MA-111 Calculus II

(D3 & D4 )

Lecture 5

B.K. Das

Department of Mathematics
Indian Institute of Technology Bombay
Powai, Mumbai - 76

February 1, 2022
Evaluating integrals over Elementary regions

The integral in polar coordinate

Recap

I If D is a bounded region in R2 and f : D → R is bounded, then

consider any rectangle R containing the region D in R2 and extend
f to the rectangle by 0 outside D and denote it by f ∗ . The integral
of f over D is defined by the integral of f ∗ on the rectangle R.
I The above definition is consistent because the definition of integral
of f on D is independent of the choice of rectangle R.
I To determine the integrability of f over region D, conditions on f
and D? The boundary of D should be ‘well-behaved’. The set
containing points of discontinuity of f is of ‘content zero’.
I Algebraic properties of integrals on D are similar to that of the
integrals on rectangle.
We will now discuss two types of regions for which D is a bounded set in
R2 with its boundary ∂D of content zero and the integral can be
evaluated easily.
We will describe two simple types of regions known as elementary regions.
Existence of Integrals over bounded sets in R2

Theorem
Let D ⊂ R2 be a bounded set whose boundary ∂D is given by the finitely
continous closed curve then any bounded and continuous function
f : D → R is integrable over D.
Example. Let D = {(x, y ) | x 2 + y 2 ≤ 1} and
f (x, y ) = x 2 + y 2 , ∀ (x, y ) ∈ D. Then f is integrable over D.

A slightly more general theorem is as follows:

Let D be a bounded set in R2 such that ∂D is of content zero. Let
f : D → R be a bounded function whose points of discontinuity have
‘content zero’. Then f is integrable over D.
Elementary region: Type 1
Let h1 , h2 : [a, b] → R be two continuous
functions such that h1 (x) ≤ h2 (x) for all
x ∈ [a, b]. Consider the set of points

D1 = {(x, y ) | a ≤ x ≤ b and h1 (x) ≤ y ≤ h2 (x)}.

Such a region is said to be of Type 1 and

for every x ∈ R vertical cross-section of D1
is an interval.
Example. D1 = {(x, y ) ∈ R2 | 0 ≤ x ≤ 2, x 2 ≤ y ≤ 2x}. Here for all
x ∈ [0, 2], h1 (x) = x 2 and h2 (x) = 2x. Note h1 (x) ≤ h2 (x) for x ∈ [0, 2].
Type 1 contd.
Example. The closed disc Dr of radius r around the origin,
Dr := {(x, y ) ∈ R2 | x 2 + y 2 ≤ r 2 }.
√ √
Take h1 (x) = − r 2 − x 2 and h2 (x) = r 2 − x 2 . We see that Dr is of
Type 1.
For Type 1, let D1 = {(x, y ) | a ≤ x ≤ b and h1 (x) ≤ y ≤ h2 (x)}, where
h1 and h2 are continuous.
The region D1 is bounded by continuous curves (the straight lines x = a
and x = b and the graphs of the curves y = h1 (x) and y = h2 (x) for
x ∈ [a, b]).
Thus ∂D1 is of ‘content zero’ in R2 .
Hence any continuous function defined on D1 is integrable over the
elementary region D1 .
Evaluating integrals on regions of Type 1
Let D be a region of Type 1 and assume that f : D → R is continuous.
Let D ⊂ R = [α, β] × [γ, δ] and let f ∗ be the corresponding function on
R (obtained by extending f by zero).
The region D is bounded by continuous curves (the straight lines x = a
and x = b and the graphs of the curves y = h1 (x) and y = h2 (x)).
Hence we can conclude that f ∗ is integrable on R. Applying Fubini’s
theorem on f ∗ we get,
Z Z Z Z Z β "Z δ #
∗ ∗
f (x, y )dxdy := f (x, y )dxdy = f (x, y )dy dx.
D R α γ

In turn, this gives

Z β "Z h2 (x) # Z b
"Z
h2 (x)
#
∗
f (x, y )dy dx = f (x, y )dy dx,
α h1 (x) a h1 (x)

since f ∗ (x, y ) = 0 if y < h1 (x) or y > h2 (x). Finally, we get

Z Z Z "Z b h2 (x)
#
f (x, y )dxdy = f (x, y )dy dx.
D a h1 (x)
Examples
Example Let D = {(x, Ry )R | 0 ≤ x ≤ 2, x 2 ≤ y ≤ 2x} and
f (x, y ) = x + y . Find D
f (x, y )dxdy .
Ans Note D is a bounded set in R2 enclosed by the graphs of the curves
y = x 2 and y = 2x and hence ∂D is of content zero. Since f is
continuous over D and D is bounded with ∂D of content zero, f is
integrable over D.

2 2x Z 2
y2
Z Z Z Z 
f (x, y ) dxdy = (x + y ) dy dx = [xy + ]yy =2x
2 dx
D 0Z x2 0 2 =x
2
x2 x4
= [2x 2 + 4 − x 3 − ] dx
0 2 2
2 2
Example Letp D = {(x, y ) |R xR + y ≤ 1, x ≥ 0, y ≥ 0} and
f (x, y ) = 1 − y 2 . Find D
f (x, y )dxdy .
√
Ans Type 1, i.e, D = {(x, y ) | 0 ≤ x ≤ 1, 0 ≤ y ≤ 1 − x 2 }. Then
Z Z Z 1  Z √1−x 2 p 
f (x, y ) dxdy = 1 − y 2 dy dx.
D 0 0

Not easy to compute!

Elementary region: Type 2

Similarly, if k1 , k2 : [c, d] → R are two

continuous functions such that
k1 (y ) ≤ k2 (y ), for all y ∈ [c, d]. The set of
points

D2 = {(x, y ) | c ≤ y ≤ d and k1 (y ) ≤ x ≤ k2 (y )}

is called a region of Type 2 and for every

y ∈ R horizontal cross-section of D2 is an
interval.
p
Example pD2 = {(x, y ) | x 2 + y 2 ≤ 1}. If we take k1 (y ) = − 1 − y 2 and
k2 (y ) = 1 − y 2 , we see that D2 is of Type 2.
Evaluating integrals on regions of type 2

Note that the boundary of D2 is of content zero in R2 . Hence any

continuous function defined on D2 is integrable over the elementary
region.
Using exactly the same reasoning as in the previous case (basically,
interchanging the roles of x and y ) we can obtain a formula for regions of
Type 2.

Let D be a bounded set of Type 2 in R2 . Let f : D → R be a continuous

function on D. We get
"Z #
Z Z Z d k2 (y )
f (x, y )dxdy = f (x, y )dx dy .
D c k1 (y )
Example Let D = {(x, y ) | x 2 + y 2 ≤ 1, x ≥ 0, y ≥ 0}. Evaluate the
integral Z Z p
1 − y 2 dxdy .
D
Z Z p Z 1  Z √1−y 2 p 
2
1 − y dxdy = 1 − y 2 dx dy
Ans. D Z 1 p0 0√ Z 1
1−y 2 2
= [x 1 − y 2 ]x=0 dy = (1 − y 2 ) dy = .
0 0 3
Remark

Both of these formulæ can be viewed as special cases of Cavalieri’s

principle when f (x, y ) ≥ 0. In the first case we are slicing by planes
perpendicular to the x-axis, while in the second case, we are slicing by
planes perpendicular to the y -axis.
Caution! There exist bounded subsets of R2 which are not elementary
regions; for example, star-shaped subset of R2 or an annulus.
Often we can write D as a union of regions of Types 1 and 2 and then we
call it a region of type 3.
Note the Domain Additivity theorem will then allow us to evaluate
integrals which are defined over finite union of such sets.
We could also view the disc as a region of type 3, by dividing it into four
quadrants.
Remark contd.
What about the annulus A = {(x, y ) ∈ R2 | s 2 ≤ x 2 + y 2 ≤ r 2 }?
Is it a type 3 region? yes
Example1: Compute the integral of f (x, y ) = x 2 + y 2 on
D = {(x, y ) ∈ R2 | x 2 + y 2 ≤ 1}.

Can we compute this integral using iterated integrals ?

2
+y 2
Example2: Compute the integral of g (x, y ) = e x on
D = {(x, y ) ∈ R2 | x 2 + y 2 ≤ 1}.

Can we use substitution like we did in one variable?

Let us see what happens when we use polar coordinates.

Polar Coordinates
Change of variables from Cartesian coordinate system to polar coordinate
system, any (x, y ) ∈ R2 in Cartesian coordinate can be written as
x = r cos(θ), y = r sin(θ), r > 0, θ ∈ [0, 2π].

Transformation of region under change of variables:

Ex. D := {(x, y ) ∈ R2 | x 2 + y 2 ≤ a2 } is transformed in polar coordinate
system as a rectangle
D ∗ = {(r , θ) | 0 ≤ r ≤ a, θ ∈ [0, 2π]}.
The integral in polar coordinates
Let D ∗ be a subset of R2 in polar coordinate system, such that for all
(r , θ) ∈ D ∗ , (r cos(θ), r sin(θ)) ∈ D, for 0 ≤ r ≤ 1, and

g (r , θ) := f (r cos(θ), r sin(θ)), (r , θ) ∈ D ∗ .

To integrate the function g on a domain D ∗ we need to cut up D ∗ into

small rectangles, but these will be rectangles in the r -θ coordinate
system.

What shape does a rectangle

[r , r + ∆r ] × [θ, θ + ∆θ] represent in
the x-y plane? A part of a sector of
a circle.

Then we will be integrating over this sector instead of rectangle.

What is the area of this part of a sector?
Ans: It is 1
2 · [(r + ∆r )2 ∆θ − r 2 ∆θ] ∼ r ∗ ∆r ∆θ, r ≤ r ∗ ≤ r + ∆r .
Partitioning the region into subrectangles is equivalent to partitioning the
region into parts of sectors as shown earlier.
It follows that the integral we want is approximated by a sum of the form
XX
g (ri∗ , θj∗ )ri∗ ∆ri ∆θj ,
i j

where {(ri∗ , θj∗ } is a tag for the partition of the “rectangle” in polar
coordinates and
Z Z Z Z
f (x, y )dxdy = f (r cos θ, r sin θ)rdrdθ,
D D∗

where D is the image of the region D ∗ .

This is the change of variable formula for polar coordinates.
Examples

Example1: Integrate f (x, y ) = x 2 + y 2 on D = {(x, y ) | x 2 + y 2 ≤ 1}.

Solution: Let us use polar coordinates. Let

D ∗ = {(r , θ) | 0 ≤ r ≤ 1, 0 ≤ θ ≤ 2π}.

Denoting x = r cos θ and y = r sin θ, the polar coordinates will transform

D ∗ to D and
g (r , θ) = f (r cos θ, r sin θ) = r 2 .
Z Z Z Z Z Z
f (x, y ) dxdy = g (r , θ) r drdθ = r 2 .r drdθ
D D∗ [0,1]×[0,2π]
2π 1 2π
r4
Z Z Z 1 π
= r 3 drdθ = dθ =
0 0 0 4 0 2
Examples contd.

2 2
Example 2: Integrate f (x, y ) = e x +y on D = {(x, y ) | x 2 + y 2 ≤ 1}.
Solution: Using the same transformation as above

x = r cos θ, y = r sin θ,

we get
Z Z Z Z Z Z
2
f (x, y ) dxdy = g (r , θ) r drdθ = e r r drdθ
D D∗ [0,1]×[0,2π]

2π 1 2π 2
er
Z Z Z 1
2
= e r r drdθ = dθ = π(e − 1)
0 0 0 2 0
An Applicaton:The integral of the Gaussian

We would like to evaluate the following integral:

Z ∞
2
I = e −x dx.
−∞

What does this integral mean? - so far we have only looked at Riemann
integrals inside closed bounded intervals, so the end points were always
finite numbers a and b.
An integral like the one above is called an improper integral. We can
assign it a meaning as follows. It is defined as
Z T
2
lim e −x dx,
T →∞ −T

provided, of course, this limit exists. We will see how to evaluate this.
The most amazing trick ever

Consider Z ∞ Z ∞
2 2
I2 = e −x dx · e −y dy .
−∞ −∞

We view this product as an iterated integral!

Z ∞Z ∞
2 2
2
I = e −x e −y dxdy .
−∞ −∞

Now under polar coordinates, the plane is sent to the plane. Hence, we
can write this as Z 2π Z ∞ 
−r 2
e rdr dθ.
0 0

But we can now evaluate the inner integral. Hence, we get

Z 2π   Z 2π
1 −r 2 ∞ 1
− e dθ = dθ = π
0 2 0 0 2
The answer

√
Since I 2 = π, we see that I = π.
Using the above result you can easily conclude that
Z ∞ r
−αx 2 π
e dx = .
−∞ α

The integral above arises in a number of places in mathematics - in

probability, the study of the heat equation, the study of the Gamma
function (next semester) and in many other contexts.
There are many other ways of evaluating the integral I , but the method
above is easily the cleverest.