Math1120 - Mathematics II

Calculus (V): Linear Approximations

In Math1110 we looked at the idea of “the linearization of a function of one variable”.
The idea was that near the point x  x0 we could approximate the function f ( x) using the
line tangent to the function at x  x0 . We called this tangent the linearization of the function
x  x0 and used it to derive the “linear approximation formula”,

y f ( x0 )x

These ideas can also be applied to functions of two variables.

§1 Tangent Planes

Recall that the graph associated with the function z  f ( x, y) is a surface in 3 (that
passes the vertical line test). Wherever this surface does not have any discontinuities or cusp-
like points it will have a tangent plane. Like the tangent line to the graph of a function of one
variable, the tangent plane to the function z  f ( x, y) at the point ( x0 , y0 ) is the plane that
“just touches” the surface at the point ( x0 , y0 , z0 ) .

 x2  y 2 
Example 1: For the function f ( x, y )  5    the graph below shows the graph
 2 
of the function and it’s tangent plane at the point ( x, y)  (2,1) .

Figure 1
☼

Page 45
 Calculus (V) - Lecture Notes

To find the equation of the plane tangent to the function z  f ( x, y) at the point
 x0 , y0 , z0  , firstly recall that the equation of a plane is given by (in Cartesian form)

ax  by  cz  k , a, b, c, k  ,

or, as illustrated in Figure 2, (in normal form) by

n  (r  r0 )  0 (1)

where n  a, b, c is a normal vector to the plane, r0 is the position vector of a point on the
plane and r  x, y, z .

Figure 2
 x2  y 2 
Example 2: Find the equation of tangent plane to the function f ( x, y )  5    at
 2 
( x, y)  (2,1) .

Now, we know that the partial derivative f x (2,1) gives the slope of the tangent at,
x  2 , to the curve of intersection of the surface associated with f and the plane y  1 .

Figure 3

Page 46
 Math1120 - Mathematics II

Since this tangent line lies in the plane tangent to f at ( x, y)  (2,1) the vector

1,0, f x (2,1)

will be a vector that is parallel to the tangent plane, (or lies in the tangent plane if we
place it’s tail at the point  2,1,5 2  ). Similarly, the vector

0,1, f y (2,1)

will be another vector parallel to the tangent plane. Since these two non-parallel vectors
are parallel to the tangent plane, their vector product will give a vector normal to the
tangent plane, i.e.
n  1, 0, f x (2,1)  0,1, f (2,1)
  f x (2,1),  f y (2,1),1
 2,1,1
 x2  y 2 
Thus, using equation (1), the equation of the plane tangent to f ( x, y )  5    at
 2 
( x, y)  (2,1) is
 5 
2,1,1   x, y, z  2,1,   0
 2 
which simplifies to
15
2x  y  z  .
2
☼

In general, to find the equation of the plane tangent to the function f ( x, y) at the point
 x0 , y0  , note that the vectors
1,0, f x ( x0 , y0 and 0,1, f y ( x0 , y0
lie in the tangent plane and hence a normal to the plane is

n  1, 0, f x ( x0 , y0 )  0,1, f ( x0 , y0 )
  f x ( x0 , y0 ),  f y ( x0 , y0 ),1
Since the point  x0 , y0 , z0  lies on the plane, using equation (1), the equation of the tangent
plane is
 f x ( x0 , y0 ),  f y ( x0 , y0 ),1   x, y, z  x0 , y0 , z0   0 .

On expanding and rearranging this we get the following result.

Page 47
 Calculus (V) - Lecture Notes

The equation of the plane tangent to the graph of the function f ( x, y) at the point
 x0 , y0 , z0  (where z0  f ( x0 , y0 ) ) is:
z  z0  f x ( x0 , y0 )( x  x0 )  f y ( x0 , y0 )( y  y0 )

Example 3: Find the equation of the plane tangent to z  x 2  2 y 2 at the point

 x, y   1, 2 .

Figure 4

Firstly, note that when  x, y   1, 2  , z  9 . Now

zx  2 x and so zx (1, 2)  2
and
z y  4 y and so z y (1, 2)  8 .
Thus the equation of the tangent plane is
z  9  2( x  1)  8( y  2)
which simplifies to
z  2x  8 y  9 .
☼

Example 4: Find the equation of the line normal to the graph of function z  x 2  2 y 2 at
the point  x, y   1, 2  .

Recall that the vector equation of a line in 3-d is

r  r0  td , t 
where r  x, y, z is the position vector of a general point, r0 is the position vector of
a point that lies on the line and d is a direction vector for the line (i.e. a vector that is
parallel to the line).

Page 48
 Math1120 - Mathematics II

We know that a vector normal to the surface z  f ( x, y) at the point  x0 , y0 , z0 

is given by
n   f x ( x0 , y0 ),  f y ( x0 , y0 ),1
For the function z  x 2  2 y 2 , zx  2 x and z y  4 y . Thus a direction vector for
the line normal to z  x 2  2 y 2 at the point  x, y   1, 2  will be
d  2, 8,1
Since the normal line passes through the point 1, 2,9  its equation is
r  1, 2,9  t 2, 8,1 .
☼

Example Tasks

ET 1: Find the equation of the tangent plane to z  x2  5xy  2 y 2 at the point

 x, y    1, 2 .
ET 2: Find the equation of the tangent plane and normal line to z  x  y sin( xy) at
the point  x, y    0,1 .

ET 3:* Show that every line that is normal to the sphere

x2  y 2  z 2  1
passes through the origin.

§2 Linear Approximations

In §1 above we found the equation of the tangent plane to the function f ( x, y) at the
point  x0 , y0 , f ( x0 , y0 )  . This equation is a linear equation.

We call the function

L  x, y   f ( x0 , y0 )  f x ( x0 , y0 )( x  x0 )  f y ( x0 , y0 )( y  y0 )
the linearisation of f at the point  x0 , y0  .

  
Example 5: Find the linearisation of z  sin( x) cos( y) at the point  x, y    ,  .
4 4

Begin by calculating the partial derivatives of z.

zx  cos( x) cos( y) and z y   sin( x)sin( y)
Thus
       1
z  ,   sin   cos   
4 4 4 4 2

Page 49
 Calculus (V) - Lecture Notes

       1
z x  ,   cos   cos   
4 4 4 4 2
       1
z y  ,    sin   sin    
4 4 4 4 2
and so the linearisation is
1 1   1 
L( x, y)   x   y 
2 2 4  2 4
☼

When we use the linearisation of f at the point  x0 , y0  to approximate the function near
the point  x0 , y0  we call this the linear (or tangent plane) approximation of f at the point
 x0 , y0  . Notice that if we let the independent variables change by the amounts x and y
then the linearisation will change from L( x0 , y0 ) to L( x0  x, y0  y) . Thus we can
approximate the change in the function value z  f ( x, y) by

z L( x0  x, y0  y)  L( x0 , y0 ) .

On using the linearisation formula given above, we end up with the following result.

The Linear Approximation Formula

The linear approximation to the change, z , in the function z  f ( x, y) when the

independent variables change from  x0 , y0  to ( x0  x, y0  y) is
z f x ( x0 , y0 )x  f y ( x0 , y0 )y .

This result is sometimes called the “small change” formula for functions of two variables.

Example 6: For the function f ( x, y)  3x 4  2 y 4 , f (1, 2)  35 . Use a linear

approximation to estimate f 1.01, 2.03 .

Via a linear approximation

f 1.01, 2.03 f (1, 2)  f .
Here
f x  12 x3 and f y  8 y 3
and so
f x (1, 2)  12 and f y (1, 2)  64 .
Thus, with x  0.01 and y  0.03 , via the linear approximation formula

f  12  0.01  64  0.03  2.04

and hence
f (1.01, 2.03) 35  2.04  37.04 .
☼

Page 50
 Math1120 - Mathematics II

Example 7: A steel ball has a mass, m, of 6300  50 g and has volume, V, 800  10 cm3.
Find the density of the ball, including an estimate of the error.

The density,  , is given by

m
 .
V
Thus
6300
  7.875 g/cm3
800
Using a linear approximation to estimate the error
 
 m  V .
m V
Now
 1  m
 and  2
m V V V
and so at m  6300 , V  800 and with m  50 and V  10 (to get the maximum
value of  )
1 6300
  50  10  0.1609375 .
800 8002
Thus
  7.875  0.1609375 g/cm3 .
☼

Example 8: Use a linear approximation to estimate the value of z at

( x, y)  1.1, 0.02  for surface z  f ( x, y) defined implicitly by z  yz 3  x  2 .

Firstly notice that when x  1 and y  0 , z  3 . Thus z (1,0)  3 . Now, via a

linear approximation
z(1.1, 0.02) z(1,0)  z
where
z zx (1, 0)x  z y (1, 0)y ,
and x  0.1 , y  0.02 . To find the partial derivatives we need to use implicit
differentiation. Differentiating with respect to x:

 
( z  yz 3 )  ( x  2)
x x
z x  3 yz z x  1
2

1
zx 
1  3 yz 2 .
Thus
1
z x (1, 0)   1.
1  3  0  32

Differentiating with respect to y:

Page 51
 Calculus (V) - Lecture Notes

 
( z  yz 3 )  ( x  2)
y y
z y  (3 yz 2 z y  z 3 )  0
z3
zy 
1  3 yz 2 .
Thus
33
z y (1, 0)   27 .
1  3  0  32
Putting this together gives
z 1 0.1  27  (0.02)  0.44
and hence
z(1.1, 0.02) 3  0.44  2.56
☼

Example Tasks

ET 1: Use a linear approximation to find the value of z  2.96, 0.95 when

z  x 2  xy  3 y 2

ET 2: Use a linear approximation to estimate the value of (2.01)2  0.98 .

ET 3:* A right angled triangle ABC with right angle at B is measured with
AB  10  0.02 cm and BC  3.4  0.02 cm. What is the angle at A, including the error?

ET 4:* In the figure below a rectangle initially with sides x and y has been made larger so
that the sides are now x  x and y  y .

Shade on the diagram the regions that represent:

(a) The increase in area.
(b) The linear approximation to the increase to the area. Explain your answer.

Page 52
 Math1120 - Mathematics II

§3 Quadratic Approximations

We have seen previously that for functions of one variable the idea of the linearisation
of the function could be extended by considering the Taylor polynomial for the function.

Example 9: For the function f ( x)  e x the Maclaurin polynomial of degree n is

n
xk x2 xn

k 0 k !
 1 x  
2!
 .
n!

We saw that the linearisation of e x at x  0 was the Maclaurin polynomial of degree 1,

i.e.
T1 ( x)  1  x .
The Maclaurin polynomial of degree 2, i.e.
x2
T2 ( x)  1  x 
2
gives a quadratic approximation to e about x  0 and so on.
x

For functions of one variable we derived the Taylor series by trying to find a power
series in ( x  a) that matched the function and all its derivatives at x  a . For functions of
two variables we can use a similar idea to derive the Taylor series about the point
( x, y)  (a, b) . This series will be a power series in ( x  a) and ( y  b) that matches the
function and its partial derivatives at the point ( x, y)  (a, b) . The formula for the Taylor
series of two variables is quite lengthy to write and so we will not reproduce it here.
However, as with functions of one variable truncations of this series are called Taylor
polynomials and the Taylor polynomial of degree 1 is the linearisation of the function.
Similarly, the Taylor polynomial of degree 2 will give us a quadratic approximation to the
function.

The Taylor polynomial of degree 2 for the function of two variables f ( x, y) about the
point ( x, y)  (a, b) is

T2 ( x, y )  f (a, b)  f x (a, b)( x  a)  f y (a, b)( y  b)

f xx (a, b) f ( a, b)
 ( x  a)2  f xy (a, b)( x  a)( y  b)  yy ( y  b) 2
2! 2!

The question of how good an approximation this polynomial is goes beyond what we
will cover in this course but if f ( x, y) has continuous partial derivatives and if ( x, y) is
“sufficiently close to” (a, b) then the approximation should be useful.

Example 10: Find the quadratic approximation to the function f ( x, y)  x 2 y about the
point ( x, y)  (1, 2) .

Page 53
 Calculus (V) - Lecture Notes

First calculate the partial derivatives:

f x  2 xy , f y  x 2 , f xx  2 y , f xy  2 x , f yy  0 .
Now evaluate these at ( x, y)  (1, 2)
f x (1, 2)  4 , f y (1, 2)  1, f xx (1, 2)  4 , f xy (1, 2)  2 , f yy (1, 2)  0 .
Thus, using f (1, 2)  2 , we have
T2 ( x, y)  2  4( x  1)  ( y  2)  2( x  1)2  2( x  1)( y  2) .
☼

Example 11: Using both a linear and a quadratic approximation, estimate the difference
in the volume between a box with a square base of side length 1m and height 2m and a
box with square base of side length 1.1m and height 2.05m.
If we let the side length of the base of a box be x and the height be y then volume
V of the box is given by the formula
V ( x, y)  x 2 y .
Thus the difference in the volume between the boxes will be the change in V when x
changes by 0.1 and y changes by 0.05. Using the results obtained in the example above,
via a linear approximation

V  4x  y  4  0.1  0.05  0.45 .

Via a quadratic approximation

V  4x  y  2x2  2xy  4  0.1  0.05  2  0.12  2  0.1 0.05  0.48

☼

Example Tasks

f ( x, y)  e x  y
2
ET 1:* Find the Taylor polynomial of degree 2 for about
( x, y)  (0,0) .

ET 2:* Find the Taylor polynomial of degree 2 for f ( x, y)  xy about

( x, y)  (2,3) .

ET 3:* If x  10  0.5 , y  15  0.05 use a linear approximation and a quadratic

approximation to find the value of the dependent variable z and an associated
error bound when
z  y ln( x)

Page 54