Physics 310

Notes on Coordinate Systems and Unit Vectors

A general system of coordinates uses a set of parameters to dene a vector. For example, x, y
and z are the parameters that dene a vector r in Cartesian coordinates:
r =x + y +

kz (1)
Similarly a vector in cylindrical polar coordinates is described in terms of the parameters r,
and z since a vector r can be written as r = r r + z

k. The dependence on is not obvious here,

but the unit vector r is actually a function of the polar angle, . If you want, you can make this
dependence explicit by writing
r = r r() +

kz (2)
Finally, a vector in spherical coordinates is described in terms of the parameters r, the polar
angle and the azimuthal angle as follows:
r = r r(, ) (3)
where the dependence of the unit vector r on the parameters and has been made explicit.
It can be very useful to express the unit vectors in these various coordinate systems in terms of
their components in a Cartesian coordinate system. For example, in cylindrical polar coordinates,
x = r cos
y = r sin (4)
z = z
while in spherical coordinates
x = r sin cos
y = r sin sin (5)
z = r cos .
Using these representations, we can construct the components of all unit vectors in these coordinate
systems and in this way dene explicitly the unit vectors r,

,

, etc.
If a vector, r depends on a parameters u, then a vector that points in the direction of
increasing u is dened by
e
u
=
r
u
. (6)
This vector is not necessarily normalized to have unit length, but from it we can always construct
the unit vector
e
u
=
e
u
|e
u
|
(7)
We will apply this denition to the Cartesian, cylindrical and spherical coordinate systems to
illustrate the construction of their unit vectors.
The case of Cartesian coordinates is almost trivial:
e
x
=
r
x
= (8)
e
y
=
r
y
= (9)
e
z
=
r
z
=

k. (10)
It also turns out that each of these vectors is already normalized to have unit length.
In the case of cylindrical polar coordinates, using Equations 2 and 4,
e
r
=
r
r
= r()
= cos + sin , (11)
e

=
r

= r
r

= r sin + r cos , (12)

e
z
=
r
z
= k (13)
The unit vectors r and

are the constructed using Equation 7 as follows:
r =
e
r

cos
2
+ sin
2

= e
r
(14)

=
e

sin
2
+ cos
2

=
e

r
(15)
so it turns out that e
r
was already normalized to unit length.
For the last example, in spherical coordinates, using Equations 3 and 5,
e
r
=
r
r
= r(, )
= sin cos + sin sin +

k cos , (16)
e

=
r

= r
r

= r sin sin + r sin cos , (17)

e

=
r

= r
r

= r cos cos + r cos sin

kr sin (18)
The unit vectors r,

and

are the constructed using Equation 7 as follows:
r =
e
r

sin
2
(sin
2
+ cos
2
) + cos
2

=
e
r

sin
2
+ cos
2

= e
r
(19)

=
e

r sin

sin
2
+ cos
2

=
e

r sin
(20)

=
e

sin
2
(sin
2
+ cos
2
) + cos
2

=
e

sin
2
+ cos
2

=
e

r
(21)
so it turns out that e
r
was already normalized to unit length. From Equation 20 you can see that
the direction of

becomes completely undened when = 0 or = .
We usually express time derivatives of the unit vectors in a particular coordinate system in
terms of the unit vectors themselves. Since all unit vectors in a Cartesian coordinate system are
constant, their time derivatives vanish, but in the case of polar and spherical coordinates they do
not.
In polar coordinates,
d r
dt
= ( sin + cos )
d
dt
=

(22)
d

dt
= ( cos sin )
d
dt
= r

(23)
In spherical coordinates,
d r
dt
=
d r
d
d
dt
+
d r
d
d
dt
= ( cos cos + cos sin

k sin )
d
dt
+ ( sin sin + sin cos )
d
dt
=

+

sin

(24)
d

dt
= r

sin

cos (25)
d

dt
= r

cos (26)