It is for this reason that when vector fields are thought of a force fields, gradient fields

are also called conservative fields.

Comment:
In many applications in the real world, conservation of energy does not have the simple
form of (2.35), because, for example, of energy losses due to friction – think of the space
shuttle re-entering the atmosphere. Dissipative, i.e. non-conservative forces have also to be
considered. It is nevertheless important to be able to find out whether a given force field is
conservative, and this is the problem we now consider.

2.4 Green’s Theorem

In this section we introduce Green’s theorem, and discuss a number of applications, including
how to spot conservative/gradient vector fields in R2 .

2.4.1 The theorem

We need some additional terminology related to curves.
Consider a curve C in Rn given by

x = g(t), a ≤ t ≤ b,

with g continuous.

i) C is a closed curve means that g(a) = g(b).

ii) C is a simple closed curve means that g(a) = g(b) and g is a one-to-one function on
the interval a ≤ t < b. (Note the strict inequality t < b.) In geometric terms a simple
closed curve has no self-intersections.

g(a) = g(b) g(a) = g(b)

Figure 2.7: A simple closed curve. A non-simple closed curve.

iii) In what follows we shall consider a bounded open subset D of R2 , whose boundary,
denoted by ∂D, is a simple closed curve. In this situation we assume that the curve
∂D is oriented counter-clockwise, so that if you walk around the boundary, the region
D is on your left.

48
 D
∂ D

Figure 2.8: A bounded open subset D and its boundary ∂D oriented counter-clockwise.

Theorem 2.3 (Green’s theorem):

Let D be a bounded subset of R2 whose boundary ∂D is a piecewise C 1 simple closed curve
oriented counter-clockwise. If F = (F1 , F2 ) is of class C 1 on D ∪ ∂D then
Z ZZ  
∂F2 ∂F1
F·dx = − dx dy. (2.36)
∂x ∂y
∂D D

Digression on iterated integrals:

If D is described by inequalities of the form
f (x) ≤ y ≤ g(x),
a ≤ x ≤ b,
ZZ
and H(x, y) is continuous on D ∪ ∂D, then the double integral H(x, y)dx dy can be
D
expressed as an iterated integral:

y
C2
(x,g(x))

D
C1

(x,f(x))
x
a x b

Figure 2.9: A region D and its boundary ∂D = C1 ∪ (−C2 ).

"Z #
ZZ Z b g(x)
H(x, y)dx dy = H(x, y)dy dx. (2.37)
x=a y=f (x)
D

49
Proof of Green’s theorem:
We give a proof subject to the assumption that D can be described by inequalities of the
form
f (x) ≤ y ≤ g(x), a ≤ x ≤ b, (2.38)
and
h(y) ≤ x ≤ k(y), c ≤ y ≤ d, (2.39)
where f, g, h and k are C 1 functions.
It is sufficient to prove two special cases of (2.36), namely
Z ZZ
∂F1
(F1 , 0) · dx = − dx dy, (2.40)
∂y
∂D D

and Z ZZ
∂F2
(0, F2 ) · dx = dx dy. (2.41)
∂x
∂D D

The sum of (2.40) and (2.41) gives (2.36).

We prove (2.40), using the inequalities (2.38). Consider the curves C1 , C2 (see Figure 2.9)
given by
x = (x, f (x)) and x = (x, g(x))
respectively, with a ≤ x ≤ b, i.e. we use x as parameter. The boundary ∂D is then the
union ∂D = C1 ∪ (−C2 ). By definition of the line integral,
Z Z Z !
R R
(F1 , 0) · dx = (F1 , 0) · dx − (F1 , 0) · dx since =−
−C2 C2
∂D C1 C2
Z b Z b
′
= {F1 (x, f (x)), 0) · (1, f (x)}dx − {F1 (x, g(x)), 0) · (1, g ′ (x)}dx
a a
Z b
= [F1 (x, f (x)) − F1 (x, g(x))] dx (2.42)
a

We now apply (2.37) to the right side of (2.40):

ZZ Z b "Z g(x) #
∂F1 ∂F1
− dx dy = − dy dx
∂y x=a y=f (x) ∂y
D (2.43)
Z b
=− [F1 (x, g(x)) − F1 (x, f (x))] dx,
a

by the FTCII. Equation (2.40) follows, on comparing (2.42) and (2.43). A similar argument
based on (2.39) yields (2.41) (do it!), which completes the proof. 

Example 2.4:
Verify Green’s theorem for the vector field F(x) = (xy, 2xy) and the triangluar region D
with vertices (0, 0), (1, 0) and (0, 1).

50
Solution:
a) The boundary of D taken counterclockwise is the piecewise C 1 curve
∂D = C1 ∪ C2 ∪ C3 .

Z Z Z Z
F · dx = F · dx + F·x+ F · dx. (2.44)
∂D C1 C2 C3
Observe that F = 0 on C1 (y = 0) and on C3 (x = 0), which implies that
Z Z
F · dx = 0 = F · dx. (2.45)
C1 C3

A vector function for C2 (x + y = 1) is

x = g(t) = (1 − t, t), 0 ≤ t ≤ 1.
It follows that
g′ (t) = (−1, 1) and F(g(t)) = t(1 − t)(1, 2),
after taking out a common factor. By definition of line integral,
Z Z 1
F · dx = F(g(t)) · g′ (t)dt
C2 0
Z 1
= t(1 − t)(1, 2) · (−1, 1)dt
0
Z 1
= t(1 − t)dt = · · · = 16 (2.46)
0

Substituting (2.45) and (2.46) in (2.44) gives

Z
F · dx = 16 .
∂D

51
 b) We calculate
∂F2 ∂F1
− = −x + 2y.
∂x ∂y
The right side of Green’s theorem is
ZZ   ZZ
∂F2 ∂F1
− dxdy = (−x + 2y)dxdy
∂x ∂y
D D
Z1 Z 1−x 
= (−x + 2y)dy dx
y=0
x=0
= · · · = 16 .

Green’s theorem is verified. 

The limits of integration for the triangular region D.

Aside: D is defined by the inequalities

0 ≤ y ≤ 1 − x,

0 ≤ x ≤ 1,
which give the limits of integration.

Green’s theorem can be used to express a given line integral as a double integral, or
conversely, to express a given double integral as a line integral, provided you can choose a
suitable vector field. Here is an example of the latter.

Example 2.5:
Use a line integral to calculate the area enclosed by the ellipse
x2 y 2
+ 2 = 1.
a2 b

Solution: The area A of a plane region D can be expressed as a double integral:

ZZ
A= (1)dx dy.
D

∂F2 ∂F1
The vector field F = (0, x) satisfies ∂x
− ∂y
= 1, so that Green’s theorem (2.36) gives
Z
A= (0, x) · dx,
∂D

where ∂D is the ellipse (a simple closed curve) oriented counter-clockwise. Using the standard
parametrization,
x = (a cos t, b sin t), 0 ≤ t ≤ 2π,

52