Chapter 3

Vector Calculus

3.1 Scalar and vector fields

We introduce scalar and vector fields. A scalar field on Rn is the same as a real-valued function
on Rn , while a vector field on Rn is the same as a Rn valued function on Rn . These concepts may
then look like an unnecessary renaming of known concepts, however the perspective is new. This
becomes more apparent when we consider vector fields on spaces such as the circle, the sphere,
and so on.
Definition 3.1.1. Let D ⊆ Rn . A function f : D → R is called a scalar field on D. In other
words, a scalar field on D is the same as a real-valued function on D.
We have seen many examples of real-valued functions, so each one of them yields an example
of a scalar field.
Definition 3.1.2. Informally: A vector field on a space X is a choice of a tangent vector at
each point of X. The following is a way to visualize a vector field.

We have shown the tangent vectors only at three points, but one has to imagine a tangent vector
at each point of X. For X = Rn , a vector field on X is the same as a function F : Rn → Rn . We
focus on the cases n = 1, n = 2, n = 3 below.
Example 3.1.1. 1. (Vector fields on the line). For X = R, a vector field on X is the same
as a function F : R → R. For example, let F (x) = 1. Then the vector at each point on the
line points one unit to the right as illustrated below.
→−→−→−→−→−→
Now let F (x) = x. Can you imagine this vector field? It is zero at the origin. At positive
real numbers, vectors point to the right, while at negative real numbers, vectors point to the
left. Their magnitudes increase linearly as we move away from the origin on either side.

71
Chapter 3. Vector Calculus

2. (Vector fields on the plane). For X = R2 , a vector field on X is the same as a function
F : R2 → R2 . Many examples along with illustrations are shown below.
(i) Constant vector fields. Let F (x, y) = (2, 1).

This is an example of a constant vector field.

(ii) Horizontal vector fields. Let F (x, y) = (x, 0). See left picture below. Vectors on y-axis
are 0 . For positive values of x, vectors point to the right. For negative values of x, vectors
point to the left. Lengths of vectors increase as we move to the right or to the left.

Let F (x, y) = (1, 0). This is a constant horizontal vector field where all vectors point one
unit to the right. See right picture above.
(iii) Radial vector fields. Let F (x, y) = (x, y). See left picture below. Vectors point radially
outward. Their lengths increase as we move outward.

 
Let F (x, y) = √ x
, √ y
on R2 \{(0, 0)}. This is a variant of the previous example
x2 +y 2 x2 +y 2
where vectors point radially outward and have length 1. See right picture above. We may
also consider F (x, y) = (−x, −y), where vectors point radially inward, and its normalized
version on R2 \{(0, 0)} where vectors also have length 1.
(iv) Rotational vector fields. Let F (x, y) = (−y, x). See left picture below. Vectors are
orthogonal to the radial direction, and point in the anticlockwise direction. Their lengths
increase as we move outward.

72
 Chapter 3. Vector Calculus

 
Let F (x, y) = √ −y ,√ x
on R2 \{(0, 0)}. This is a variant of the previous example
x2 +y 2 x2 +y 2
where vectors in addition have length 1 . See right picture above.

3. (Vector fields in space). For X = R3 , a vector field on X is the same as a function

F : R3 → R3 . A vector field on R2 yields a vector field on R3 by setting the z-coordinate to
zero. We can visualize it by translating the vector field in the xy-plane along the z-axis. For
instance, F (x, y, z) = (y, −x, 0).

4. (Vector fields on a curve). Consider the unit circle S 1 = {(x, y) ∈ R2 : x2 + y 2 = 1} A

vector field on S 1 is shown below on the left. For simplicity, only four tangent vectors have
been drawn.

The picture on the right cannot be part of a vector field on S 1 since there are vectors (two
of them are shown) which are not tangential to the circle. A way to visualize a vector field
on a general curve is shown below.

Suppose a particle is moving along a curve X. Then the velocity vector of the particle yields
a vector field on X.

5. (Vector fields on a surface). Consider a surface X in R3 . For example, X can be the

unit sphere S 2 , or X can be the graph z = f (x, y) of a function f of two variables. For each
point p on X, there is a plane tangent to X at p. To specify a vector field on X, we choose
a vector from this tangent plane at p, and do this for each point p on X.

73
Chapter 3. Vector Calculus

3.2 Gradient, curl, divergence

We now introduce three fundamental operations related to scalar and vector fields . These are
called the gradient, curl, divergence. In this section, D denotes an open subset of Rn .

Definition 3.2.1. Let f be a scalar field on D. The gradient vector field of f is the vector
field on D defined by
!
∂f ∂f ∂f
grad f := ∇f = , ,... = (fx1 , fx2 , . . . fxn ) . (3.1)
∂x1 ∂x2 ∂xn

In this case, we say F = grad f is a conservative vector field with potential function f .

Example 3.2.1. (1) Let D = R3 . Consider the vector field F (x, y, z) = (yz, zx, xy) on D. It is
the gradient vector field of f (x, y, z) = xyz, that is, F = grad f .

Definition 3.2.2. Let F = (P, Q, R) be a vector field on D ⊂ R3 . The curl vector field of F is
the vector field on D defined by
!
∂R ∂Q ∂P ∂R ∂Q ∂P
curl F := ∇ × F := − , − , − . (3.2)
∂y ∂z ∂z ∂x ∂x ∂y

The notation ∇ × F is to be interpreted as

 
i j k
∇ × F = ∇ × (P, Q, R) = det 
 ∂ ∂ ∂
.

∂x ∂y ∂z
P Q R

Note that curl is defined only in 3-dimensions.

Example 3.2.2. Let F = ω(−y, x, 0), where we interpret w as angular speed. Then curl F =
(0, 0, 2ω). This is a constant vector field pointing along the positive z-axis.

Definition 3.2.3. Let F = (f1 , f2 , . . . fn ) be a vector field on D. The divergence field of F is the
scalar field on D defined by
∂f1 ∂f2 ∂fn
div F := ∇ · F = + + ··· + . (3.3)
∂x1 ∂x2 ∂xn

Example 3.2.3. div(x, y, z) = 3 and div (xy 2 , z sin y, xey ) = y 2 + z cos y.

Remark 3.2.1. Gradient, curl, divergence can be assembled into a sequence of maps

scalar vector vector scalar

( ) ( ) ( ) ( )
grad curl div
−−−→ −−−→ −−→ .
fields fields fields fields

Lemma 3.2.1. Suppose f is a scalar field and F is a vector field on D ⊂ R3 . Then we have

curl(grad f ) = ∇ × (∇f ) = 0,
div(curl F ) = ∇ · (∇ × F ) = 0.

74
 Chapter 3. Vector Calculus

Proof. We compute:  
i j k
curl(grad(f )) = det  = 0.
 ∂ ∂ ∂ 
∂x ∂y ∂z 
fx fy fz
Similarly,
div(curl F ) = (Ry − Qz )x + (Pz − Rx )y + (Qx − Py )z
= Ryx − Qzx + Pzy − Rxy + Qxz − Pyz = 0.
This proves the result.
Question 3.2.1. When is a vector field a gradient vector field? Given a vector field F on D, we
would like to know if F is the gradient vector field of some scalar field f on D ?
We will see a necessary condition for this and later we will see a sufficient condition.
Lemma 3.2.2. (Necessary condition for a gradient vector field). Suppose F = grad f ,
that is, F = (P, Q, R) and P = fx , Q = fy , R = fz . Then curl F = 0, that is, Qx = Py , Ry =
Qz , Pz = Rx .
Proof. The proof follows directly from the Lemma 3.2.1.
Example 3.2.4. (1) Consider the inward radial vector field F (x, y) = (−x, −y). It is the gradient
vector field of f (x, y) = − 21 (x2 + y 2 ), that is, F = grad f .
(2) Consider the rotational vector field F (x, y) = (−y, x). This is not a gradient vector field since
F = (P, Q) and Qx = 1 = −Py which violates the necessary condition above. Now consider
!
−y x
F (x, y) = , 2
x + y x + y2
2 2

on R2 \{(0, 0)}. Now

x2 − y 2
F = (P, Q) and Qx = − 2 = Py .
(x + y 2 )2
So the necessary condition is satisfied, yet F is not a gradient vector field.
We mention that the converse of Lemma 3.2.2 holds which we will see later.

3.3 Line integrals and FTC

Now we look at line integrals of scalar fields and of vector fields along a parametrized curve.
Definition 3.3.1. A path or a parametrized curve in Rn is a map

γ : [a, b] → Rn , t 7→ (x1 (t), . . . , xn (t)) .

We say γ(a) is the initial point and γ(b) is the final point of the path γ. We say γ is closed
if γ(a) = γ(b). Define
γ(t + h) − γ(t)
γ ′ (t) := lim = (x′1 (t), . . . , x′n (t)) .
h→0 h
It is the tangent vector to γ at t. We say γ is regular if γ ′ (t) ̸= 0 for all a ≤ t ≤ b

75
Chapter 3. Vector Calculus

The picture below shows a parametrized curve in R2 , with the tangent vector marked at one
point.

In physics terms, one may think of a path as a particle moving along a curve in Rn with t as the
time variable. The particle is at the initial point at t = a, at the final point at t = b, and in general
at point γ(t) at time t. The vector γ ′ (t) is then the velocity of the particle at time t.

Example 3.3.1. (1) (Graph). Let f : [a, b] → R be a function. Then the graph of f is a
parametrized curve γ : [a, b] → R2 , γ(t) = (t, f (t)), and γ ′ (t) = (1, f ′ (t)).
2
(2) Consider the graph of the function y = x 3 . This curve can be parametrized by γ(t) = (t3 , t2 )
for t ∈ [−1, 1]. Then γ ′ (t) = (3t3 , 2t). Observe that γ ′ (0) = (0, 0). Thus, γ is not regular at t = 0.

Definition 3.3.2. For a parametrized curve γ, define the length of γ to be

Z b Z bq
ℓ(γ) := ∥γ (t)∥ dt =
′
x′1 (t)2 + · · · + x′m (t)2 dt.
a a

Example 3.3.2. (Helix). Consider the curve γ(t) = (cos t, sin t, t) for a ≤ t ≤ b. Then
Z bq √
ℓ(γ) = (− sin t)2 + (cos t)2 + 1dt = 2(b − a).
a

3.3.1 Line integral of a scalar field

Let C = γ([a, b]). Let f : C → R be a scalar field. Define the line integral of f along a path γ by
Z Z b
f |ds| := f (γ(t)) ∥γ ′ (t)∥ dt. (3.4)
γ a

If f ≡ 1, the the above formula coincides with the arc length formula of γ.

Remark 3.3.1 (Differential notation). Put

q
|ds| = ∥γ ′ (t)∥ dt = dx21 + · · · + dx2m .

Then Z Z q Z b
f |ds| = f dx21 + ··· + dx2m = f (γ(t)) ∥γ ′ (t)∥ dt.
γ γ a

Definition 3.3.3. Let γ(t) be a parametrized curve, where t ∈ [a, b]. A reparametrization is
⃗
another parametrization R(s) = γ(t(s)), where t = t(s) is a smooth, one-to-one, and monotonic
function.

76
 Chapter 3. Vector Calculus

Definition 3.3.4 (Arc length parametrization). For t ∈ [a, b], define

Z t
u(t) := ∥γ ′ (t)∥ dt. (3.5)
a

This yields a bijective map from [a, b] to [0, ℓ(γ)]. Composing its inverse with γ yields
γ
γ̃ : [0, ℓ(γ)] → [a, b] →
− Rn .
This parametrization of C is known as its arc length parametrization, and u is called the arc
length parameter.
Example 3.3.3. For the circle of radius a centered at the origin, an arc length parametrization is
u u
    
γ̃ : [0, 2πa] → R2 , u 7→ a cos , a sin .
a a
Starting with the usual parametrization (a cos t, a sin t), formula (3.5) yields u = at which leads to
the above map.

3.3.2 Line integral of a vector field

Define the line integral of a vector field F on Rn along a path γ in Rn by
Z Z b
F · ds := F (γ(t)) · γ ′ (t)dt. (3.6)
γ a

The vector field F is defined everywhere, but for the above definition, only its values on points of
the curve matter. In physics terms, one may think of the above formula as the work done by the
force field F along the curve γ.
Example 3.3.4. Consider the radial vector field F (x, y, z) = (x, y, z) on R3 and the path
γ : [0, 2π] → R3 , γ(t) = (cos t, sin t, t).
The latter is a parametrized helix. Then
Z Z 2π Z 2π
F · ds = (cos t, sin t, t) · (− sin t, cos t, 1)dt = tdt = 2π 2 .
γ 0 0

Remark 3.3.2 (Differential notation). Let F = (P, Q, R) and γ(t) = (x(t), y(t), z(t)). Put
ds = γ ′ (t)dt = (x′ (t)dt, y ′ (t)dt, z ′ (t)dt) = (dx, dy, dz).
Then
F · ds = P x′ (t)dt + Qy ′ (t)dt + Rz ′ (t)dt = P dx + Qdy + Rdz
and Z Z Z b
F · ds = P dx + Qdy + Rdz = P x′ (t)dt + Qy ′ (t)dt + Rz ′ (t)dt.
γ γ a

Example 3.3.5. Consider the vector field F (x, y, z) = (cos z, ex , ey ) on R3 and the path γ : [0, 2] →
R3 , γ(t) = (1, t, et ) Then
Z Z
F · ds = cos zdx + ex dy + ey dz
γ γ
Z 2 h       i
= cos et (0) + e1 (1) + et et dt
0
1 4 
= 2e + e −1 .
2
77
Chapter 3. Vector Calculus

Line integral of a gradient vector field:

Recall The Fundamental Theorem of Calculus (FTC) Part I and Part II:
FTC Part I: Let f be Riemann integrable on [a, b]. For x ∈ [a, b], define
Z x
F (x) := f (t)dt.
a

Then F is continuous on [a, b]. Moreover, if f is continuous at c ∈ [a, b], then F is differentiable
at c, and F ′ (c) = f (c).
FTC Part II: Let f be Riemann integrable on [a, b]. Let F be continuous on [a, b], differentiable
on (a, b), and F ′ = f on (a, b). Then
Z b
f (x)dx = F (b) − F (a).
a

The following is a generalization of FTC, Part II. We may refer to it as FTC on a parametrized
curve; the earlier FTC is on a line segment (which is the simplest example of a parametrized curve
with γ : [a, b] → R defined by γ(x) = x).
Theorem 3.3.1. Let F = grad f be a gradient vector field on an open subset D of Rn . Let
γ : [a, b] → Rn be a path which lies in D. Then
Z
F · ds = f (γ(b)) − f (γ(a)). (3.7)
γ

Proof. Let γ(t) = (x1 (t), x2 (t), . . . , xn (t)). Then

Z Z b
F · ds = ∇f (γ(t)) · γ ′ (t)dt,
γ a
Z b
= [fx1 (γ(t))x′1 (t) + fx2 (γ(t))x′2 (t) + · · · + fxn (γ(t))x′n (t)] dt,
a
Z b
= (f ◦ γ)′ (t)dt,
a
= f (γ(b)) − f (γ(a)).
The third equality is by the chain rule. In matrix form,
x′1 (t)
 

h i x′2 (t) 
(f ◦ γ) (t) =
′
fx1 (γ(t)) fx2 (γ(t)) . . . fxn (γ(t)) ..
 
 .
.
 
 
x′n (t).
The last equality is by FTC, Part II on an interval [a, b].
Example 3.3.6. Consider the vector field on R3 given by F (x, y, z) = (yz, zx, xy). It is the
gradient vector field of f (x, y, z) = xyz. Let
 
γ : [0, π/4] → R3 , γ(t) = cos4 t, sin4 t, tan4 t .

By formula (3.7),
1 1 1
Z  
F · ds = f (γ(π/4)) − f (γ(0)) = f , , 1 − f (1, 0, 0) = .
γ 4 4 16

78
 Chapter 3. Vector Calculus

Definition 3.3.5 (Path independence of line integrals). Let F be a gradient vector field on
an open subset D of Rn . We say line integrals of F are path independent in D if
Z Z
F · ds = F · ds
γ1 γ2

for any paths γ1 and γ2 lying in D withR the same initial point and same final point. The above
condition is equivalent to requiring that γ F · ds = 0 for any closed path γ lying in D.

Proposition 3.3.1. Let D be an open subset of Rn . If F is a gradient vector field then line
integrals of F are path independent in D. Moreover, if D is path connected and suppose the line
integrals of F are path independent in D, then F is a gradient vector field on D.

Proof. The first part follows from formula (3.7). For the second part we proceed as follows: We
illustrate with n = 3. Let F = (P, Q, R). Fix a point (a0 , b0 , c0 ) ∈ D. Put
Z (x,y,z) Z
F · ds := F · ds
(a0 ,b0 ,c0 ) γ

where γ is any path in D from (a0 , b0 , c0 ) to (x, y, z). This is well-defined by the hypothesis that
line integrals of F only depend on endpoints of γ. Define
Z (x,y,z)
f (x, y, z) := F · ds.
(a0 ,b0 ,c0 )

We claim that f is the required potential function. That is, ∇f = F , or equivalently, fx = P, fy =

Q, fz = R. Let us first deal with the claim fx = P . We calculate:
R (a+h,b,c)
f (a + h, b, c) − f (a, b, c) (a,b,c) F · ds
fx (a, b, c) = lim = lim ,
h→0 h h→0 h
P (a + x, b, c)dx
Rh
= lim 0 = P (a, b, c).
h→0 h
In the third step, we took the straight line path in the x-direction, namely,

γ : [0, h] → R3 , γ(x) = (a + x, b, c),

and so γ ′ (x) = (1, 0, 0). The last step can be deduced from FTC, Part I. This shows fx = P .
Analogously, we can show fy = Q, fz = R.
 
Example 3.3.7. Consider the vector field F (x, y) = −y
, x
x2 +y 2 x2 +y 2
on R2 \{(0, 0)}, and the closed
path γ : [0, 2π] → R2 , γ(t) = (cos t, sin t). Then
Z Z 2π Z 2π
F · ds = (− sin t, cos t) · (− sin t, cos t)dt = dt = 2π.
γ 0 0

Hence F is not a gradient vector field (since its line integral along the above closed path γ is not
zero).

79
Chapter 3. Vector Calculus

3.3.3 Geometric curve

A geometric curve is the image of a parametrized curve. An orientation of a geometric curve
C is an assignment of a unit tangent vector, that is, a specification of a direction at each point of
C in a continuous manner.
This can be done in several ways.
(1) If C is not closed, then specify the ’start’ and ’finish’ points.
(2) If C is closed and is the boundary of a planar region S, then specify whether S lies to the left
(or to the right) as C is traversed (with head in upright position).
′ (t)
(3) If γ : [a, b] → R2 is a ’nice’ parametrized curve, then for each t, ∥γγ ′ (t)∥ is a unit tangent vector
to C at γ(t), and we obtain an orientation of C.
An oriented geometric curve is a geometric curve with a specified orientation. If C is
′ (t)
the image of a parametrized curve γ, then we say γ is orientation preserving if ∥γγ ′ (t)∥ equals the
prescribed unit tangent vector. In this case, we define
Z Z
F · ds = F · ds.
C γ

If this is not the case, that is, γ is orientation reversing, then −γ is orientation preserving, and we
define Z Z
F · ds = − F · ds. (3.8)
−γ γ

3.4 Green’s theorem

We now generalize FTC, Part II to two dimensions, where the two-dimensional region is a subset of
R2 . This result relates an integral on a two-dimensional region D to a line integral on its boundary
∂D. All curves in this section are in R2 .

Definition 3.4.1 (Orienting the boundary curve). Let D be a bounded subset of R2 such
that ∂D consists of a finite number of simple closed non-intersecting piecewise smooth geometric
curves. Orient each such curve so that as one travels along that curve, the set D lies to the left
(with head in upright position). Then we say ∂D is positively oriented. See illustrations below.

When viewed from above, the outer boundary of D is traversed anticlockwise, and each of the inner
boundary curves is traversed clockwise.

Theorem 3.4.1 (Green’s theorem). Let D ⊆ R2 be such that ∂D is positively oriented. If P

and Q are scalar fields defined on an open set containing D, then
Z ZZ
P dx + Qdy = (Qx − Py ) d(x, y). (3.9)
∂D D

80
 Chapter 3. Vector Calculus

Proof. We break the proof in three steps.

(i) D is a rectangle. Let D = [a, b] × [c, d]. Let γ1 , γ2 , γ3 , γ4 be oriented line segments corre-
sponding to the four sides of the rectangle, as shown below.

Using FTC, Part II,

ZZ Z dZ b Z d Z Z
Qx d(x, y) = Qx d(x, y), = [Q(b, y) − Q(a, y)]dy, = Qdy + Qdy.
D c a c γ2 γ4

Similarly,
ZZ Z bZ d Z b Z Z
−Py d(x, y) = −Py d(x, y), = [−P (x, d) + P (x, c)]dx, = P dx + P dx.
D a c a γ3 γ1

Adding the two proves formula (3.9) when D is a rectangle.

(ii) D is a "union" of rectangles. Now suppose D is as shown below.

Overlapping line segments have opposite orientations, so line integrals over them cancel by (3.8).
Thus, formula (3.9) holds when D is formed by adjoining rectangles.
(iii) D is a general region. Draw a grid inside D as shown below.

Let D′ be the largest region that fits inside D which is of the form in item (ii) above and is made
out of the sides of the grid. By item (ii), formula (3.9) holds for D′ . Finally, we approach formula
(3.9) for D by considering finer and finer grids.

81
Chapter 3. Vector Calculus

Example 3.4.1. We verify Green’s theorem in the plane C (xy + y 2 ) dx + x2 dy where C is the
H

closed curve of the region bounded by y = x and y = x2 . Along y = x2 , from (0, 0) to (1, 1) the
line integral equals
Z 1 Z 1
19
(xx + x ) dx + x (2x dx) =
2 4 2
(3x3 + x4 ) dx = .
0 0 20
Along y = x from (1, 1) to (0, 0) the line integral equals
Z 0 Z 0
(xx + x ) dx + x (dx) =
2 2
3x3 dx = −1.
1 1

Then the required line integral is 19

20
− 1 = − 20
1
. Now, by Green’s Theorem:
!
I ZZ
∂Ψ ∂Φ
(Φ dx + Ψ dy) = − dx dy,
C R ∂x ∂y
∂x2 ∂(xy + y 2 )
ZZ ! ZZ ZZ
− dx dy = (2x − (x + 2y)) dx dy = (x − 2y) dx dy,
R ∂x ∂y R R
Z 1 Z x Z 1h ix Z 1
= dx (x − 2y) dy = xy − y 2 2
dx = (x2 − x4 ) dx,
x=0 y=x2 0 y=x 0
4 1
1 1 1
" #
Z 1 5
x x
= (x4 − x3 ) dx = − = − =− .
0 5 4 0
5 4 20

Hence, both line and double integrals match, verifying Green’s Theorem.

3.4.1 Area calculation.

Corollary 3.4.1. Let C be a simple closed geometric curve oriented anticlockwise that encloses a
region D. Then
1Z Z Z
Area(D) = xdy − ydx = xdy = − ydx. (3.10)
2 C C C

Proof. Put P (x, y) = −y

2
and Q(x, y) = x2 . Then Qx − Py = 1
2
+ 1
2
= 1. By formula (2.5),
ZZ ZZ Z
1Z
Area(D) = d(x, y) = (Qx − Py ) d(x, y) = P dx + Qdy = xdy − ydx.
D D ∂D=C 2 C
The third equality is by (3.9).

Remark 3.4.1. If in (3.10) we replace xdy − ydx by xdy + ydx, then we get a line integral of a
gradient vector field over a closed curve which is zero by formula (3.7).

Corollary 3.4.2. For a simple closed curve C parametrized by (x(t), y(t)) for a ≤ t ≤ b which is
traversed anticlockwise as t goes from a to b,

1Z b
Area(D) = x(t)y ′ (t) − y(t)x′ (t)dt, (3.11)
2 a
1Z b
" #
x(t) y(t)
= det dt.
2 a x′ (t) y ′ (t)

82
 Chapter 3. Vector Calculus

2 2
Example 3.4.2. (1) (Ellipse). Let C be the ellipse defined by xa2 + yb2 = 1. Parametrize C by
x = a cos t and y = b sin t for 0 ≤ t ≤ 2π. Then, by formula (3.11), area of the elliptical region
enclosed by C is
1 Z 2π
(a cos t)(b cos t) − (b sin t)(−a sin t)dt = πab.
2 0
(2) (Hypocycloid). Let C be the hypocycloid given by x = a cos3 t and y = a sin3 t for 0 ≤ t ≤ 2π.
Then " #
x(t) y(t)
det = 3a2 cos2 t sin2 t.
x′ (t) y ′ (t)
Thus, by formula (3.11), area enclosed by C is
1 Z 2π 2 3a2 Z 2π sin 2t 2 3a2 Z 2π 1 − cos 4t 3πa2
 
3a cos2 t sin2 tdt = dt = dt = .
2 0 2 0 2 8 0 2 8

(a) Ellipse
(b) Hypocycloid

Figure 3.1

3.5 Surface integrals

Now we look at surface integrals of scalar fields and of vector fields across a parametrized surface.
Definition 3.5.1 (Parametrized Surface). A parametrized surface in R3 is a map
Φ : D → R3 , (u, v) 7→ (x(u, v), y(u, v), z(u, v)) (3.12)
where D ⊆ R2 . Define
Φu := (xu , yu , zu ) and Φv := (xv , yv , zv ) (3.13)
where xu is the partial derivative of x wrt u, and so on.
Definition 3.5.2 (Fundamental vector product). The fundamental vector product of the
parametrization Φ is defined by
 
i j k " # " # " #!
yu zu zu xu xu yu
(Φu × Φv ) (u, v) = det  xu yu zu  = det , det , det ,
 
yv zv zv xv xv yv
xv yv zv
!
∂(y, z) ∂(z, x) ∂(x, y)
=: , , . (3.14)
∂(u, v) ∂(u, v) ∂(u, v)
This is a function of (u, v).

83
Chapter 3. Vector Calculus

Example 3.5.1. (1) (Cylinder). Let D = [0, 2π] × [0, h]. Fix a > 0. Define

Φ : D → R3 , Φ(θ, v) := (a cos θ, a sin θ, v), (3.15)

Now
Φθ = (−a sin θ, a cos θ, 0) and Φv = (0, 0, 1)
and  
i j k
(Φθ × Φv ) (θ, v) = det  −a sin θ a cos θ 0 

 = (a cos θ, a sin θ, 0).
0 0 1
(2) (Sphere). Let D = [0, π] × [0, 2π]. Fix a > 0. Define

Φ : D → R3 , Φ(φ, θ) := (a sin φ cos θ, a sin φ sin θ, a cos φ). (3.16)

Now
Φφ = (a cos φ cos θ, a cos φ sin θ, −a sin φ), Φθ = (−a sin φ sin θ, a sin φ cos θ, 0)
and
 
i j k
(Φφ × Φθ ) (φ, θ) = det  a cos φ cos θ a cos φ sin θ −a sin φ 

 = a sin φΦ(φ, θ).
−a sin φ sin θ a sin φ cos θ 0

(3) (Graph). Let f : D → R be a function of two variables. Then the graph of f is a parametrized
surface
Φ : D → R3 , Φ(u, v) = (u, v, f (u, v)). (3.17)
Now
Φu = (1, 0, fu ) and Φv = (0, 1, fv )
and  
i j k
(Φu × Φv ) (u, v) = det  1 0 fu  = (−fu , −fv , 1) .
 

0 1 fv

3.5.1 Area of a Parametrized Surface

Recall: For vectors u, v ∈ R3 , area of the parallelogram spanned by u and v is given by ∥u × v∥.
Let Φ : D → R3 be a parametrized surface. Define
ZZ
Area(Φ) := ∥(Φu × Φv ) (u, v)∥ d(u, v), (3.18)
D

where Φu × Φv is the fundamental vector product in (3.14).

Example 3.5.2. (1) (Cylinder). Consider the cylinder parametrized by (3.15). By formula (3.18),
ZZ ZZ
Area(Φ) = ∥(a cos θ, a sin θ, 0)∥d(θ, v) = ad(θ, v) = 2πah.
[0,2π]×[0,h] [0,2π]×[0,h]

This is the familiar formula for the surface area of cylinder of radius a and height h.

84
 Chapter 3. Vector Calculus

(2)(Sphere). Consider the sphere parametrized by (3.16). By formula (3.18),

ZZ Z π
Area(Φ) = ∥a sin φΦ(φ, θ)∥d(φ, θ) = 2πa2 sin φdφ = 4πa2 .
[0,π]×[0,2π] 0

This is the familiar formula for the surface area of sphere of radius a.
(3) (Graph). Consider the surface parametrized by (3.17). By formula (3.18),
ZZ ZZ q
Area(Φ) = ∥(−fu , −fv , 1)∥ d(u, v) = 1 + fu2 + fv2 d(u, v).
D D

As an example, consider the paraboloid parametrized by f (u, v) = u2 + v 2 for (u, v) ∈ D, where

D is the disc of radius a around (0, 0). Then
ZZ √ Z 2π Z a√
π 
 3/2 
Area(Φ) = 1 + 4u2 + 4v 2 d(u, v) = dθ 1 + 4r2 rdr = 1 + 4a2 −1 .
D 0 0 6

3.5.2 Surface integral of a scalar field

Let S = Φ(D). Let f : S → R be a scalar field. Define the surface integral of f across Φ by
ZZ ZZ
f |dS| := f (Φ(u, v)) ∥(Φu × Φv ) (u, v)∥ d(u, v), (3.19)
Φ D

where Φu × Φv is the fundamental vector product in (3.14). If f ≡ 1, then formula (3.19) coincides
with (3.18) and yields the area of the surface parametrized by Φ.
h i
Example 3.5.3. (1) Consider the hemisphere parametrized by (3.16) on D = 0, π2 × [0, 2π]. Let
f (x, y, z) = z. Then
ZZ ZZ Z π
2
f |dS| = a cos φ∥a sin φΦ(φ, θ)∥d(φ, θ) = 2πa 3
cos φ sin φdφ = πa3 .
Φ [0, π2 ]×[0,2π] 0

(2) Let f (x, y, z) = xy + z. Let S be the portion of the cylinder y 2 + z 2 = 9 in the first octant
with 0 ≤ y ≤ 2 and 0 ≤ x ≤ 4. We want to evaluate (3.19).
√
Observe: S is the graph of the function z = g(x, y) = 9 − y 2 with 0 ≤ y ≤ 2 and 0 ≤ x ≤ 4.
Thus,
Z 2Z 4
3 Z 2Z 4
ZZ q q 
f |dS| = (xy + g(x, y)) 1 + gx2 + gy2 dxdy = dxdy xy + 9 − y2 √
Φ 0 0 0 0 9 − y2
Z 2Z 4
3xy Z 2Z 4 √ √
= √ dxdy + 3dxdy = 24(3 − 5) + 24 = 24(4 − 5).
0 0 9 − y2 0 0

h Φ
Lemma 3.5.1. Let E → → R3 , with J(h) ̸= 0, where J is the Jacobian of h. Then
− D−
ZZ ZZ
f |dS| = f |dS|. (3.20)
Φ◦h Φ

85
Chapter 3. Vector Calculus

3.5.3 Surface integral of a vector field

Define the surface integral of a vector field F on R3 across a parametrized surface Φ in R3 by
ZZ ZZ
F · dS := F (Φ(u, v)) · (Φu × Φv ) (u, v) d(u, v) (3.21)
Φ D

where Φu × Φv is the fundamental vector product in (3.14).

Remark 3.5.1 (Differential notation). Let F = (P, Q, R). Put

!
∂(y, z) ∂(z, x) ∂(x, y)
dS := (Φu × Φv ) (u, v)d(u, v) = d(u, v), d(u, v), d(u, v)
∂(u, v) ∂(u, v) ∂(u, v)
= (dy ∧ dz, dz ∧ dx, dx ∧ dy).

Then
∂(y, z) ∂(z, x) ∂(x, y)
F · dS = P d(u, v) + Q d(u, v) + R d(u, v),
∂(u, v) ∂(u, v) ∂(u, v)
= P dy ∧ dz + Qdz ∧ dx + Rdx ∧ dy.
and ZZ ZZ
F · dS = P dy ∧ dz + Qdz ∧ dx + Rdx ∧ dy,
Φ D
ZZ
∂(y, z) ∂(z, x) ∂(x, y)
= P d(u, v) + Q d(u, v) + R d(u, v).
D ∂(u, v) ∂(u, v) ∂(u, v)

Example 3.5.4. (1) (Cylinder). Consider the radial vector field F (x, y, z) = (x, y, z) on R3 and
the cylinder parametrized by (3.15). Then
ZZ ZZ Z 2π Z h
F · dS = (a cos θ, a sin θ, v) · (a cos θ, a sin θ, 0)d(θ, v) = a 2
dθ dv = 2πa2 h.
Φ [0,2π]×[0,h] 0 0

(2) (Sphere). Consider the radial vector field F (x, y, z) = (x, y, z) on R3 and the sphere parametrized
by (3.16). Then
ZZ ZZ Z π Z 2π
F · dS = Φ(φ, θ) · Φ(φ, θ)a sin φd(φ, θ) = a3 sin φd(φ, θ) = 4πa3 .
Φ [0,π]×[0,2π] 0 0

(3) (Graph). Consider the surface parametrized by (3.17). Let F = (P, Q, R). Then
ZZ ZZ ZZ
F · dS = (P, Q, R) · (−fu , −fv , 1) d(u, v) = (−P fu − Qfv + R) d(u, v).
Φ D D

Consider the paraboloid parametrized by f (u, v) = u2 + v 2 for (u, v) ∈ D, where D = [0, 1] × [0, 1]
is the unit square. Then
ZZ ZZ   
F · dS = −u(2u) − v(2v) + u2 + v 2 d(u, v),
Φ [0,1]×[0,1]
ZZ   2
=− u2 + v 2 d(u, v) = − .
[0,1]×[0,1] 3

Note: F · dS can be negative in general.

RR
Φ

86
 Chapter 3. Vector Calculus

Definition 3.5.3. Let a surface be parametrized by a vector-valued function:

Φ(u, v) = (x(u, v), y(u, v), z(u, v)), (u, v) ∈ D ⊂ R2 .

A reparametrization of the surface is a new parametrization given by:

h(s, t) = Φ(u(s, t), v(s, t))

where u = u(s, t), v = v(s, t) defines a smooth, one-to-one, map with J(h) ̸= 0 from a new subset
E ⊂ R2 to D.
h Φ
Lemma 3.5.2. Let E → → R3 , with J(h) ̸= 0. Then
− D−
ZZ ZZ
F · dS = ± F · dS. (3.22)
Φ◦h Φ

h Φ
Special case. For D ⊆ R2 , let E ⊆ R2 be defined by {(p, q) : (q, p) ∈ D}. Consider E →
− D− → R3 ,
with h(p, q) = (q, p). We denote Φ ◦ h by −Φ, and call it the negative of Φ. Since J(h) = −1 < 0,
ZZ ZZ
F · dS = − F · dS. (3.23)
−Φ Φ

Lemma 3.5.3 (Relating dS and |dS|). Suppose (Φu × Φv ) (u, v) ̸= 0. That is, the normal
vector to the surface at (u, v) under the parametrization Φ is nonzero. Put

(Φu × Φv ) (u, v)
n(Φ(u, v)) := (3.24)
∥(Φu × Φv ) (u, v)∥

the unit normal vector to the surface Φ at (u, v). Then

ZZ ZZ
F · dS = F · n|dS| (3.25)
Φ Φ

with lhs defined by (3.21) and rhs by (3.19).

Proof. This is a consequences of our definitions:
ZZ ZZ
F · dS = F (Φ(u, v)) · (Φu × Φv ) (u, v)d(u, v),
Φ ZZD
= F (Φ(u, v)) · n(Φ(u, v)) ∥(Φu × Φv ) (u, v)∥ d(u, v),
ZZD
= F · n|dS|.
Φ

In the second step, we used definition (3.24).

3.5.4 Geometric surface

A geometric surface in R3 is the image of a parametrized surface. An orientation of a geometric
surface S is an assignment of a unit normal vector at each point of S in a continuous manner.
This can be done in several ways.
(1) If S is the graph of a function of two variables, then specify whether the unit normal points
upward with positive z-component (or downward with negative z-component).

87
Chapter 3. Vector Calculus

(2) If S is a bounded surface without boundary, then specify whether the unit normal points
outward (or inward).
(3) If Φ : D → R3 is a ’nice’ parametrized surface, then for each (u, v), ∥Φ
Φu ×Φv
u ×Φv ∥
is a unit normal
vector to S at Φ(u, v), and we obtain an orientation of S.
An oriented geometric surface is a geometric surface with a specified orientation. If S is
the image of a parametrized surface Φ, then we say Φ is orientation preserving if ∥Φ Φu ×Φv
u ×Φv ∥
equals
the prescribed unit normal vector. In this case, we define
ZZ ZZ
F · dS = F · dS.
S Φ

If this is not the case, that is, Φ is orientation reversing, then −Φ is orientation preserving, and
we define ZZ ZZ ZZ
F · dS = F · dS = − F · dS.
S −Φ Φ

3.5.5 Gauss’s divergence theorem

We now generalize FTC, Part II to three dimensions, where the three-dimensional region is a
subset of R3 . This result relates an integral on a three-dimensional region W to a surface integral
on its boundary ∂W .

Definition 3.5.4 (Orienting the boundary surface). Let W be a bounded subset of R3 such that
∂W consists of a finite number of non-intersecting geometric surfaces. Orient each such surface
such that the unit normal vector points out of W . Then we say ∂W is positively oriented.

Example 3.5.5. Let W = {(x, y, z) ∈ R3 : 1 ≤ x2 + y 2 + z 2 ≤ 4}. Then ∂W consists of

n o
S1 = (x, y, z) ∈ R3 : x2 + y 2 + z 2 = 4

and n o
S2 = (x, y, z) ∈ R3 : x2 + y 2 + z 2 = 1 .
The positive orientation on ∂W is given by outward normals on S1 and inward normals on S2 .

More generally, the normal to the outer boundary is outward, and to each of the inner bound-
aries is inward.

Theorem 3.5.1 (Gauss’s Divergence Theorem). Let W ⊆ R3 be such that ∂W is positively

oriented. If P , Q, R are scalar fields defined on an open set containing W , then
ZZ ZZZ
P dy ∧ dz + Qdz ∧ dx + Rdx ∧ dy = (Px + Qy + Rz ) d(x, y, z). (3.26)
∂W W

Equivalently, for F = (P, Q, R),

ZZ ZZZ
F · dS = div F d(x, y, z).
∂W W

Proof. We break the proof in three steps as in the proof of Green’s Theorem.

88
 Chapter 3. Vector Calculus

(i) W is a cuboid. Let W = [a, b] × [c, d] × [p, q]. Let S1 be the face of W where z = p, and S2
be the face of W where z = q. We orient these surfaces by the normal going out of W . In other
words, S1 is oriented by n1 = (0, 0, −1), and S2 is oriented by n2 = (0, 0, 1). Using FTC, Part II,
ZZZ ZZ ZZ
Rz d(x, y, z) = R(x, y, q)d(x, y) − R(x, y, p)d(x, y)
W [a,b]×[c,d] [a,b]×[c,d]
ZZ ZZ
= F · n2 |dS| + F · n1 |dS|.
S2 S1

Similarly, rewrite W Qy d(x, y, z) and W Px d(x, y, z) to surface integrals over the remaining
RRR RRR

two pairs of faces. Adding these equalities proves formula (3.26) when W is a cuboid.
(ii) W is a "union" of cuboids. Now suppose W is formed by adjoining cuboids. Overlapping
faces have opposite orientations, so surface integrals over them cancel by (3.23). Thus, formula
(3.26) holds when W is formed by adjoining cuboids.
(iii) W is a general region. Draw a cuboidal grid inside W . Let W ′ be the largest region which
fits inside W which is of the form in item (ii) above and is made out of the faces of the cuboidal
grid. By item (ii), formula (3.26) holds for W ′ . Finally, we approach formula (3.26) for W by
considering finer and finer cuboidal grids.
Remark 3.5.2. Gauss’s divergence theorem gives us an interpretation of divergence: div F (a, b, c)
is the flux of the vector field F through the boundary of a small box around (a, b, c), that is, the
net flow of F out of the box. This explains the terminology "divergence".
Example 3.5.6. We verify the divergence theorem for vector field F = (x − y, x + z, z − y) and
surface S that consists of cone x2 + y 2 = z 2 , 0 ≤ z ≤ 1, and the circular top of the cone (see the
following figure). Assume this surface is positively oriented.

Let E be the solid cone enclosed by S. To compute the triple integral div F⃗ dV , note that
RRR
E
div F⃗ = Px + Qy + Rz = 2, and therefore the triple integral is
ZZZ ZZZ
div F⃗ dV = 2 dV = 2 (volume of E).
E E

89
Chapter 3. Vector Calculus

πr2 h
The volume of a right circular cone is given by 3
. In this case, h = r = 1. Therefore,
ZZZ
2π
div F⃗ dV = 2 (volume of E) = .
E 3
To compute the flux integral, first note that S is piecewise smooth; S can be written as a union
of smooth surfaces. Therefore, we break the flux integral into two pieces: one flux integral across
the circular top of the cone and one flux integral across the remaining portion of the cone. Call
the circular top S1 and the portion under the top S2 . We start by calculating the flux across the
circular top of the cone. Notice that S1 has parameterization
⃗r(u, v) = ⟨u cos v, u sin v, 1⟩, 0 ≤ u ≤ 1, 0 ≤ v ≤ 2π.
Then, the tangent vectors are ⃗tu = ⟨cos v, sin v, 0⟩ and ⃗tv = ⟨−u sin v, u cos v, 0⟩. Therefore, the
flux across S1 is
ZZ Z 1 Z 2π
⃗=
F⃗ · dS F⃗ (⃗r(u, v)) · (⃗tu × ⃗tv ) dA
S1 0 0
Z 1 Z 2π
= ⟨u cos v − u sin v, u cos v + 1, 1 − u sin v⟩ · ⟨0, 0, u⟩ dv du
0 0
Z 1 Z 2π
= u − u2 sin v dv du = π.
0 0

We now calculate the flux over S2 . A parameterization of this surface is

⃗r(u, v) = ⟨u cos v, u sin v, u⟩, 0 ≤ u ≤ 1, 0 ≤ v ≤ 2π.
The tangent vectors are ⃗tu = ⟨cos v, sin v, 1⟩ and ⃗tv = ⟨−u sin v, u cos v, 0⟩, so the cross product is
⃗tu × ⃗tv = ⟨−u cos v, −u sin v, u⟩.

Notice that the negative signs on the x and y components induce the negative (or inward) orienta-
tion of the cone. Since thesurface is positively oriented, we use vector ⃗tv × ⃗tu = ⟨u cos v, u sin v, −u⟩
in theflux integral. The flux across S2 is then
ZZ Z 1 Z 2π
⃗=
F⃗ · dS F⃗ (⃗r(u, v)) · (⃗tv × ⃗tu ) dA
S2 0 0
Z 1 Z 2π
= ⟨u cos v − u sin v, u cos v + u, u − u sin v⟩ · ⟨u cos v, u sin v, −u⟩ dv du
0 0
Z 1 Z 2π
π
= u2 cos2 v + 2u2 sin v − u2 dv du = − .
0 0 3
The total flux across S is
ZZ
⃗=
ZZ
⃗+
ZZ
π 2π ZZZ
F⃗ · dS F⃗ · dS F · dS = π − =
⃗ ⃗ = div F⃗ dV.
S S1 S2 3 3 E

3.5.6 Volume calculation.

Corollary 3.5.1. Let S be a simple closed geometric surface oriented by the outward normal which
encloses a region W . Then
1 ZZ
Vol(W ) = xdy ∧ dz + ydz ∧ dx + zdx ∧ dy,
3 S
ZZ ZZ ZZ
= xdy ∧ dz = ydz ∧ dx = zdx ∧ dy. (3.27)
S S S

90
 Chapter 3. Vector Calculus

Proof. Put P (x, y, z) = x3 , Q(x, y, z) = y3 , R(x, y, z) = z3 . Then Px + Qy + Rz = 1

3
+ 1
3
+ 1
3
= 1. By
formula (2.14),
ZZZ ZZZ
Vol(W ) = d(x, y, z) = (Px + Qy + Rz ) d(x, y, z),
ZZ W W

= P dy ∧ dz + Qdz ∧ dx + Rdx ∧ dy,

∂W =S
1 ZZ
= xdy ∧ dz + ydz ∧ dx + zdx ∧ dy.
3 S

The third equality is by (3.26).

Corollary 3.5.2. For a simple closed surface S parametrized by (x(u, v), y(u, v), z(u, v)) which
agrees with the outward normal,
1 ZZ
" #
∂(y, z) ∂(z, x) ∂(x, y)
Vol(W ) = x(u, v) + y(u, v) + z(u, v) d(u, v), (3.28)
3 D ∂(u, v) ∂(u, v) ∂(u, v)
 
x(u, v) y(u, v) z(u, v)
1 ZZ
= det  xu (u, v)

yu (u, v) zu (u, v) 
 d(u, v).
3 D
xv (u, v) yv (u, v) zv (u, v)
x2 y2 z2
Example 3.5.7. Let S be the surface defined by a2
+ b2
+ c2
= 1. Parametrize S by
x = a sin φ cos θ, y = b sin φ sin θ, z = c cos φ
for 0 ≤ φ ≤ π and 0 ≤ θ ≤ 2π. Then
a sin φ cos θ b sin φ sin θ c cos φ
   
x(φ, θ) y(φ, θ) z(φ, θ)
 xφ (φ, θ) yφ (φ, θ) zφ (φ, θ)  =  a cos φ cos θ b cos φ sin θ −c sin φ 
   

xθ (φ, θ) yθ (φ, θ) zθ (φ, θ) −a sin φ sin θ b sin φ cos θ 0

whose determinant equals abc sin φ. Thus, by formula (3.28), volume of the elliptical region en-
closed by S is
1 ZZ 4
abc sin φd(φ, θ) = πabc.
3 [0,π]×[0,2π] 3
In particular, the volume of a sphere of radius a is 34 πa3 .

3.5.7 Stokes theorem

We now generalize FTC, Part II to two dimensions, where the two-dimensional region is a surface
in R3 . In the special case when this surface lies in R2 , we recover Green’s theorem. This result
relates a surface integral on an oriented surface S to a line integral on its boundary ∂S.
Definition 3.5.5 (Orienting the boundary curve). Let S be an oriented geometric surface in
R3 . It induces an orientation on ∂S as follows. Walk along ∂S with the prescribed unit normal
vector as our upright direction so that S lies to our left. This is the induced orientation on ∂S.
Theorem 3.5.2 (Stokes theorem). Let S be an oriented geometric surface in R3 . Suppose ∂S
consists of a finite number of nonintersecting simple closed curves. Let ∂S be given the induced
orientation. If P, Q, R are scalar fields defined on a region containing S, then
Z ZZ
P dx + Qdy + Rdz = (Ry − Qz ) dy ∧ dz + (Pz − Rx ) dz ∧ dx + (Qx − Py ) dx ∧ dy. (3.29)
∂S S

91
Chapter 3. Vector Calculus

Equivalently, for F = (P, Q, R),

Z ZZ
F · ds = curl F · dS. (3.30)
∂S S

Proof. We break the proof in three steps.

(i) S is a parallelogram. We assume that the sides of the parallelogram are parallel to the x-axis
and y-axis. Let
Φ : [a, b] × [c, d] → S, (x, y) 7→ (x, y, αx + βy).
Hence, Φx × Φy = (−α, −β, 1).
ZZ ZZ
curl F · dS = −α (Ry − Qz ) − β (Pz − Rx ) + (Qx − Py ) dx ∧ dy,
S Z Z D
F · ds = (P + αR)dx + (Q + βR)dy.
∂S ∂D

Why are the two results equal? The answer is Green’s theorem on the rectangle! Let us check
this.
(Q + βR)x − (P + αR)y = Qx + αQz + β (Rx + αRz ) − (Py + βPz ) − α (Ry + βRz ) ,
= −α (Ry − Qz ) − β (Pz − Rx ) + (Qx − Py ) .

For the first equality, note very carefully that we are taking partial wrt x of Q(x, y, αx + βy), and
so on. This proves formula (3.30) when S is a parallelogram.
(ii) S is a "union" of parallelograms. This follows from item (i). The contribution along com-
mon edges cancels.
(iii) S is a general surface. Divide S into small pieces so that each piece is roughly a parallelo-
gram. Use item (ii), and then take limits.
Example 3.5.8. (1) Let us verify Stokes theorem, that is, formula (3.30) for the upper hemisphere
2 2 2
for the vector field F = (x, y, z). Observe: F = ∇f for f (x, y, z) = x +y2 +z . Thus, F is a gradient
vector field. Hence, curl F = 0 by Lemma 3.2.1. Thus, the surface integral in (3.30) is zero. Also,
Z Z 2π
F · ds = (cos t, sin t, 0) · (− sin t, cos t, 0)dt = 0.
∂S 0

Thus, the line integral in (3.30) is zero.

(2) Consider the surface defined by
n √ o
S := (x, y, z) ∈ R3 : x2 + y 2 + (z − 3)2 = 4, z ≥ 0 .
√
It is the portion of the sphere of radius 2 centered at (0, 0, 3) which lies above the xy-plane.
Orient it by the outward normal of the sphere. Observe:
n o
∂S = (x, y, z) ∈ R3 : x2 + y 2 = 1, z = 0 .

This is the unit circle in the xy-plane. The induced orientation on ∂S is anticlockwise. Hence the
parametrization (cos θ, sin θ, 0) for 0 ≤ θ ≤ 2π is orientation preserving. Consider the vector field
F = (y, −x, exz ). By (3.30),
ZZ Z Z 2π  
curl F · dS = F · ds = sin θ, − cos θ, e(cos θ)(0) · (− sin θ, cos θ, 0)dθ = −2π.
S ∂S 0

92
 Chapter 3. Vector Calculus

Remark 3.5.3. Stokes Theorem gives us an interpretation of curl: (curl F )(a, b, c) is the flow of
F around a small loop lying in the plane perpendicular to curl F . This explains the terminology
"curl".
Question 3.5.1. Suppose we are given curl F = 0. Then is F = grad f for some f ? That is, is
F a gradient vector field? or is C F · ds = 0 for every closed curve C ?
R

The answer is yes when the domain of F is simply connected. In rough terms, it means that
the domain has no holes. More precisely, it means that every closed curve in the domain can be
shrunk to a point by staying within that domain.
Example 3.5.9 (Simply connected). We illustrate this concept.
(1) R2 is simply connected. But R2 with a point removed is not simply connected.
(2) A closed disc in R2 is simply connected. But a closed disc with an interior point removed
is not simply connected. Similarly, an annulus is not simply connected.
(3) R3 is simply connected. R3 with a point removed is simply connected. But R3 with the
entire z-axis removed is not simply connected.
Answer of Question 3.5.1: Suppose the domain of F is simply connected, and curl F = 0. We
indicate why why F is a gradient vector field. Let C be any closed curve. Since the domain of F
is simply connected, one can find an orientable surface S whose boundary is C. Now apply (3.30)
to S.
 
Example 3.5.10. (1) Consider the vector field F (x, y) = x2−y , x
+y 2 x2 +y 2
on the right half plane
x > 0. This region
 is simply connected. So F is a gradient vector field on this region. Check that
f (x, y) = tan −1 y
x
is a potential function for F on this region.
 
(2) Consider the vector field F (x, y, z) = x2−y , x , 0 on R3 \{(0, 0, z)}, that is, on R3 minus
+y 2 x2 +y 2
z-axis. Note: curl F = 0. For the path γ : [0, 2π] → R3 , γ(t) = (cos t, sin t, 0),
Z Z 2π Z 2π
F · ds = (− sin t, cos t, 0) · (− sin t, cos t, 0)dt = dt = 2π.
γ 0 0

Hence F is not a gradient vector field.

3.6 Problem Sheet 4

1. Find the length of the epicycloid with k cusps parameterized by
 
σ(t) = R(k + 1) cos t − R cos((k + 1)t), R(k + 1) sin t − R sin((k + 1)t) , 0 ≤ t ≤ 2π.

2. Let C be the curve given by r(t) = t cos t î + t sin t ĵ + t2 k̂, 0≤t≤π

(a) Find the unit tangent T̂ to C at the point (−π, 0, π 2 ).

(b) Calculate the line integral Z q
x2 + y 2 ds.
C

3. LetH F(x, y, z) = ex sin y î + [aex cos y + bz] ĵ + cx k̂. For which values of the constants a, b, c
is C F · dr = 0 for all closed paths C?

93
Chapter 3. Vector Calculus

4. Let S be the part of the surface z = xy lying inside the cylinder x2 + y 2 = 3. Find the
moment of inertia of S about the z-axis, that is,
ZZ
I= (x2 + y 2 ) dS.
S

5. Find the surface area of the part of the paraboloid z = a2 − x2 − y 2 which lies above the
xy-plane.
6. Find the area of the portion of the cone z 2 = x2 + y 2 lying between the planes z = 2 and
z = 3.
7. Determine the surface area of the surface given by z = 2
3
(x3/2 + y 3/2 ), over the square
0 ≤ x ≤ 1, 0 ≤ y ≤ 1.
Compute the net outward flux of the vector field

r xî + y ĵ + z k̂
F= =√ 2
|r| x + y2 + z2
across the boundary of the region between the spheres of radius 1 and radius 2 centred at
the origin.
8. Evaluate the surface integral z 2 dS where S is the part of the cone x2 + y 2 = 4z 2 where
RR
S
0 ≤ x ≤ y and 0 ≤ z ≤ 1.
9. A physicist studies a vector field F in her lab. She knows from theoretical considerations
that F must be of the form F = ∇ × G, for some smooth vector field G. Experiments also
show that F must be of the form

F(x, y, z) = (xz + xy)î + α(yz − xy)ĵ + β(yz + xz)k̂

where α and β are constant. Determine α and β. Further experiments show that G =
xyz î − xyz ĵ + g(x, y, z)k̂. Find the unknown function g(x, y, z).
10. By applying the divergence theorem to F = ϕa, where a is an arbitrary constant vector,
show that ZZZ ZZ
∇ϕ dV = ϕ n̂ dS.
V ∂V

11. Let S be the unit sphere centered at the origin and oriented by the outward pointing normal.
If F(x, y, z) = (x, y, z 2 ) evaluate the flux of F through S directly, and by applying the
divergence theorem.
12. Evaluate C F·dr where F = x2 y 2 î+2xy ĵ and C is the boundary of the square in the xy-plane
H

having one vertex at the origin and diagonally opposite vertex at the point (3, 3), oriented
counterclockwise.
13. Evaluate I    
x sin y 2 − y 2 dx + x2 y cos y 2 + 3x dy
C
where C is the counterclockwise boundary of the trapezoid with vertices (0, −2), (1, −1),
(1, 1), and (0, 2).

94
 Chapter 3. Vector Calculus

14. Evaluate
1 2 3
I    
I= x y − x4 y dx + xy 4 + x3 y 2 dy
C 3
√
counterclockwise around the boundary of the half-disk 0 ≤ y ≤ 4 − x2 .

15. Verify the identity I I

ϕ∇ψ · dr = − ψ∇ϕ · dr
C C
for any continuously differentiable scalar fields ϕ and ψ and curve C that is the boundary of
a piecewise smooth surface.
√
16. Let C be the curve given by the parametric equations: x = cos t, y = 2 sin t, z =
cos t, 0 ≤ t ≤ 2π and let F = z î + xĵ + y 3 z 3 k̂ Use Stokes’ theorem to evaluate C F · dr.
H

17. Find the value of ∇ × F · n̂ dS where F = (z − y, x, −x) and S is the hemisphere

RR
S

{(x, y, z) ∈ R3 | x2 + y 2 + z 2 = 4, z ≥ 0}

oriented so the surface normals point away from the centre of the hemisphere.

18. Let S be the part of the surface z = 16 − (x2 + y 2 )2 which lies above the xy-plane. Let F be
the vector field F = x ln(1 + z)î + x(3 + y)ĵ + y cos z k̂. Calculate
ZZ
∇ × F · n̂ dS
S

where n̂ is the upward normal on S.

95