Spherical coordinates, also called spherical polar coordinates (Walton 1967, Arfken 1985), are a system of curvilinear coordinates that are natural for describing positions on a sphere or spheroid. Define to be the azimuthal angle in the -plane from the x-axis with (denoted when referred to as the longitude), to be the polar angle (also known as the zenith angle and colatitude, with where is the latitude) from the positive z-axis with , and to be distance (radius) from a point to the origin. This is the convention commonly used in mathematics.
In this work, following the mathematics convention, the symbols for the radial, azimuth, and zenith angle coordinates are taken as , , and , respectively. Note that this definition provides a logical extension of the usual polar coordinates notation, with remaining the angle in the -plane and becoming the angle out of that plane. The sole exception to this convention in this work is in spherical harmonics, where the convention used in the physics literature is retained (resulting, it is hoped, in a bit less confusion than a foolish rigorous consistency might engender).
Unfortunately, the convention in which the symbols and are reversed (both in meaning and in order listed) is also frequently used, especially in physics. This is especially confusing since the identical notation typically means (radial, azimuthal, polar) to a mathematician but (radial, polar, azimuthal) to a physicist. The symbol is sometimes also used in place of , instead of , and and instead of . The following table summarizes a number of conventions used by various authors. Extreme care is therefore needed when consulting the literature.
|(radial, azimuthal, polar)||this work|
|(radial, azimuthal, polar)||Apostol (1969, p. 95), Anton (1984, p. 859), Beyer (1987, p. 212)|
|(radial, polar, azimuthal)||ISO 31-11, Misner et al. (1973, p. 205)|
|(radial, polar, azimuthal)||Arfken (1985, p. 102)|
|(radial, polar, azimuthal)||Moon and Spencer (1988, p. 24)|
|(radial, polar, azimuthal)||Korn and Korn (1968, p. 60), Bronshtein et al. (2004, pp. 209-210)|
|(radial, polar, azimuthal)||Zwillinger (1996, pp. 297-299)|
The spherical coordinates are related to the Cartesian coordinates by
where , , and , and the inverse tangent must be suitably defined to take the correct quadrant of into account.
In terms of Cartesian coordinates,
The scale factors are
so the metric coefficientsare
The line element is
the area element
and the volume element
The Jacobian is
The radius vector is
so the unit vectors are
Derivatives of the unit vectors are
The gradient is
and its components are
(Misner et al. 1973, p. 213, who however use the notation convention ).
The Christoffel symbols of the second kind in the definition of Misner et al. (1973, p. 209) are given by
(Misner et al. 1973, p. 213, who however use the notation convention ). The Christoffel symbols of the second kind in the definition of Arfken (1985) are given by
(Walton 1967; Moon and Spencer 1988, p. 25a; both of whom however use the notation convention ).
The divergence is
or, in vector notation,
The covariant derivatives are given by
The commutation coefficients are givenby
so , where .
so , .
Time derivatives of the radius vector are
The speed is therefore given by
The acceleration is
Plugging these in gives
Time derivatives of the unitvectors are
The curl is
The Laplacian is
The vector Laplacian in spherical coordinatesis given by
To express partial derivatives with respect to Cartesian axes in terms of partial derivatives of the spherical coordinates,
Upon inversion, the result is
The Cartesian partial derivatives in sphericalcoordinates are therefore
(Gasiorowicz 1974, pp. 167-168; Arfken 1985, p. 108).
The Helmholtz differential equationis separable in spherical coordinates.