9.2 Tensor Basis Coordinate Areas

Tensor Basis Coordinate Areas is the classification of the distinct kinds of coordinate systems commonly paired with a choice of basis for tensor work — rectilinear (Cartesian and general affine) systems, curvilinear systems such as polar, cylindrical, and spherical coordinates, and specially adapted systems such as normal or geodesic coordinates — together with the characteristic way each kind organizes its basis vectors and metric components across the region it covers. It surveys the recurring "areas," in the sense of families or regimes, into which practical coordinate choices fall, rather than any single coordinate system in isolation.

Rectilinear Coordinate Areas

Cartesian Coordinates and Their Constant Orthonormal Basis

Cartesian coordinates are the simplest coordinate area: the coordinate basis vectors have constant length, are mutually orthogonal, and do not vary in direction from point to point, so the metric tensor takes the identity matrix form everywhere the coordinates are defined. This constancy is what makes upper and lower tensor components coincide numerically throughout a Cartesian region, and is the reason elementary vector calculus can often avoid the upper/lower index distinction entirely.

General Affine (Oblique) Coordinates

A broader rectilinear area allows the basis vectors to remain constant throughout the region but drops the requirements of unit length and mutual orthogonality; an affine or oblique coordinate system still has a constant, position-independent basis, but its metric tensor, while still constant, is no longer the identity, requiring the full metric-based machinery of raising and lowering indices even though the basis itself does not vary with position.

Curvilinear Coordinate Areas

Polar, Cylindrical, and Spherical Coordinates

Curvilinear coordinate areas are characterized by basis vectors that vary in direction, length, or both as position changes; polar coordinates in the plane, and their cylindrical and spherical generalizations in three dimensions, are the most commonly encountered members of this area, with metric components that depend explicitly on position (for instance, the r² factor multiplying the angular metric component in polar coordinates) and with coordinate singularities typically appearing at the origin or along a designated axis.

General Curvilinear Systems on a Manifold

Beyond these named examples, any coordinate system on a curved manifold, or any nonlinear reparametrization of a flat space, falls into the general curvilinear coordinate area: the defining feature is a position-dependent metric and, usually, position-dependent basis vectors, requiring covariant differentiation (via connection coefficients) to differentiate tensor fields correctly, since ordinary partial differentiation of components no longer accounts for the basis vectors' own variation.

Specially Adapted Coordinate Areas

Normal (Geodesic) Coordinates

Normal coordinates are constructed to be as close to Cartesian as possible in a small neighborhood of a single chosen point on a curved manifold, arranging for the metric to equal the identity and its first derivatives to vanish exactly at that point, even though the coordinate area's metric generally departs from this idealized form away from the chosen point; this area is used precisely because it isolates curvature effects to the second-derivative terms of the metric, simplifying calculations local to one point.

Comoving and Adapted Coordinates in Applications

In physical applications, coordinate areas are sometimes chosen to be adapted to a particular structure of interest — coordinates that move along with a physical system, or that are aligned with a preferred family of surfaces or curves — trading simplicity of the metric's general form for simplicity in describing the specific structure the coordinates were built around, illustrating that the choice of coordinate area is generally driven by which features of a problem are meant to be made as transparent as possible.

Diagram Comparing the Basis Behavior Across Coordinate Areas

Choosing a Coordinate Area for a Given Task

Matching the Area to the Symmetry of the Problem

A practical guideline underlying the choice among these coordinate areas is to match the coordinate system's own symmetry to the symmetry of the tensor field or geometric configuration under study, since a coordinate area aligned with a problem's natural symmetry (spherical coordinates for a spherically symmetric field, for instance) typically produces the simplest possible component expressions, while a poorly matched choice can obscure structure that is otherwise straightforward.

Areas Are Not Mutually Exclusive Within a Single Problem

A single extended calculation often moves between coordinate areas as convenient — using normal coordinates to analyze behavior local to one point, then switching to a curvilinear system adapted to the problem's global symmetry, then converting to Cartesian coordinates for a final numerical evaluation — with the coordinate change scope and transition transformations connecting each area to the next providing the means of moving consistently from one to another as the calculation proceeds.