equations of surfaces are often irregular at some points. For example, all but two points of the unit sphere, are the image, by the above parametrization, of exactly one pair of Euler angles (modulo 2π). For the remaining two points (the north and south poles), one has cos v = 0, and the longitude u may take any values. Also, there are surfaces for which there cannot exits a single parametrization that covers the whole surface. Therefore one often considers surfaces which are parametrized by several parametric equations, whose images covers the surface. This is formalized by the concept of manifold: in the context of manifolds, typically in topology and differential geometry, a surface is a manifold of dimension two; this means that a surface is a topological space such that every point has a neighborhood which is homeomorphic to an open subset of the Euclidean plane (see Surface (topology) and Surface (differential geometry)). This allows defining surfaces in spaces of dimension higher than three, and even abstract surfaces, which are not contained in any other space. On the other hand this excludes surfaces that have singularities, such as the vertex of a conical surface or points where a surface crosses itself.

In classical geometry, a surface is generally defined as a locus of a point or a line. For example, a sphere is the locus of a point which is at a given distance of a fixed point, called the center; a conical surface is the locus of a line passing through a fixed point and crossing a curve; a surface of revolution is the locus of a curve rotating around a line. A ruled surface is the locus of a moving line satisfying some constraints; in modern terminology, a ruled surface is a surface, which is a union of lines.

Terminology

In this article, several kinds of surfaces are considered and compared. A non-ambiguous terminology is thus necessary for distinguish them. Therefore, we call topological surfaces the surfaces that are manifolds of dimension two (the surfaces considered in Surface (topology)). We call differential surfaces the surfaces that are differentiable manifolds (the surfaces considered in Surface (differential geometry)). Every differential surface is a topological surface, but the contrary is false.

For simplicity, unless otherwise stated, "surface" will mean a surface in the Euclidean space of dimension 3 or in R³. A surface, that is not supposed to be included in another space is called an abstract surface.

Examples

• The graph of a continuous function of two variables, defined over a connected open subset of R² is a topological surface. If the function is differentiable, the graph is a differential surface.
• A plane is together an algebraic surface and a differentiable surface. It is also a ruled surface and a surface of revolution.
• A circular cylinder (that is the locus of a line crossing a circle and parallel to a given direction) is an algebraic surface and a differential surface.
• A circular cone (locus of a line crossing a circle, and passing through a fixed point, the apex, which is outside the plane of the circle) is an algebraic surface, which is not a differential surface. If one remove the apex, the remaining of the cone is the union of two differential surfaces.
• The surface of a polyhedron is a topological surface, which is not a differential surface nor an algebraic surface.
• An hyperbolic paraboloid (the graph of the function z = xy) is a differential surface and an algebraic surface. It is also a ruled surface, and, for this reason, is often used in architecture.
• A two-sheet hyperboloid is an algebraic surface and the union of two non-intersecting differential surfaces.

Parametric surface

A parametric surface is the image of an open subset of the Euclidean plane (typically $\Bbb R^2$) by a continuous function, in a topological space, generally an Euclidean space of dimension at least three. Usually the function is supposed to be continuously differentiable, and this will be always the case in this article.

Specifically, a parametric surface in $\Bbb R^3$ is given by three functions of two variables u and v, called parameters

$\begin{align} x&=f_1(u,v)\\ y&=f_2(u,v)\\ z&=f_3(u,v)\,. \end{align}$

As the image of such a function may be a curve (for example if the three functions are constant with respect to v), a further condition is required, generally that, for almost all values of the parameters, the Jacobian matrix

$\begin{bmatrix} \dfrac{\partial f_1}{\partial u} & \dfrac{\partial f_2}{\partial u} & \dfrac{\partial f_1}{\partial u} \\[12pt] \dfrac{\partial f_1}{\partial v} & \dfrac{\partial f_2}{\partial v} & \dfrac{\partial f_1}{\partial v} \end{bmatrix}$

has rank two. Here "almost all" means that the values of the parameters where the rank is two contain a dense open subset of the range of the parametrization. For surfaces in a space of higher dimension, the condition is the same, except for the number of columns of the Jacobian matrix.

Tangent plane and normal vector

A point p where the above Jacobian matrix has rank two is said regular, or, more properly, the parametrization is said regular ar p.

The tangent plane at a regular point p is the unique plane passing through p, and having a direction parallel to the two row vectors of the Jacobian matrix. The tangent plane is an affine concept, because its definition is independent of the choice of a metric. In other words, any affine transformation maps the tangent plane to the surface at a point to the tangent plane to the image of the surface at the image of the point.

The normal line, or simply normal at a point of a surface is the unique line passing through the point and perpendicular to the tangent plane. A normal vector is a vector which is parallel to the normal.

For other differential invariants of surfaces, in the neighborhood of a point, see Differential geometry of surfaces

Irregular point and singular point

A point of a parametric surface, which is not regular is irregular. There are several kinds of irregular points.

It may occur that an irregular point becomes regular, if one changes of parametrization. This is the case of the poles in the parametrization of the unit sphere by Euler angles: it suffices to permute the role of the different coordinate axes for changing the poles.

On the other hand, let us consider the circular cone of parametric equation

$\begin{align} x&= t\cos(u)\\ y&=t\sin(u)\\ z&=t\,. \end{align}$

The apex of the cone is the origin (0, 0, 0), and is obtained for t = 0. It is an irregular point that remains irregular, whichever parametrization is chosen (otherwise, there wold exist a unique tangent plane). Such an irregular point, where the tangent plane is undefined, is said singular.

There is another kind of singular points. There are the self-crossing points, that is the points where the surface crosses itself. In other words, these are the points which are obtained for (at least) two different values of the parameters.

Graph of a bivariate function

Let z = f(x, y) be a function of two real variables. This is a parametric surface, parametrized as

$\begin{align} x&= t\\ y&=u\\ z&=f(t,u)\,. \end{align}$

Every point of this surface is regular, as the two first columns of the Jacobian matrix form the identity matrix of rank two.

Rational surface

A rational surface is a surface that may be parametrized by rational functions of two variables. That is, if f_i(t, u) are, for i = 0, 1, 2, 3, polynomials in two indeterminates, then the parametric surface, defined by

$\begin{align} x&= \frac{f_1(t,u)}{f_0(t,u)}\\ y&=\frac{f_2(t,u)}{f_0(t,u)}\\ z&=\frac{f_1(t,u)}{f_0(t,u)}\,, \end{align}$

is a rational surface.

A rational surface is an algebraic surface, but most algebraic surfaces are not rational.

Implicit surface

An implicit surface in a Euclidean space (or, more generally, in an Affine space) of dimension 3 is the set of the common zeros of a differentiable function of three variables

$f(x, y, z)=0$

Implicit means, that the equation defines implicitly one of the variables as a function of the other variables. This is made more explicit by the implicit function theorem: if f(x₀, y₀, z₀) = 0, and the partial derivative in z of f is not zero at (x₀, y₀, z₀), then there exists a differentiable function φ(x, y) such that

$f(x,y,\varphi(x,y))=0$

in a neighbourhood of (x₀, y₀, z₀). In other words, the implicit surface is the graph of a function nearby a point of the surface where the partial derivative in z is nonzero. It has thus, locally, a parametric representation, except at the points of the surface where the three partial derivatives are zero.

Regular points and tangent plane

A point of the surface, where at least one partial derivatives of f is nonzero is said regular. At such a point $(x_0, y_0, z_0)$, the tangent plane and the direction of the normal are well defined, and may be deduced, with the implicit function theorem from the definition given above, in § Tangent plane and normal vector. The direction of the normal is the gradient, that is the vector

$\left[\frac{\partial f}{\partial x}(x_0, y_0, z_0), \frac{\partial f}{\partial y}(x_0, y_0, z_0), \frac{\partial f}{\partial z}(x_0, y_0, z_0)\right].$

The tangent plane is defined by its implicit equation

$\frac{\partial f}{\partial x}(x_0, y_0, z_0)(x-x_0) + \frac{\partial f}{\partial y}(x_0, y_0, z_0) (y-y_0)+ \frac{\partial f}{\partial z}(x_0, y_0, z_0)(z-z_0) = 0.$

Singular point

A singular point of an implicit surface (in $\Bbb R^3$ is a point of the surface, where implicit equation and its three partial derivatives are all zero. Therefore the singular points are the solutions of a system of four equations in three indeterminates. As most such systems have no solution, many surfaces do not have any singular point. A surface without singular point is said regular or non-singular.

The study of surfaces nearby their singular points, and the classification of the singular points is the Singularity theory. A singular point is isolated if there is no other singular point in a neighborhood of it. Otherwise, the singular points may form a curve. This is in particular the case for self-crossing surfaces.

Algebraic surface

Topological 2-manifold

Differential 2-manifold

Fractal surface

In nature

The concept of surface is widely used in physics, engineering, computer graphics, and many other disciplines, primarily in representing the surfaces of physical objects. For example, in analyzing the aerodynamic properties of an airplane, the central consideration is the flow of air along its surface.

A surface may be the idealized limit between two fluids (the surface of the sea) or the idealized boundary of a solid (the surface of a ball). Many surfaces considered in Physics and Chemistry are interfaces. However, they are surfaces only at macroscopic scale. At microscopic scale, they may have some thickness. At atomic scale, they do not look at all as a surface, because of holes formed by spaces between atoms or molecules.

Other surfaces considered in physics are equipotential surfaces and wavefronts. The wavefront, which has been first studied by Fresnel, is called wave surface by mathematicians.

The shape of soap bubbles is mathematically explained by the theory of minimal surfaces.

The surface of the reflector of a telescope is a paraboloid of revolution.

Atmospheric boundaries (tropopause, edge of space, plasmapause, etc.) are also natural surfaces.

In computer graphics

See also

• Coordinate surfaces
• Polyhedral surface
• Signed distance function
• Surface integral

Notes

1. Here implicit does not refer to a property of the surface, which may be defined by other means, but it refers to how it is defined. Thus this term is an abbreviation of "surface defined by an implicit equation".